diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,448033 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.548872180451128, + "eval_steps": 500, + "global_step": 64000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 6.015037593984962e-05, + "grad_norm": 1.25, + "learning_rate": 6.000000000000001e-07, + "loss": 2.2776, + "step": 1 + }, + { + "epoch": 0.00012030075187969925, + "grad_norm": 1.3203125, + "learning_rate": 1.2000000000000002e-06, + "loss": 2.2798, + "step": 2 + }, + { + "epoch": 0.00018045112781954887, + "grad_norm": 1.421875, + "learning_rate": 1.8e-06, + "loss": 2.2798, + "step": 3 + }, + { + "epoch": 0.0002406015037593985, + "grad_norm": 1.3046875, + "learning_rate": 2.4000000000000003e-06, + "loss": 2.2462, + "step": 4 + }, + { + "epoch": 0.0003007518796992481, + "grad_norm": 1.5078125, + "learning_rate": 3e-06, + "loss": 2.2743, + "step": 5 + }, + { + "epoch": 0.00036090225563909775, + "grad_norm": 1.203125, + "learning_rate": 3.6e-06, + "loss": 2.3459, + "step": 6 + }, + { + "epoch": 0.0004210526315789474, + "grad_norm": 1.3046875, + "learning_rate": 4.2000000000000004e-06, + "loss": 2.243, + "step": 7 + }, + { + "epoch": 0.000481203007518797, + "grad_norm": 1.1953125, + "learning_rate": 4.800000000000001e-06, + "loss": 2.3092, + "step": 8 + }, + { + "epoch": 0.0005413533834586466, + "grad_norm": 1.2421875, + "learning_rate": 5.4e-06, + "loss": 2.2723, + "step": 9 + }, + { + "epoch": 0.0006015037593984962, + "grad_norm": 1.234375, + "learning_rate": 6e-06, + "loss": 2.3156, + "step": 10 + }, + { + "epoch": 0.0006616541353383459, + "grad_norm": 1.46875, + "learning_rate": 6.6e-06, + "loss": 2.272, + "step": 11 + }, + { + "epoch": 0.0007218045112781955, + "grad_norm": 1.34375, + "learning_rate": 7.2e-06, + "loss": 2.2152, + "step": 12 + }, + { + "epoch": 0.0007819548872180451, + "grad_norm": 1.1171875, + "learning_rate": 7.8e-06, + "loss": 2.2544, + "step": 13 + }, + { + "epoch": 0.0008421052631578948, + "grad_norm": 1.234375, + "learning_rate": 8.400000000000001e-06, + "loss": 2.2287, + "step": 14 + }, + { + "epoch": 0.0009022556390977444, + "grad_norm": 1.1796875, + "learning_rate": 9e-06, + "loss": 2.2287, + "step": 15 + }, + { + "epoch": 0.000962406015037594, + "grad_norm": 1.2421875, + "learning_rate": 9.600000000000001e-06, + "loss": 2.2078, + "step": 16 + }, + { + "epoch": 0.0010225563909774435, + "grad_norm": 1.2109375, + "learning_rate": 1.02e-05, + "loss": 2.2484, + "step": 17 + }, + { + "epoch": 0.0010827067669172932, + "grad_norm": 1.1171875, + "learning_rate": 1.08e-05, + "loss": 2.2406, + "step": 18 + }, + { + "epoch": 0.001142857142857143, + "grad_norm": 1.1015625, + "learning_rate": 1.1400000000000001e-05, + "loss": 2.2305, + "step": 19 + }, + { + "epoch": 0.0012030075187969924, + "grad_norm": 1.09375, + "learning_rate": 1.2e-05, + "loss": 2.1616, + "step": 20 + }, + { + "epoch": 0.0012631578947368421, + "grad_norm": 0.96875, + "learning_rate": 1.26e-05, + "loss": 2.1752, + "step": 21 + }, + { + "epoch": 0.0013233082706766918, + "grad_norm": 0.72265625, + "learning_rate": 1.32e-05, + "loss": 2.1882, + "step": 22 + }, + { + "epoch": 0.0013834586466165413, + "grad_norm": 0.92578125, + "learning_rate": 1.3800000000000002e-05, + "loss": 2.1026, + "step": 23 + }, + { + "epoch": 0.001443609022556391, + "grad_norm": 0.828125, + "learning_rate": 1.44e-05, + "loss": 2.0831, + "step": 24 + }, + { + "epoch": 0.0015037593984962407, + "grad_norm": 0.70703125, + "learning_rate": 1.5e-05, + "loss": 2.1132, + "step": 25 + }, + { + "epoch": 0.0015639097744360902, + "grad_norm": 0.66015625, + "learning_rate": 1.56e-05, + "loss": 2.1031, + "step": 26 + }, + { + "epoch": 0.0016240601503759399, + "grad_norm": 0.51953125, + "learning_rate": 1.62e-05, + "loss": 2.0964, + "step": 27 + }, + { + "epoch": 0.0016842105263157896, + "grad_norm": 0.5078125, + "learning_rate": 1.6800000000000002e-05, + "loss": 2.1556, + "step": 28 + }, + { + "epoch": 0.001744360902255639, + "grad_norm": 0.34765625, + "learning_rate": 1.74e-05, + "loss": 2.1243, + "step": 29 + }, + { + "epoch": 0.0018045112781954887, + "grad_norm": 0.427734375, + "learning_rate": 1.8e-05, + "loss": 2.0181, + "step": 30 + }, + { + "epoch": 0.0018646616541353384, + "grad_norm": 0.5078125, + "learning_rate": 1.86e-05, + "loss": 2.0345, + "step": 31 + }, + { + "epoch": 0.001924812030075188, + "grad_norm": 0.361328125, + "learning_rate": 1.9200000000000003e-05, + "loss": 2.0882, + "step": 32 + }, + { + "epoch": 0.0019849624060150374, + "grad_norm": 0.4921875, + "learning_rate": 1.98e-05, + "loss": 2.1383, + "step": 33 + }, + { + "epoch": 0.002045112781954887, + "grad_norm": 0.51953125, + "learning_rate": 2.04e-05, + "loss": 2.017, + "step": 34 + }, + { + "epoch": 0.002105263157894737, + "grad_norm": 0.6484375, + "learning_rate": 2.1e-05, + "loss": 2.0846, + "step": 35 + }, + { + "epoch": 0.0021654135338345865, + "grad_norm": 0.57421875, + "learning_rate": 2.16e-05, + "loss": 2.0597, + "step": 36 + }, + { + "epoch": 0.002225563909774436, + "grad_norm": 0.53515625, + "learning_rate": 2.22e-05, + "loss": 1.961, + "step": 37 + }, + { + "epoch": 0.002285714285714286, + "grad_norm": 0.51953125, + "learning_rate": 2.2800000000000002e-05, + "loss": 2.0369, + "step": 38 + }, + { + "epoch": 0.002345864661654135, + "grad_norm": 0.42578125, + "learning_rate": 2.3400000000000003e-05, + "loss": 2.0801, + "step": 39 + }, + { + "epoch": 0.002406015037593985, + "grad_norm": 0.482421875, + "learning_rate": 2.4e-05, + "loss": 2.0794, + "step": 40 + }, + { + "epoch": 0.0024661654135338345, + "grad_norm": 0.32421875, + "learning_rate": 2.4599999999999998e-05, + "loss": 2.0745, + "step": 41 + }, + { + "epoch": 0.0025263157894736842, + "grad_norm": 0.2421875, + "learning_rate": 2.52e-05, + "loss": 2.017, + "step": 42 + }, + { + "epoch": 0.002586466165413534, + "grad_norm": 0.32421875, + "learning_rate": 2.58e-05, + "loss": 1.9946, + "step": 43 + }, + { + "epoch": 0.0026466165413533836, + "grad_norm": 0.236328125, + "learning_rate": 2.64e-05, + "loss": 1.9838, + "step": 44 + }, + { + "epoch": 0.002706766917293233, + "grad_norm": 0.26953125, + "learning_rate": 2.7000000000000002e-05, + "loss": 1.9634, + "step": 45 + }, + { + "epoch": 0.0027669172932330826, + "grad_norm": 0.337890625, + "learning_rate": 2.7600000000000003e-05, + "loss": 2.0195, + "step": 46 + }, + { + "epoch": 0.0028270676691729323, + "grad_norm": 0.306640625, + "learning_rate": 2.8199999999999998e-05, + "loss": 1.9705, + "step": 47 + }, + { + "epoch": 0.002887218045112782, + "grad_norm": 0.4453125, + "learning_rate": 2.88e-05, + "loss": 1.9827, + "step": 48 + }, + { + "epoch": 0.0029473684210526317, + "grad_norm": 0.314453125, + "learning_rate": 2.94e-05, + "loss": 2.0132, + "step": 49 + }, + { + "epoch": 0.0030075187969924814, + "grad_norm": 0.28515625, + "learning_rate": 3e-05, + "loss": 2.0447, + "step": 50 + }, + { + "epoch": 0.0030676691729323306, + "grad_norm": 0.3359375, + "learning_rate": 3e-05, + "loss": 1.9493, + "step": 51 + }, + { + "epoch": 0.0031278195488721803, + "grad_norm": 0.271484375, + "learning_rate": 3e-05, + "loss": 1.9729, + "step": 52 + }, + { + "epoch": 0.00318796992481203, + "grad_norm": 0.21484375, + "learning_rate": 3e-05, + "loss": 1.9715, + "step": 53 + }, + { + "epoch": 0.0032481203007518797, + "grad_norm": 0.29296875, + "learning_rate": 3e-05, + "loss": 2.0652, + "step": 54 + }, + { + "epoch": 0.0033082706766917294, + "grad_norm": 0.23828125, + "learning_rate": 3e-05, + "loss": 2.056, + "step": 55 + }, + { + "epoch": 0.003368421052631579, + "grad_norm": 0.2294921875, + "learning_rate": 3e-05, + "loss": 1.9081, + "step": 56 + }, + { + "epoch": 0.0034285714285714284, + "grad_norm": 0.25390625, + "learning_rate": 3e-05, + "loss": 1.9679, + "step": 57 + }, + { + "epoch": 0.003488721804511278, + "grad_norm": 0.2236328125, + "learning_rate": 3e-05, + "loss": 1.9523, + "step": 58 + }, + { + "epoch": 0.003548872180451128, + "grad_norm": 0.2373046875, + "learning_rate": 3e-05, + "loss": 1.9232, + "step": 59 + }, + { + "epoch": 0.0036090225563909775, + "grad_norm": 0.1884765625, + "learning_rate": 3e-05, + "loss": 1.9294, + "step": 60 + }, + { + "epoch": 0.003669172932330827, + "grad_norm": 0.353515625, + "learning_rate": 3e-05, + "loss": 1.9844, + "step": 61 + }, + { + "epoch": 0.003729323308270677, + "grad_norm": 0.2197265625, + "learning_rate": 3e-05, + "loss": 1.9354, + "step": 62 + }, + { + "epoch": 0.003789473684210526, + "grad_norm": 0.28125, + "learning_rate": 3e-05, + "loss": 1.9713, + "step": 63 + }, + { + "epoch": 0.003849624060150376, + "grad_norm": 0.1962890625, + "learning_rate": 3e-05, + "loss": 1.9175, + "step": 64 + }, + { + "epoch": 0.003909774436090226, + "grad_norm": 0.263671875, + "learning_rate": 3e-05, + "loss": 1.9055, + "step": 65 + }, + { + "epoch": 0.003969924812030075, + "grad_norm": 0.2265625, + "learning_rate": 3e-05, + "loss": 2.0434, + "step": 66 + }, + { + "epoch": 0.0040300751879699245, + "grad_norm": 0.2080078125, + "learning_rate": 3e-05, + "loss": 1.9329, + "step": 67 + }, + { + "epoch": 0.004090225563909774, + "grad_norm": 0.27734375, + "learning_rate": 3e-05, + "loss": 1.9121, + "step": 68 + }, + { + "epoch": 0.004150375939849624, + "grad_norm": 0.251953125, + "learning_rate": 3e-05, + "loss": 1.9155, + "step": 69 + }, + { + "epoch": 0.004210526315789474, + "grad_norm": 0.1982421875, + "learning_rate": 3e-05, + "loss": 1.9192, + "step": 70 + }, + { + "epoch": 0.004270676691729323, + "grad_norm": 0.24609375, + "learning_rate": 3e-05, + "loss": 1.9565, + "step": 71 + }, + { + "epoch": 0.004330827067669173, + "grad_norm": 0.23046875, + "learning_rate": 3e-05, + "loss": 1.8499, + "step": 72 + }, + { + "epoch": 0.004390977443609023, + "grad_norm": 0.2119140625, + "learning_rate": 3e-05, + "loss": 1.9604, + "step": 73 + }, + { + "epoch": 0.004451127819548872, + "grad_norm": 0.1953125, + "learning_rate": 3e-05, + "loss": 1.9066, + "step": 74 + }, + { + "epoch": 0.004511278195488722, + "grad_norm": 0.1826171875, + "learning_rate": 3e-05, + "loss": 1.9346, + "step": 75 + }, + { + "epoch": 0.004571428571428572, + "grad_norm": 0.2236328125, + "learning_rate": 3e-05, + "loss": 1.9166, + "step": 76 + }, + { + "epoch": 0.0046315789473684215, + "grad_norm": 0.283203125, + "learning_rate": 3e-05, + "loss": 1.9411, + "step": 77 + }, + { + "epoch": 0.00469172932330827, + "grad_norm": 0.2216796875, + "learning_rate": 3e-05, + "loss": 1.9147, + "step": 78 + }, + { + "epoch": 0.00475187969924812, + "grad_norm": 0.193359375, + "learning_rate": 3e-05, + "loss": 1.9052, + "step": 79 + }, + { + "epoch": 0.00481203007518797, + "grad_norm": 0.2314453125, + "learning_rate": 3e-05, + "loss": 1.89, + "step": 80 + }, + { + "epoch": 0.004872180451127819, + "grad_norm": 0.28515625, + "learning_rate": 3e-05, + "loss": 1.8953, + "step": 81 + }, + { + "epoch": 0.004932330827067669, + "grad_norm": 0.2275390625, + "learning_rate": 3e-05, + "loss": 1.9368, + "step": 82 + }, + { + "epoch": 0.004992481203007519, + "grad_norm": 0.2197265625, + "learning_rate": 3e-05, + "loss": 1.9272, + "step": 83 + }, + { + "epoch": 0.0050526315789473685, + "grad_norm": 0.205078125, + "learning_rate": 3e-05, + "loss": 1.8736, + "step": 84 + }, + { + "epoch": 0.005112781954887218, + "grad_norm": 0.1748046875, + "learning_rate": 3e-05, + "loss": 1.9185, + "step": 85 + }, + { + "epoch": 0.005172932330827068, + "grad_norm": 0.2265625, + "learning_rate": 3e-05, + "loss": 1.8789, + "step": 86 + }, + { + "epoch": 0.005233082706766918, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.7715, + "step": 87 + }, + { + "epoch": 0.005293233082706767, + "grad_norm": 0.150390625, + "learning_rate": 3e-05, + "loss": 1.8915, + "step": 88 + }, + { + "epoch": 0.005353383458646617, + "grad_norm": 0.2041015625, + "learning_rate": 3e-05, + "loss": 1.9373, + "step": 89 + }, + { + "epoch": 0.005413533834586466, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.9004, + "step": 90 + }, + { + "epoch": 0.0054736842105263155, + "grad_norm": 0.2041015625, + "learning_rate": 3e-05, + "loss": 1.9075, + "step": 91 + }, + { + "epoch": 0.005533834586466165, + "grad_norm": 0.232421875, + "learning_rate": 3e-05, + "loss": 1.8594, + "step": 92 + }, + { + "epoch": 0.005593984962406015, + "grad_norm": 0.21875, + "learning_rate": 3e-05, + "loss": 1.8501, + "step": 93 + }, + { + "epoch": 0.005654135338345865, + "grad_norm": 0.22265625, + "learning_rate": 3e-05, + "loss": 1.8639, + "step": 94 + }, + { + "epoch": 0.005714285714285714, + "grad_norm": 0.189453125, + "learning_rate": 3e-05, + "loss": 1.8549, + "step": 95 + }, + { + "epoch": 0.005774436090225564, + "grad_norm": 0.25, + "learning_rate": 3e-05, + "loss": 1.8578, + "step": 96 + }, + { + "epoch": 0.005834586466165414, + "grad_norm": 0.287109375, + "learning_rate": 3e-05, + "loss": 1.7607, + "step": 97 + }, + { + "epoch": 0.005894736842105263, + "grad_norm": 0.2275390625, + "learning_rate": 3e-05, + "loss": 1.9212, + "step": 98 + }, + { + "epoch": 0.005954887218045113, + "grad_norm": 0.283203125, + "learning_rate": 3e-05, + "loss": 1.9041, + "step": 99 + }, + { + "epoch": 0.006015037593984963, + "grad_norm": 0.173828125, + "learning_rate": 3e-05, + "loss": 1.8672, + "step": 100 + }, + { + "epoch": 0.0060751879699248125, + "grad_norm": 0.267578125, + "learning_rate": 3e-05, + "loss": 1.8691, + "step": 101 + }, + { + "epoch": 0.006135338345864661, + "grad_norm": 0.244140625, + "learning_rate": 3e-05, + "loss": 1.9092, + "step": 102 + }, + { + "epoch": 0.006195488721804511, + "grad_norm": 0.263671875, + "learning_rate": 3e-05, + "loss": 1.8306, + "step": 103 + }, + { + "epoch": 0.006255639097744361, + "grad_norm": 0.2119140625, + "learning_rate": 3e-05, + "loss": 1.8957, + "step": 104 + }, + { + "epoch": 0.00631578947368421, + "grad_norm": 0.2314453125, + "learning_rate": 3e-05, + "loss": 1.8819, + "step": 105 + }, + { + "epoch": 0.00637593984962406, + "grad_norm": 0.21875, + "learning_rate": 3e-05, + "loss": 1.8803, + "step": 106 + }, + { + "epoch": 0.00643609022556391, + "grad_norm": 0.248046875, + "learning_rate": 3e-05, + "loss": 1.8567, + "step": 107 + }, + { + "epoch": 0.0064962406015037595, + "grad_norm": 0.298828125, + "learning_rate": 3e-05, + "loss": 1.9232, + "step": 108 + }, + { + "epoch": 0.006556390977443609, + "grad_norm": 0.1943359375, + "learning_rate": 3e-05, + "loss": 1.8756, + "step": 109 + }, + { + "epoch": 0.006616541353383459, + "grad_norm": 0.21484375, + "learning_rate": 3e-05, + "loss": 1.8778, + "step": 110 + }, + { + "epoch": 0.0066766917293233086, + "grad_norm": 0.2578125, + "learning_rate": 3e-05, + "loss": 1.8253, + "step": 111 + }, + { + "epoch": 0.006736842105263158, + "grad_norm": 0.236328125, + "learning_rate": 3e-05, + "loss": 1.8727, + "step": 112 + }, + { + "epoch": 0.006796992481203007, + "grad_norm": 0.2412109375, + "learning_rate": 3e-05, + "loss": 1.8098, + "step": 113 + }, + { + "epoch": 0.006857142857142857, + "grad_norm": 0.19140625, + "learning_rate": 3e-05, + "loss": 1.8305, + "step": 114 + }, + { + "epoch": 0.0069172932330827065, + "grad_norm": 0.33203125, + "learning_rate": 3e-05, + "loss": 1.8331, + "step": 115 + }, + { + "epoch": 0.006977443609022556, + "grad_norm": 0.205078125, + "learning_rate": 3e-05, + "loss": 1.859, + "step": 116 + }, + { + "epoch": 0.007037593984962406, + "grad_norm": 0.244140625, + "learning_rate": 3e-05, + "loss": 1.9831, + "step": 117 + }, + { + "epoch": 0.007097744360902256, + "grad_norm": 0.259765625, + "learning_rate": 3e-05, + "loss": 1.8687, + "step": 118 + }, + { + "epoch": 0.007157894736842105, + "grad_norm": 0.1787109375, + "learning_rate": 3e-05, + "loss": 1.8164, + "step": 119 + }, + { + "epoch": 0.007218045112781955, + "grad_norm": 0.2421875, + "learning_rate": 3e-05, + "loss": 1.8327, + "step": 120 + }, + { + "epoch": 0.007278195488721805, + "grad_norm": 0.25390625, + "learning_rate": 3e-05, + "loss": 1.8048, + "step": 121 + }, + { + "epoch": 0.007338345864661654, + "grad_norm": 0.1787109375, + "learning_rate": 3e-05, + "loss": 1.8108, + "step": 122 + }, + { + "epoch": 0.007398496240601504, + "grad_norm": 0.287109375, + "learning_rate": 3e-05, + "loss": 1.8337, + "step": 123 + }, + { + "epoch": 0.007458646616541354, + "grad_norm": 0.244140625, + "learning_rate": 3e-05, + "loss": 1.8215, + "step": 124 + }, + { + "epoch": 0.007518796992481203, + "grad_norm": 0.2197265625, + "learning_rate": 3e-05, + "loss": 1.8629, + "step": 125 + }, + { + "epoch": 0.007578947368421052, + "grad_norm": 0.2236328125, + "learning_rate": 3e-05, + "loss": 1.795, + "step": 126 + }, + { + "epoch": 0.007639097744360902, + "grad_norm": 0.17578125, + "learning_rate": 3e-05, + "loss": 1.8196, + "step": 127 + }, + { + "epoch": 0.007699248120300752, + "grad_norm": 0.228515625, + "learning_rate": 3e-05, + "loss": 1.8427, + "step": 128 + }, + { + "epoch": 0.007759398496240601, + "grad_norm": 0.27734375, + "learning_rate": 3e-05, + "loss": 1.7926, + "step": 129 + }, + { + "epoch": 0.007819548872180452, + "grad_norm": 0.240234375, + "learning_rate": 3e-05, + "loss": 1.8711, + "step": 130 + }, + { + "epoch": 0.0078796992481203, + "grad_norm": 0.271484375, + "learning_rate": 3e-05, + "loss": 1.7993, + "step": 131 + }, + { + "epoch": 0.00793984962406015, + "grad_norm": 0.2021484375, + "learning_rate": 3e-05, + "loss": 1.8236, + "step": 132 + }, + { + "epoch": 0.008, + "grad_norm": 0.1943359375, + "learning_rate": 3e-05, + "loss": 1.8404, + "step": 133 + }, + { + "epoch": 0.008060150375939849, + "grad_norm": 0.2080078125, + "learning_rate": 3e-05, + "loss": 1.8235, + "step": 134 + }, + { + "epoch": 0.0081203007518797, + "grad_norm": 0.251953125, + "learning_rate": 3e-05, + "loss": 1.8123, + "step": 135 + }, + { + "epoch": 0.008180451127819548, + "grad_norm": 0.2255859375, + "learning_rate": 3e-05, + "loss": 1.8288, + "step": 136 + }, + { + "epoch": 0.008240601503759399, + "grad_norm": 0.2392578125, + "learning_rate": 3e-05, + "loss": 1.8401, + "step": 137 + }, + { + "epoch": 0.008300751879699248, + "grad_norm": 0.25390625, + "learning_rate": 3e-05, + "loss": 1.8826, + "step": 138 + }, + { + "epoch": 0.008360902255639098, + "grad_norm": 0.2138671875, + "learning_rate": 3e-05, + "loss": 1.7892, + "step": 139 + }, + { + "epoch": 0.008421052631578947, + "grad_norm": 0.20703125, + "learning_rate": 3e-05, + "loss": 1.805, + "step": 140 + }, + { + "epoch": 0.008481203007518798, + "grad_norm": 0.263671875, + "learning_rate": 3e-05, + "loss": 1.8363, + "step": 141 + }, + { + "epoch": 0.008541353383458647, + "grad_norm": 0.2001953125, + "learning_rate": 3e-05, + "loss": 1.8105, + "step": 142 + }, + { + "epoch": 0.008601503759398495, + "grad_norm": 0.28125, + "learning_rate": 3e-05, + "loss": 1.8415, + "step": 143 + }, + { + "epoch": 0.008661654135338346, + "grad_norm": 0.193359375, + "learning_rate": 3e-05, + "loss": 1.8107, + "step": 144 + }, + { + "epoch": 0.008721804511278195, + "grad_norm": 0.2451171875, + "learning_rate": 3e-05, + "loss": 1.8503, + "step": 145 + }, + { + "epoch": 0.008781954887218045, + "grad_norm": 0.1708984375, + "learning_rate": 3e-05, + "loss": 1.8049, + "step": 146 + }, + { + "epoch": 0.008842105263157894, + "grad_norm": 0.2314453125, + "learning_rate": 3e-05, + "loss": 1.8397, + "step": 147 + }, + { + "epoch": 0.008902255639097745, + "grad_norm": 0.2890625, + "learning_rate": 3e-05, + "loss": 1.7222, + "step": 148 + }, + { + "epoch": 0.008962406015037594, + "grad_norm": 0.2099609375, + "learning_rate": 3e-05, + "loss": 1.8182, + "step": 149 + }, + { + "epoch": 0.009022556390977444, + "grad_norm": 0.26171875, + "learning_rate": 3e-05, + "loss": 1.8216, + "step": 150 + }, + { + "epoch": 0.009082706766917293, + "grad_norm": 0.224609375, + "learning_rate": 3e-05, + "loss": 1.8333, + "step": 151 + }, + { + "epoch": 0.009142857142857144, + "grad_norm": 0.1884765625, + "learning_rate": 3e-05, + "loss": 1.7936, + "step": 152 + }, + { + "epoch": 0.009203007518796992, + "grad_norm": 0.28515625, + "learning_rate": 3e-05, + "loss": 1.7836, + "step": 153 + }, + { + "epoch": 0.009263157894736843, + "grad_norm": 0.2216796875, + "learning_rate": 3e-05, + "loss": 1.8154, + "step": 154 + }, + { + "epoch": 0.009323308270676692, + "grad_norm": 0.34765625, + "learning_rate": 3e-05, + "loss": 1.8469, + "step": 155 + }, + { + "epoch": 0.00938345864661654, + "grad_norm": 0.283203125, + "learning_rate": 3e-05, + "loss": 1.8744, + "step": 156 + }, + { + "epoch": 0.009443609022556391, + "grad_norm": 0.302734375, + "learning_rate": 3e-05, + "loss": 1.7953, + "step": 157 + }, + { + "epoch": 0.00950375939849624, + "grad_norm": 0.2119140625, + "learning_rate": 3e-05, + "loss": 1.7618, + "step": 158 + }, + { + "epoch": 0.00956390977443609, + "grad_norm": 0.267578125, + "learning_rate": 3e-05, + "loss": 1.8207, + "step": 159 + }, + { + "epoch": 0.00962406015037594, + "grad_norm": 0.291015625, + "learning_rate": 3e-05, + "loss": 1.8043, + "step": 160 + }, + { + "epoch": 0.00968421052631579, + "grad_norm": 0.2294921875, + "learning_rate": 3e-05, + "loss": 1.7941, + "step": 161 + }, + { + "epoch": 0.009744360902255639, + "grad_norm": 0.1904296875, + "learning_rate": 3e-05, + "loss": 1.8011, + "step": 162 + }, + { + "epoch": 0.00980451127819549, + "grad_norm": 0.2099609375, + "learning_rate": 3e-05, + "loss": 1.8587, + "step": 163 + }, + { + "epoch": 0.009864661654135338, + "grad_norm": 0.263671875, + "learning_rate": 3e-05, + "loss": 1.8238, + "step": 164 + }, + { + "epoch": 0.009924812030075189, + "grad_norm": 0.2412109375, + "learning_rate": 3e-05, + "loss": 1.8004, + "step": 165 + }, + { + "epoch": 0.009984962406015038, + "grad_norm": 0.2275390625, + "learning_rate": 3e-05, + "loss": 1.7581, + "step": 166 + }, + { + "epoch": 0.010045112781954886, + "grad_norm": 0.271484375, + "learning_rate": 3e-05, + "loss": 1.8197, + "step": 167 + }, + { + "epoch": 0.010105263157894737, + "grad_norm": 0.17578125, + "learning_rate": 3e-05, + "loss": 1.7919, + "step": 168 + }, + { + "epoch": 0.010165413533834586, + "grad_norm": 0.20703125, + "learning_rate": 3e-05, + "loss": 1.7977, + "step": 169 + }, + { + "epoch": 0.010225563909774436, + "grad_norm": 0.23828125, + "learning_rate": 3e-05, + "loss": 1.7916, + "step": 170 + }, + { + "epoch": 0.010285714285714285, + "grad_norm": 0.228515625, + "learning_rate": 3e-05, + "loss": 1.7808, + "step": 171 + }, + { + "epoch": 0.010345864661654136, + "grad_norm": 0.333984375, + "learning_rate": 3e-05, + "loss": 1.7565, + "step": 172 + }, + { + "epoch": 0.010406015037593985, + "grad_norm": 0.251953125, + "learning_rate": 3e-05, + "loss": 1.7849, + "step": 173 + }, + { + "epoch": 0.010466165413533835, + "grad_norm": 0.267578125, + "learning_rate": 3e-05, + "loss": 1.7897, + "step": 174 + }, + { + "epoch": 0.010526315789473684, + "grad_norm": 0.2138671875, + "learning_rate": 3e-05, + "loss": 1.7833, + "step": 175 + }, + { + "epoch": 0.010586466165413535, + "grad_norm": 0.2412109375, + "learning_rate": 3e-05, + "loss": 1.8133, + "step": 176 + }, + { + "epoch": 0.010646616541353383, + "grad_norm": 0.2197265625, + "learning_rate": 3e-05, + "loss": 1.7686, + "step": 177 + }, + { + "epoch": 0.010706766917293234, + "grad_norm": 0.265625, + "learning_rate": 3e-05, + "loss": 1.7717, + "step": 178 + }, + { + "epoch": 0.010766917293233083, + "grad_norm": 0.23828125, + "learning_rate": 3e-05, + "loss": 1.8137, + "step": 179 + }, + { + "epoch": 0.010827067669172932, + "grad_norm": 0.27734375, + "learning_rate": 3e-05, + "loss": 1.7506, + "step": 180 + }, + { + "epoch": 0.010887218045112782, + "grad_norm": 0.2412109375, + "learning_rate": 3e-05, + "loss": 1.7817, + "step": 181 + }, + { + "epoch": 0.010947368421052631, + "grad_norm": 0.255859375, + "learning_rate": 3e-05, + "loss": 1.7778, + "step": 182 + }, + { + "epoch": 0.011007518796992482, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.7745, + "step": 183 + }, + { + "epoch": 0.01106766917293233, + "grad_norm": 0.318359375, + "learning_rate": 3e-05, + "loss": 1.7618, + "step": 184 + }, + { + "epoch": 0.011127819548872181, + "grad_norm": 0.21875, + "learning_rate": 3e-05, + "loss": 1.7994, + "step": 185 + }, + { + "epoch": 0.01118796992481203, + "grad_norm": 0.2490234375, + "learning_rate": 3e-05, + "loss": 1.8065, + "step": 186 + }, + { + "epoch": 0.01124812030075188, + "grad_norm": 0.2255859375, + "learning_rate": 3e-05, + "loss": 1.7646, + "step": 187 + }, + { + "epoch": 0.01130827067669173, + "grad_norm": 0.302734375, + "learning_rate": 3e-05, + "loss": 1.7319, + "step": 188 + }, + { + "epoch": 0.01136842105263158, + "grad_norm": 0.265625, + "learning_rate": 3e-05, + "loss": 1.7873, + "step": 189 + }, + { + "epoch": 0.011428571428571429, + "grad_norm": 0.267578125, + "learning_rate": 3e-05, + "loss": 1.7828, + "step": 190 + }, + { + "epoch": 0.011488721804511277, + "grad_norm": 0.22265625, + "learning_rate": 3e-05, + "loss": 1.7718, + "step": 191 + }, + { + "epoch": 0.011548872180451128, + "grad_norm": 0.29296875, + "learning_rate": 3e-05, + "loss": 1.7422, + "step": 192 + }, + { + "epoch": 0.011609022556390977, + "grad_norm": 0.275390625, + "learning_rate": 3e-05, + "loss": 1.7189, + "step": 193 + }, + { + "epoch": 0.011669172932330827, + "grad_norm": 0.251953125, + "learning_rate": 3e-05, + "loss": 1.7502, + "step": 194 + }, + { + "epoch": 0.011729323308270676, + "grad_norm": 0.201171875, + "learning_rate": 3e-05, + "loss": 1.7095, + "step": 195 + }, + { + "epoch": 0.011789473684210527, + "grad_norm": 0.2431640625, + "learning_rate": 3e-05, + "loss": 1.7606, + "step": 196 + }, + { + "epoch": 0.011849624060150376, + "grad_norm": 0.240234375, + "learning_rate": 3e-05, + "loss": 1.8027, + "step": 197 + }, + { + "epoch": 0.011909774436090226, + "grad_norm": 0.2314453125, + "learning_rate": 3e-05, + "loss": 1.8044, + "step": 198 + }, + { + "epoch": 0.011969924812030075, + "grad_norm": 0.32421875, + "learning_rate": 3e-05, + "loss": 1.751, + "step": 199 + }, + { + "epoch": 0.012030075187969926, + "grad_norm": 0.26171875, + "learning_rate": 3e-05, + "loss": 1.7597, + "step": 200 + }, + { + "epoch": 0.012090225563909774, + "grad_norm": 0.25, + "learning_rate": 3e-05, + "loss": 1.7558, + "step": 201 + }, + { + "epoch": 0.012150375939849625, + "grad_norm": 0.2177734375, + "learning_rate": 3e-05, + "loss": 1.7819, + "step": 202 + }, + { + "epoch": 0.012210526315789474, + "grad_norm": 0.2236328125, + "learning_rate": 3e-05, + "loss": 1.7393, + "step": 203 + }, + { + "epoch": 0.012270676691729323, + "grad_norm": 0.28515625, + "learning_rate": 3e-05, + "loss": 1.789, + "step": 204 + }, + { + "epoch": 0.012330827067669173, + "grad_norm": 0.28125, + "learning_rate": 3e-05, + "loss": 1.6968, + "step": 205 + }, + { + "epoch": 0.012390977443609022, + "grad_norm": 0.203125, + "learning_rate": 3e-05, + "loss": 1.7387, + "step": 206 + }, + { + "epoch": 0.012451127819548873, + "grad_norm": 0.287109375, + "learning_rate": 3e-05, + "loss": 1.7594, + "step": 207 + }, + { + "epoch": 0.012511278195488721, + "grad_norm": 0.2333984375, + "learning_rate": 3e-05, + "loss": 1.7599, + "step": 208 + }, + { + "epoch": 0.012571428571428572, + "grad_norm": 0.2041015625, + "learning_rate": 3e-05, + "loss": 1.7407, + "step": 209 + }, + { + "epoch": 0.01263157894736842, + "grad_norm": 0.2265625, + "learning_rate": 3e-05, + "loss": 1.7307, + "step": 210 + }, + { + "epoch": 0.012691729323308271, + "grad_norm": 0.29296875, + "learning_rate": 3e-05, + "loss": 1.7407, + "step": 211 + }, + { + "epoch": 0.01275187969924812, + "grad_norm": 0.25390625, + "learning_rate": 3e-05, + "loss": 1.8118, + "step": 212 + }, + { + "epoch": 0.01281203007518797, + "grad_norm": 0.2578125, + "learning_rate": 3e-05, + "loss": 1.6475, + "step": 213 + }, + { + "epoch": 0.01287218045112782, + "grad_norm": 0.291015625, + "learning_rate": 3e-05, + "loss": 1.7583, + "step": 214 + }, + { + "epoch": 0.012932330827067668, + "grad_norm": 0.271484375, + "learning_rate": 3e-05, + "loss": 1.7533, + "step": 215 + }, + { + "epoch": 0.012992481203007519, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.9151, + "step": 216 + }, + { + "epoch": 0.013052631578947368, + "grad_norm": 0.318359375, + "learning_rate": 3e-05, + "loss": 1.7442, + "step": 217 + }, + { + "epoch": 0.013112781954887218, + "grad_norm": 0.251953125, + "learning_rate": 3e-05, + "loss": 1.7334, + "step": 218 + }, + { + "epoch": 0.013172932330827067, + "grad_norm": 0.2353515625, + "learning_rate": 3e-05, + "loss": 1.7729, + "step": 219 + }, + { + "epoch": 0.013233082706766918, + "grad_norm": 0.1943359375, + "learning_rate": 3e-05, + "loss": 1.7158, + "step": 220 + }, + { + "epoch": 0.013293233082706767, + "grad_norm": 0.234375, + "learning_rate": 3e-05, + "loss": 1.7036, + "step": 221 + }, + { + "epoch": 0.013353383458646617, + "grad_norm": 0.33984375, + "learning_rate": 3e-05, + "loss": 1.7715, + "step": 222 + }, + { + "epoch": 0.013413533834586466, + "grad_norm": 0.25, + "learning_rate": 3e-05, + "loss": 1.7, + "step": 223 + }, + { + "epoch": 0.013473684210526317, + "grad_norm": 0.302734375, + "learning_rate": 3e-05, + "loss": 1.7707, + "step": 224 + }, + { + "epoch": 0.013533834586466165, + "grad_norm": 0.248046875, + "learning_rate": 3e-05, + "loss": 1.7638, + "step": 225 + }, + { + "epoch": 0.013593984962406014, + "grad_norm": 0.341796875, + "learning_rate": 3e-05, + "loss": 1.7372, + "step": 226 + }, + { + "epoch": 0.013654135338345865, + "grad_norm": 0.2353515625, + "learning_rate": 3e-05, + "loss": 1.6875, + "step": 227 + }, + { + "epoch": 0.013714285714285714, + "grad_norm": 0.265625, + "learning_rate": 3e-05, + "loss": 1.7194, + "step": 228 + }, + { + "epoch": 0.013774436090225564, + "grad_norm": 0.228515625, + "learning_rate": 3e-05, + "loss": 1.7732, + "step": 229 + }, + { + "epoch": 0.013834586466165413, + "grad_norm": 0.296875, + "learning_rate": 3e-05, + "loss": 1.7756, + "step": 230 + }, + { + "epoch": 0.013894736842105264, + "grad_norm": 0.25390625, + "learning_rate": 3e-05, + "loss": 1.7299, + "step": 231 + }, + { + "epoch": 0.013954887218045112, + "grad_norm": 0.2255859375, + "learning_rate": 3e-05, + "loss": 1.7446, + "step": 232 + }, + { + "epoch": 0.014015037593984963, + "grad_norm": 0.255859375, + "learning_rate": 3e-05, + "loss": 1.7556, + "step": 233 + }, + { + "epoch": 0.014075187969924812, + "grad_norm": 0.2578125, + "learning_rate": 3e-05, + "loss": 1.7291, + "step": 234 + }, + { + "epoch": 0.014135338345864662, + "grad_norm": 0.25, + "learning_rate": 3e-05, + "loss": 1.7251, + "step": 235 + }, + { + "epoch": 0.014195488721804511, + "grad_norm": 0.333984375, + "learning_rate": 3e-05, + "loss": 1.6971, + "step": 236 + }, + { + "epoch": 0.014255639097744362, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.8038, + "step": 237 + }, + { + "epoch": 0.01431578947368421, + "grad_norm": 0.2275390625, + "learning_rate": 3e-05, + "loss": 1.7253, + "step": 238 + }, + { + "epoch": 0.01437593984962406, + "grad_norm": 0.279296875, + "learning_rate": 3e-05, + "loss": 1.7841, + "step": 239 + }, + { + "epoch": 0.01443609022556391, + "grad_norm": 0.2080078125, + "learning_rate": 3e-05, + "loss": 1.756, + "step": 240 + }, + { + "epoch": 0.014496240601503759, + "grad_norm": 0.2353515625, + "learning_rate": 3e-05, + "loss": 1.7245, + "step": 241 + }, + { + "epoch": 0.01455639097744361, + "grad_norm": 0.2890625, + "learning_rate": 3e-05, + "loss": 1.7889, + "step": 242 + }, + { + "epoch": 0.014616541353383458, + "grad_norm": 0.31640625, + "learning_rate": 3e-05, + "loss": 1.7328, + "step": 243 + }, + { + "epoch": 0.014676691729323309, + "grad_norm": 0.2470703125, + "learning_rate": 3e-05, + "loss": 1.7135, + "step": 244 + }, + { + "epoch": 0.014736842105263158, + "grad_norm": 0.267578125, + "learning_rate": 3e-05, + "loss": 1.7802, + "step": 245 + }, + { + "epoch": 0.014796992481203008, + "grad_norm": 0.287109375, + "learning_rate": 3e-05, + "loss": 1.7414, + "step": 246 + }, + { + "epoch": 0.014857142857142857, + "grad_norm": 0.30078125, + "learning_rate": 3e-05, + "loss": 1.8049, + "step": 247 + }, + { + "epoch": 0.014917293233082708, + "grad_norm": 0.263671875, + "learning_rate": 3e-05, + "loss": 1.6679, + "step": 248 + }, + { + "epoch": 0.014977443609022556, + "grad_norm": 0.34375, + "learning_rate": 3e-05, + "loss": 1.6957, + "step": 249 + }, + { + "epoch": 0.015037593984962405, + "grad_norm": 0.306640625, + "learning_rate": 3e-05, + "loss": 1.6623, + "step": 250 + }, + { + "epoch": 0.015097744360902256, + "grad_norm": 0.37109375, + "learning_rate": 3e-05, + "loss": 1.7749, + "step": 251 + }, + { + "epoch": 0.015157894736842105, + "grad_norm": 0.283203125, + "learning_rate": 3e-05, + "loss": 1.7521, + "step": 252 + }, + { + "epoch": 0.015218045112781955, + "grad_norm": 0.375, + "learning_rate": 3e-05, + "loss": 1.6808, + "step": 253 + }, + { + "epoch": 0.015278195488721804, + "grad_norm": 0.2236328125, + "learning_rate": 3e-05, + "loss": 1.7248, + "step": 254 + }, + { + "epoch": 0.015338345864661655, + "grad_norm": 0.28515625, + "learning_rate": 3e-05, + "loss": 1.7642, + "step": 255 + }, + { + "epoch": 0.015398496240601503, + "grad_norm": 0.3515625, + "learning_rate": 3e-05, + "loss": 1.7051, + "step": 256 + }, + { + "epoch": 0.015458646616541354, + "grad_norm": 0.26171875, + "learning_rate": 3e-05, + "loss": 1.7672, + "step": 257 + }, + { + "epoch": 0.015518796992481203, + "grad_norm": 0.255859375, + "learning_rate": 3e-05, + "loss": 1.7489, + "step": 258 + }, + { + "epoch": 0.015578947368421053, + "grad_norm": 0.333984375, + "learning_rate": 3e-05, + "loss": 1.7027, + "step": 259 + }, + { + "epoch": 0.015639097744360904, + "grad_norm": 0.267578125, + "learning_rate": 3e-05, + "loss": 1.7285, + "step": 260 + }, + { + "epoch": 0.01569924812030075, + "grad_norm": 0.2578125, + "learning_rate": 3e-05, + "loss": 1.7465, + "step": 261 + }, + { + "epoch": 0.0157593984962406, + "grad_norm": 0.279296875, + "learning_rate": 3e-05, + "loss": 1.7669, + "step": 262 + }, + { + "epoch": 0.015819548872180452, + "grad_norm": 0.357421875, + "learning_rate": 3e-05, + "loss": 1.7315, + "step": 263 + }, + { + "epoch": 0.0158796992481203, + "grad_norm": 0.26953125, + "learning_rate": 3e-05, + "loss": 1.7238, + "step": 264 + }, + { + "epoch": 0.01593984962406015, + "grad_norm": 0.3046875, + "learning_rate": 3e-05, + "loss": 1.7508, + "step": 265 + }, + { + "epoch": 0.016, + "grad_norm": 0.328125, + "learning_rate": 3e-05, + "loss": 1.7337, + "step": 266 + }, + { + "epoch": 0.01606015037593985, + "grad_norm": 0.2470703125, + "learning_rate": 3e-05, + "loss": 1.7426, + "step": 267 + }, + { + "epoch": 0.016120300751879698, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.7853, + "step": 268 + }, + { + "epoch": 0.01618045112781955, + "grad_norm": 0.3046875, + "learning_rate": 3e-05, + "loss": 1.6809, + "step": 269 + }, + { + "epoch": 0.0162406015037594, + "grad_norm": 0.3203125, + "learning_rate": 3e-05, + "loss": 1.6284, + "step": 270 + }, + { + "epoch": 0.01630075187969925, + "grad_norm": 0.296875, + "learning_rate": 3e-05, + "loss": 1.7128, + "step": 271 + }, + { + "epoch": 0.016360902255639097, + "grad_norm": 0.337890625, + "learning_rate": 3e-05, + "loss": 1.7358, + "step": 272 + }, + { + "epoch": 0.016421052631578947, + "grad_norm": 0.263671875, + "learning_rate": 3e-05, + "loss": 1.6829, + "step": 273 + }, + { + "epoch": 0.016481203007518798, + "grad_norm": 0.3046875, + "learning_rate": 3e-05, + "loss": 1.7019, + "step": 274 + }, + { + "epoch": 0.016541353383458645, + "grad_norm": 0.2890625, + "learning_rate": 3e-05, + "loss": 1.6973, + "step": 275 + }, + { + "epoch": 0.016601503759398496, + "grad_norm": 0.341796875, + "learning_rate": 3e-05, + "loss": 1.7035, + "step": 276 + }, + { + "epoch": 0.016661654135338346, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.735, + "step": 277 + }, + { + "epoch": 0.016721804511278197, + "grad_norm": 0.2353515625, + "learning_rate": 3e-05, + "loss": 1.715, + "step": 278 + }, + { + "epoch": 0.016781954887218044, + "grad_norm": 0.2734375, + "learning_rate": 3e-05, + "loss": 1.7219, + "step": 279 + }, + { + "epoch": 0.016842105263157894, + "grad_norm": 0.271484375, + "learning_rate": 3e-05, + "loss": 1.7534, + "step": 280 + }, + { + "epoch": 0.016902255639097745, + "grad_norm": 0.28515625, + "learning_rate": 3e-05, + "loss": 1.7312, + "step": 281 + }, + { + "epoch": 0.016962406015037595, + "grad_norm": 0.384765625, + "learning_rate": 3e-05, + "loss": 1.7204, + "step": 282 + }, + { + "epoch": 0.017022556390977443, + "grad_norm": 0.30859375, + "learning_rate": 3e-05, + "loss": 1.7176, + "step": 283 + }, + { + "epoch": 0.017082706766917293, + "grad_norm": 0.234375, + "learning_rate": 3e-05, + "loss": 1.681, + "step": 284 + }, + { + "epoch": 0.017142857142857144, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.7672, + "step": 285 + }, + { + "epoch": 0.01720300751879699, + "grad_norm": 0.37109375, + "learning_rate": 3e-05, + "loss": 1.7319, + "step": 286 + }, + { + "epoch": 0.01726315789473684, + "grad_norm": 0.298828125, + "learning_rate": 3e-05, + "loss": 1.7019, + "step": 287 + }, + { + "epoch": 0.017323308270676692, + "grad_norm": 0.275390625, + "learning_rate": 3e-05, + "loss": 1.6849, + "step": 288 + }, + { + "epoch": 0.017383458646616542, + "grad_norm": 0.259765625, + "learning_rate": 3e-05, + "loss": 1.6997, + "step": 289 + }, + { + "epoch": 0.01744360902255639, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.7571, + "step": 290 + }, + { + "epoch": 0.01750375939849624, + "grad_norm": 0.279296875, + "learning_rate": 3e-05, + "loss": 1.7316, + "step": 291 + }, + { + "epoch": 0.01756390977443609, + "grad_norm": 0.34375, + "learning_rate": 3e-05, + "loss": 1.7039, + "step": 292 + }, + { + "epoch": 0.01762406015037594, + "grad_norm": 0.255859375, + "learning_rate": 3e-05, + "loss": 1.7279, + "step": 293 + }, + { + "epoch": 0.01768421052631579, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.6951, + "step": 294 + }, + { + "epoch": 0.01774436090225564, + "grad_norm": 0.228515625, + "learning_rate": 3e-05, + "loss": 1.7086, + "step": 295 + }, + { + "epoch": 0.01780451127819549, + "grad_norm": 0.29296875, + "learning_rate": 3e-05, + "loss": 1.6914, + "step": 296 + }, + { + "epoch": 0.01786466165413534, + "grad_norm": 0.28125, + "learning_rate": 3e-05, + "loss": 1.6988, + "step": 297 + }, + { + "epoch": 0.017924812030075187, + "grad_norm": 0.267578125, + "learning_rate": 3e-05, + "loss": 1.7392, + "step": 298 + }, + { + "epoch": 0.017984962406015038, + "grad_norm": 0.298828125, + "learning_rate": 3e-05, + "loss": 1.7332, + "step": 299 + }, + { + "epoch": 0.01804511278195489, + "grad_norm": 0.28125, + "learning_rate": 3e-05, + "loss": 1.6555, + "step": 300 + }, + { + "epoch": 0.018105263157894735, + "grad_norm": 0.34765625, + "learning_rate": 3e-05, + "loss": 1.6738, + "step": 301 + }, + { + "epoch": 0.018165413533834586, + "grad_norm": 0.3046875, + "learning_rate": 3e-05, + "loss": 1.7092, + "step": 302 + }, + { + "epoch": 0.018225563909774437, + "grad_norm": 0.376953125, + "learning_rate": 3e-05, + "loss": 1.6711, + "step": 303 + }, + { + "epoch": 0.018285714285714287, + "grad_norm": 0.337890625, + "learning_rate": 3e-05, + "loss": 1.6292, + "step": 304 + }, + { + "epoch": 0.018345864661654134, + "grad_norm": 0.251953125, + "learning_rate": 3e-05, + "loss": 1.7315, + "step": 305 + }, + { + "epoch": 0.018406015037593985, + "grad_norm": 0.33203125, + "learning_rate": 3e-05, + "loss": 1.724, + "step": 306 + }, + { + "epoch": 0.018466165413533835, + "grad_norm": 0.359375, + "learning_rate": 3e-05, + "loss": 1.7008, + "step": 307 + }, + { + "epoch": 0.018526315789473686, + "grad_norm": 0.396484375, + "learning_rate": 3e-05, + "loss": 1.664, + "step": 308 + }, + { + "epoch": 0.018586466165413533, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.6987, + "step": 309 + }, + { + "epoch": 0.018646616541353384, + "grad_norm": 0.30859375, + "learning_rate": 3e-05, + "loss": 1.6894, + "step": 310 + }, + { + "epoch": 0.018706766917293234, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.7064, + "step": 311 + }, + { + "epoch": 0.01876691729323308, + "grad_norm": 0.27734375, + "learning_rate": 3e-05, + "loss": 1.7065, + "step": 312 + }, + { + "epoch": 0.018827067669172932, + "grad_norm": 0.345703125, + "learning_rate": 3e-05, + "loss": 1.6923, + "step": 313 + }, + { + "epoch": 0.018887218045112782, + "grad_norm": 0.3125, + "learning_rate": 3e-05, + "loss": 1.7099, + "step": 314 + }, + { + "epoch": 0.018947368421052633, + "grad_norm": 0.25390625, + "learning_rate": 3e-05, + "loss": 1.695, + "step": 315 + }, + { + "epoch": 0.01900751879699248, + "grad_norm": 0.26171875, + "learning_rate": 3e-05, + "loss": 1.6492, + "step": 316 + }, + { + "epoch": 0.01906766917293233, + "grad_norm": 0.3515625, + "learning_rate": 3e-05, + "loss": 1.7011, + "step": 317 + }, + { + "epoch": 0.01912781954887218, + "grad_norm": 0.259765625, + "learning_rate": 3e-05, + "loss": 1.7115, + "step": 318 + }, + { + "epoch": 0.01918796992481203, + "grad_norm": 0.296875, + "learning_rate": 3e-05, + "loss": 1.6541, + "step": 319 + }, + { + "epoch": 0.01924812030075188, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.7012, + "step": 320 + }, + { + "epoch": 0.01930827067669173, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.6688, + "step": 321 + }, + { + "epoch": 0.01936842105263158, + "grad_norm": 0.310546875, + "learning_rate": 3e-05, + "loss": 1.7098, + "step": 322 + }, + { + "epoch": 0.019428571428571427, + "grad_norm": 0.306640625, + "learning_rate": 3e-05, + "loss": 1.6908, + "step": 323 + }, + { + "epoch": 0.019488721804511278, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.7348, + "step": 324 + }, + { + "epoch": 0.019548872180451128, + "grad_norm": 0.314453125, + "learning_rate": 3e-05, + "loss": 1.7534, + "step": 325 + }, + { + "epoch": 0.01960902255639098, + "grad_norm": 0.3515625, + "learning_rate": 3e-05, + "loss": 1.6591, + "step": 326 + }, + { + "epoch": 0.019669172932330826, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.6947, + "step": 327 + }, + { + "epoch": 0.019729323308270676, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.6182, + "step": 328 + }, + { + "epoch": 0.019789473684210527, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.6764, + "step": 329 + }, + { + "epoch": 0.019849624060150377, + "grad_norm": 0.2578125, + "learning_rate": 3e-05, + "loss": 1.7097, + "step": 330 + }, + { + "epoch": 0.019909774436090225, + "grad_norm": 0.25390625, + "learning_rate": 3e-05, + "loss": 1.6522, + "step": 331 + }, + { + "epoch": 0.019969924812030075, + "grad_norm": 0.341796875, + "learning_rate": 3e-05, + "loss": 1.6445, + "step": 332 + }, + { + "epoch": 0.020030075187969926, + "grad_norm": 0.35546875, + "learning_rate": 3e-05, + "loss": 1.6974, + "step": 333 + }, + { + "epoch": 0.020090225563909773, + "grad_norm": 0.28515625, + "learning_rate": 3e-05, + "loss": 1.6834, + "step": 334 + }, + { + "epoch": 0.020150375939849623, + "grad_norm": 0.37890625, + "learning_rate": 3e-05, + "loss": 1.6474, + "step": 335 + }, + { + "epoch": 0.020210526315789474, + "grad_norm": 0.314453125, + "learning_rate": 3e-05, + "loss": 1.6811, + "step": 336 + }, + { + "epoch": 0.020270676691729324, + "grad_norm": 0.27734375, + "learning_rate": 3e-05, + "loss": 1.6846, + "step": 337 + }, + { + "epoch": 0.02033082706766917, + "grad_norm": 0.26171875, + "learning_rate": 3e-05, + "loss": 1.6993, + "step": 338 + }, + { + "epoch": 0.020390977443609022, + "grad_norm": 0.4375, + "learning_rate": 3e-05, + "loss": 1.7186, + "step": 339 + }, + { + "epoch": 0.020451127819548873, + "grad_norm": 0.3203125, + "learning_rate": 3e-05, + "loss": 1.7092, + "step": 340 + }, + { + "epoch": 0.020511278195488723, + "grad_norm": 0.328125, + "learning_rate": 3e-05, + "loss": 1.6459, + "step": 341 + }, + { + "epoch": 0.02057142857142857, + "grad_norm": 0.337890625, + "learning_rate": 3e-05, + "loss": 1.7021, + "step": 342 + }, + { + "epoch": 0.02063157894736842, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.7139, + "step": 343 + }, + { + "epoch": 0.02069172932330827, + "grad_norm": 0.353515625, + "learning_rate": 3e-05, + "loss": 1.6841, + "step": 344 + }, + { + "epoch": 0.02075187969924812, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.7118, + "step": 345 + }, + { + "epoch": 0.02081203007518797, + "grad_norm": 0.392578125, + "learning_rate": 3e-05, + "loss": 1.7099, + "step": 346 + }, + { + "epoch": 0.02087218045112782, + "grad_norm": 0.279296875, + "learning_rate": 3e-05, + "loss": 1.6352, + "step": 347 + }, + { + "epoch": 0.02093233082706767, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.6537, + "step": 348 + }, + { + "epoch": 0.020992481203007517, + "grad_norm": 0.318359375, + "learning_rate": 3e-05, + "loss": 1.7208, + "step": 349 + }, + { + "epoch": 0.021052631578947368, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.6622, + "step": 350 + }, + { + "epoch": 0.02111278195488722, + "grad_norm": 0.318359375, + "learning_rate": 3e-05, + "loss": 1.6665, + "step": 351 + }, + { + "epoch": 0.02117293233082707, + "grad_norm": 0.296875, + "learning_rate": 3e-05, + "loss": 1.6822, + "step": 352 + }, + { + "epoch": 0.021233082706766916, + "grad_norm": 0.27734375, + "learning_rate": 3e-05, + "loss": 1.7216, + "step": 353 + }, + { + "epoch": 0.021293233082706767, + "grad_norm": 0.361328125, + "learning_rate": 3e-05, + "loss": 1.6833, + "step": 354 + }, + { + "epoch": 0.021353383458646617, + "grad_norm": 0.3671875, + "learning_rate": 3e-05, + "loss": 1.6547, + "step": 355 + }, + { + "epoch": 0.021413533834586468, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.6886, + "step": 356 + }, + { + "epoch": 0.021473684210526315, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.6601, + "step": 357 + }, + { + "epoch": 0.021533834586466166, + "grad_norm": 0.40625, + "learning_rate": 3e-05, + "loss": 1.693, + "step": 358 + }, + { + "epoch": 0.021593984962406016, + "grad_norm": 0.42578125, + "learning_rate": 3e-05, + "loss": 1.6726, + "step": 359 + }, + { + "epoch": 0.021654135338345863, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.7261, + "step": 360 + }, + { + "epoch": 0.021714285714285714, + "grad_norm": 0.36328125, + "learning_rate": 3e-05, + "loss": 1.6836, + "step": 361 + }, + { + "epoch": 0.021774436090225564, + "grad_norm": 0.302734375, + "learning_rate": 3e-05, + "loss": 1.6804, + "step": 362 + }, + { + "epoch": 0.021834586466165415, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.6727, + "step": 363 + }, + { + "epoch": 0.021894736842105262, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.6405, + "step": 364 + }, + { + "epoch": 0.021954887218045113, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.6474, + "step": 365 + }, + { + "epoch": 0.022015037593984963, + "grad_norm": 0.287109375, + "learning_rate": 3e-05, + "loss": 1.6932, + "step": 366 + }, + { + "epoch": 0.022075187969924814, + "grad_norm": 0.3125, + "learning_rate": 3e-05, + "loss": 1.6955, + "step": 367 + }, + { + "epoch": 0.02213533834586466, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.7171, + "step": 368 + }, + { + "epoch": 0.02219548872180451, + "grad_norm": 0.291015625, + "learning_rate": 3e-05, + "loss": 1.6309, + "step": 369 + }, + { + "epoch": 0.022255639097744362, + "grad_norm": 0.388671875, + "learning_rate": 3e-05, + "loss": 1.6478, + "step": 370 + }, + { + "epoch": 0.02231578947368421, + "grad_norm": 0.3203125, + "learning_rate": 3e-05, + "loss": 1.6229, + "step": 371 + }, + { + "epoch": 0.02237593984962406, + "grad_norm": 0.3671875, + "learning_rate": 3e-05, + "loss": 1.6583, + "step": 372 + }, + { + "epoch": 0.02243609022556391, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.6396, + "step": 373 + }, + { + "epoch": 0.02249624060150376, + "grad_norm": 0.365234375, + "learning_rate": 3e-05, + "loss": 1.6591, + "step": 374 + }, + { + "epoch": 0.022556390977443608, + "grad_norm": 0.34765625, + "learning_rate": 3e-05, + "loss": 1.6788, + "step": 375 + }, + { + "epoch": 0.02261654135338346, + "grad_norm": 0.328125, + "learning_rate": 3e-05, + "loss": 1.6644, + "step": 376 + }, + { + "epoch": 0.02267669172932331, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.6267, + "step": 377 + }, + { + "epoch": 0.02273684210526316, + "grad_norm": 0.349609375, + "learning_rate": 3e-05, + "loss": 1.6582, + "step": 378 + }, + { + "epoch": 0.022796992481203007, + "grad_norm": 0.259765625, + "learning_rate": 3e-05, + "loss": 1.6489, + "step": 379 + }, + { + "epoch": 0.022857142857142857, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.6659, + "step": 380 + }, + { + "epoch": 0.022917293233082708, + "grad_norm": 0.349609375, + "learning_rate": 3e-05, + "loss": 1.6392, + "step": 381 + }, + { + "epoch": 0.022977443609022555, + "grad_norm": 0.32421875, + "learning_rate": 3e-05, + "loss": 1.6542, + "step": 382 + }, + { + "epoch": 0.023037593984962405, + "grad_norm": 0.306640625, + "learning_rate": 3e-05, + "loss": 1.6822, + "step": 383 + }, + { + "epoch": 0.023097744360902256, + "grad_norm": 0.28125, + "learning_rate": 3e-05, + "loss": 1.6539, + "step": 384 + }, + { + "epoch": 0.023157894736842106, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.7668, + "step": 385 + }, + { + "epoch": 0.023218045112781954, + "grad_norm": 0.3671875, + "learning_rate": 3e-05, + "loss": 1.6502, + "step": 386 + }, + { + "epoch": 0.023278195488721804, + "grad_norm": 0.349609375, + "learning_rate": 3e-05, + "loss": 1.6883, + "step": 387 + }, + { + "epoch": 0.023338345864661655, + "grad_norm": 0.306640625, + "learning_rate": 3e-05, + "loss": 1.6394, + "step": 388 + }, + { + "epoch": 0.023398496240601505, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.7065, + "step": 389 + }, + { + "epoch": 0.023458646616541352, + "grad_norm": 0.349609375, + "learning_rate": 3e-05, + "loss": 1.6415, + "step": 390 + }, + { + "epoch": 0.023518796992481203, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.6338, + "step": 391 + }, + { + "epoch": 0.023578947368421053, + "grad_norm": 0.375, + "learning_rate": 3e-05, + "loss": 1.6269, + "step": 392 + }, + { + "epoch": 0.0236390977443609, + "grad_norm": 0.2890625, + "learning_rate": 3e-05, + "loss": 1.6805, + "step": 393 + }, + { + "epoch": 0.02369924812030075, + "grad_norm": 0.2890625, + "learning_rate": 3e-05, + "loss": 1.6593, + "step": 394 + }, + { + "epoch": 0.0237593984962406, + "grad_norm": 0.345703125, + "learning_rate": 3e-05, + "loss": 1.6691, + "step": 395 + }, + { + "epoch": 0.023819548872180452, + "grad_norm": 0.34765625, + "learning_rate": 3e-05, + "loss": 1.7221, + "step": 396 + }, + { + "epoch": 0.0238796992481203, + "grad_norm": 0.361328125, + "learning_rate": 3e-05, + "loss": 1.697, + "step": 397 + }, + { + "epoch": 0.02393984962406015, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.6601, + "step": 398 + }, + { + "epoch": 0.024, + "grad_norm": 0.330078125, + "learning_rate": 3e-05, + "loss": 1.6351, + "step": 399 + }, + { + "epoch": 0.02406015037593985, + "grad_norm": 0.3359375, + "learning_rate": 3e-05, + "loss": 1.6772, + "step": 400 + }, + { + "epoch": 0.024120300751879698, + "grad_norm": 0.328125, + "learning_rate": 3e-05, + "loss": 1.6756, + "step": 401 + }, + { + "epoch": 0.02418045112781955, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.6499, + "step": 402 + }, + { + "epoch": 0.0242406015037594, + "grad_norm": 0.373046875, + "learning_rate": 3e-05, + "loss": 1.6487, + "step": 403 + }, + { + "epoch": 0.02430075187969925, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.7147, + "step": 404 + }, + { + "epoch": 0.024360902255639097, + "grad_norm": 0.328125, + "learning_rate": 3e-05, + "loss": 1.6577, + "step": 405 + }, + { + "epoch": 0.024421052631578948, + "grad_norm": 0.384765625, + "learning_rate": 3e-05, + "loss": 1.6751, + "step": 406 + }, + { + "epoch": 0.024481203007518798, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.6736, + "step": 407 + }, + { + "epoch": 0.024541353383458645, + "grad_norm": 0.369140625, + "learning_rate": 3e-05, + "loss": 1.6512, + "step": 408 + }, + { + "epoch": 0.024601503759398496, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.6538, + "step": 409 + }, + { + "epoch": 0.024661654135338346, + "grad_norm": 0.283203125, + "learning_rate": 3e-05, + "loss": 1.6633, + "step": 410 + }, + { + "epoch": 0.024721804511278197, + "grad_norm": 0.353515625, + "learning_rate": 3e-05, + "loss": 1.6943, + "step": 411 + }, + { + "epoch": 0.024781954887218044, + "grad_norm": 0.2734375, + "learning_rate": 3e-05, + "loss": 1.6559, + "step": 412 + }, + { + "epoch": 0.024842105263157895, + "grad_norm": 0.3984375, + "learning_rate": 3e-05, + "loss": 1.6143, + "step": 413 + }, + { + "epoch": 0.024902255639097745, + "grad_norm": 0.38671875, + "learning_rate": 3e-05, + "loss": 1.6313, + "step": 414 + }, + { + "epoch": 0.024962406015037596, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.7871, + "step": 415 + }, + { + "epoch": 0.025022556390977443, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.6886, + "step": 416 + }, + { + "epoch": 0.025082706766917293, + "grad_norm": 0.34375, + "learning_rate": 3e-05, + "loss": 1.6331, + "step": 417 + }, + { + "epoch": 0.025142857142857144, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.6778, + "step": 418 + }, + { + "epoch": 0.02520300751879699, + "grad_norm": 0.388671875, + "learning_rate": 3e-05, + "loss": 1.6649, + "step": 419 + }, + { + "epoch": 0.02526315789473684, + "grad_norm": 0.3203125, + "learning_rate": 3e-05, + "loss": 1.6713, + "step": 420 + }, + { + "epoch": 0.025323308270676692, + "grad_norm": 0.40625, + "learning_rate": 3e-05, + "loss": 1.6047, + "step": 421 + }, + { + "epoch": 0.025383458646616543, + "grad_norm": 0.3515625, + "learning_rate": 3e-05, + "loss": 1.6532, + "step": 422 + }, + { + "epoch": 0.02544360902255639, + "grad_norm": 0.33203125, + "learning_rate": 3e-05, + "loss": 1.6341, + "step": 423 + }, + { + "epoch": 0.02550375939849624, + "grad_norm": 0.4140625, + "learning_rate": 3e-05, + "loss": 1.6775, + "step": 424 + }, + { + "epoch": 0.02556390977443609, + "grad_norm": 0.373046875, + "learning_rate": 3e-05, + "loss": 1.6547, + "step": 425 + }, + { + "epoch": 0.02562406015037594, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.6208, + "step": 426 + }, + { + "epoch": 0.02568421052631579, + "grad_norm": 0.34765625, + "learning_rate": 3e-05, + "loss": 1.694, + "step": 427 + }, + { + "epoch": 0.02574436090225564, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.6579, + "step": 428 + }, + { + "epoch": 0.02580451127819549, + "grad_norm": 0.3515625, + "learning_rate": 3e-05, + "loss": 1.6262, + "step": 429 + }, + { + "epoch": 0.025864661654135337, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.6499, + "step": 430 + }, + { + "epoch": 0.025924812030075187, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.5616, + "step": 431 + }, + { + "epoch": 0.025984962406015038, + "grad_norm": 0.349609375, + "learning_rate": 3e-05, + "loss": 1.6801, + "step": 432 + }, + { + "epoch": 0.02604511278195489, + "grad_norm": 0.365234375, + "learning_rate": 3e-05, + "loss": 1.656, + "step": 433 + }, + { + "epoch": 0.026105263157894736, + "grad_norm": 0.306640625, + "learning_rate": 3e-05, + "loss": 1.6269, + "step": 434 + }, + { + "epoch": 0.026165413533834586, + "grad_norm": 0.27734375, + "learning_rate": 3e-05, + "loss": 1.6915, + "step": 435 + }, + { + "epoch": 0.026225563909774437, + "grad_norm": 0.40625, + "learning_rate": 3e-05, + "loss": 1.6256, + "step": 436 + }, + { + "epoch": 0.026285714285714287, + "grad_norm": 0.353515625, + "learning_rate": 3e-05, + "loss": 1.6531, + "step": 437 + }, + { + "epoch": 0.026345864661654134, + "grad_norm": 0.3828125, + "learning_rate": 3e-05, + "loss": 1.6314, + "step": 438 + }, + { + "epoch": 0.026406015037593985, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.669, + "step": 439 + }, + { + "epoch": 0.026466165413533835, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.6893, + "step": 440 + }, + { + "epoch": 0.026526315789473683, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.6608, + "step": 441 + }, + { + "epoch": 0.026586466165413533, + "grad_norm": 0.36328125, + "learning_rate": 3e-05, + "loss": 1.7154, + "step": 442 + }, + { + "epoch": 0.026646616541353384, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.6203, + "step": 443 + }, + { + "epoch": 0.026706766917293234, + "grad_norm": 0.27734375, + "learning_rate": 3e-05, + "loss": 1.6863, + "step": 444 + }, + { + "epoch": 0.02676691729323308, + "grad_norm": 0.357421875, + "learning_rate": 3e-05, + "loss": 1.6646, + "step": 445 + }, + { + "epoch": 0.026827067669172932, + "grad_norm": 0.38671875, + "learning_rate": 3e-05, + "loss": 1.6631, + "step": 446 + }, + { + "epoch": 0.026887218045112782, + "grad_norm": 0.3359375, + "learning_rate": 3e-05, + "loss": 1.6432, + "step": 447 + }, + { + "epoch": 0.026947368421052633, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.6573, + "step": 448 + }, + { + "epoch": 0.02700751879699248, + "grad_norm": 0.34765625, + "learning_rate": 3e-05, + "loss": 1.5616, + "step": 449 + }, + { + "epoch": 0.02706766917293233, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.7278, + "step": 450 + }, + { + "epoch": 0.02712781954887218, + "grad_norm": 0.384765625, + "learning_rate": 3e-05, + "loss": 1.6633, + "step": 451 + }, + { + "epoch": 0.02718796992481203, + "grad_norm": 0.30859375, + "learning_rate": 3e-05, + "loss": 1.6304, + "step": 452 + }, + { + "epoch": 0.02724812030075188, + "grad_norm": 0.30859375, + "learning_rate": 3e-05, + "loss": 1.5982, + "step": 453 + }, + { + "epoch": 0.02730827067669173, + "grad_norm": 0.384765625, + "learning_rate": 3e-05, + "loss": 1.6142, + "step": 454 + }, + { + "epoch": 0.02736842105263158, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.6318, + "step": 455 + }, + { + "epoch": 0.027428571428571427, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.5753, + "step": 456 + }, + { + "epoch": 0.027488721804511278, + "grad_norm": 0.40234375, + "learning_rate": 3e-05, + "loss": 1.626, + "step": 457 + }, + { + "epoch": 0.02754887218045113, + "grad_norm": 0.396484375, + "learning_rate": 3e-05, + "loss": 1.6517, + "step": 458 + }, + { + "epoch": 0.02760902255639098, + "grad_norm": 0.287109375, + "learning_rate": 3e-05, + "loss": 1.6437, + "step": 459 + }, + { + "epoch": 0.027669172932330826, + "grad_norm": 0.341796875, + "learning_rate": 3e-05, + "loss": 1.6395, + "step": 460 + }, + { + "epoch": 0.027729323308270677, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.6362, + "step": 461 + }, + { + "epoch": 0.027789473684210527, + "grad_norm": 0.36328125, + "learning_rate": 3e-05, + "loss": 1.6608, + "step": 462 + }, + { + "epoch": 0.027849624060150378, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.672, + "step": 463 + }, + { + "epoch": 0.027909774436090225, + "grad_norm": 0.349609375, + "learning_rate": 3e-05, + "loss": 1.6565, + "step": 464 + }, + { + "epoch": 0.027969924812030075, + "grad_norm": 0.396484375, + "learning_rate": 3e-05, + "loss": 1.5993, + "step": 465 + }, + { + "epoch": 0.028030075187969926, + "grad_norm": 0.349609375, + "learning_rate": 3e-05, + "loss": 1.7512, + "step": 466 + }, + { + "epoch": 0.028090225563909773, + "grad_norm": 0.359375, + "learning_rate": 3e-05, + "loss": 1.6191, + "step": 467 + }, + { + "epoch": 0.028150375939849624, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.6314, + "step": 468 + }, + { + "epoch": 0.028210526315789474, + "grad_norm": 0.279296875, + "learning_rate": 3e-05, + "loss": 1.6485, + "step": 469 + }, + { + "epoch": 0.028270676691729325, + "grad_norm": 0.330078125, + "learning_rate": 3e-05, + "loss": 1.656, + "step": 470 + }, + { + "epoch": 0.028330827067669172, + "grad_norm": 0.376953125, + "learning_rate": 3e-05, + "loss": 1.6116, + "step": 471 + }, + { + "epoch": 0.028390977443609022, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.5927, + "step": 472 + }, + { + "epoch": 0.028451127819548873, + "grad_norm": 0.376953125, + "learning_rate": 3e-05, + "loss": 1.6472, + "step": 473 + }, + { + "epoch": 0.028511278195488723, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.6097, + "step": 474 + }, + { + "epoch": 0.02857142857142857, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.6408, + "step": 475 + }, + { + "epoch": 0.02863157894736842, + "grad_norm": 0.296875, + "learning_rate": 3e-05, + "loss": 1.6245, + "step": 476 + }, + { + "epoch": 0.02869172932330827, + "grad_norm": 0.3046875, + "learning_rate": 3e-05, + "loss": 1.6627, + "step": 477 + }, + { + "epoch": 0.02875187969924812, + "grad_norm": 0.38671875, + "learning_rate": 3e-05, + "loss": 1.7246, + "step": 478 + }, + { + "epoch": 0.02881203007518797, + "grad_norm": 0.3359375, + "learning_rate": 3e-05, + "loss": 1.6175, + "step": 479 + }, + { + "epoch": 0.02887218045112782, + "grad_norm": 0.34375, + "learning_rate": 3e-05, + "loss": 1.6194, + "step": 480 + }, + { + "epoch": 0.02893233082706767, + "grad_norm": 0.35546875, + "learning_rate": 3e-05, + "loss": 1.6492, + "step": 481 + }, + { + "epoch": 0.028992481203007518, + "grad_norm": 0.42578125, + "learning_rate": 3e-05, + "loss": 1.6463, + "step": 482 + }, + { + "epoch": 0.029052631578947368, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.646, + "step": 483 + }, + { + "epoch": 0.02911278195488722, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.6725, + "step": 484 + }, + { + "epoch": 0.02917293233082707, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.6494, + "step": 485 + }, + { + "epoch": 0.029233082706766916, + "grad_norm": 0.40234375, + "learning_rate": 3e-05, + "loss": 1.6608, + "step": 486 + }, + { + "epoch": 0.029293233082706767, + "grad_norm": 0.302734375, + "learning_rate": 3e-05, + "loss": 1.5914, + "step": 487 + }, + { + "epoch": 0.029353383458646617, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.701, + "step": 488 + }, + { + "epoch": 0.029413533834586465, + "grad_norm": 0.353515625, + "learning_rate": 3e-05, + "loss": 1.6541, + "step": 489 + }, + { + "epoch": 0.029473684210526315, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.6144, + "step": 490 + }, + { + "epoch": 0.029533834586466166, + "grad_norm": 0.384765625, + "learning_rate": 3e-05, + "loss": 1.6412, + "step": 491 + }, + { + "epoch": 0.029593984962406016, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.5989, + "step": 492 + }, + { + "epoch": 0.029654135338345863, + "grad_norm": 0.345703125, + "learning_rate": 3e-05, + "loss": 1.6327, + "step": 493 + }, + { + "epoch": 0.029714285714285714, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.6262, + "step": 494 + }, + { + "epoch": 0.029774436090225564, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.6482, + "step": 495 + }, + { + "epoch": 0.029834586466165415, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.5741, + "step": 496 + }, + { + "epoch": 0.029894736842105262, + "grad_norm": 0.357421875, + "learning_rate": 3e-05, + "loss": 1.6247, + "step": 497 + }, + { + "epoch": 0.029954887218045113, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.5678, + "step": 498 + }, + { + "epoch": 0.030015037593984963, + "grad_norm": 0.35546875, + "learning_rate": 3e-05, + "loss": 1.65, + "step": 499 + }, + { + "epoch": 0.03007518796992481, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.6096, + "step": 500 + }, + { + "epoch": 0.03013533834586466, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5983, + "step": 501 + }, + { + "epoch": 0.03019548872180451, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.6518, + "step": 502 + }, + { + "epoch": 0.030255639097744362, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.57, + "step": 503 + }, + { + "epoch": 0.03031578947368421, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.6363, + "step": 504 + }, + { + "epoch": 0.03037593984962406, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.6915, + "step": 505 + }, + { + "epoch": 0.03043609022556391, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.5727, + "step": 506 + }, + { + "epoch": 0.03049624060150376, + "grad_norm": 0.36328125, + "learning_rate": 3e-05, + "loss": 1.6456, + "step": 507 + }, + { + "epoch": 0.030556390977443608, + "grad_norm": 0.375, + "learning_rate": 3e-05, + "loss": 1.5998, + "step": 508 + }, + { + "epoch": 0.03061654135338346, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.6063, + "step": 509 + }, + { + "epoch": 0.03067669172932331, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.7144, + "step": 510 + }, + { + "epoch": 0.030736842105263156, + "grad_norm": 0.37890625, + "learning_rate": 3e-05, + "loss": 1.5877, + "step": 511 + }, + { + "epoch": 0.030796992481203007, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.6705, + "step": 512 + }, + { + "epoch": 0.030857142857142857, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.6083, + "step": 513 + }, + { + "epoch": 0.030917293233082708, + "grad_norm": 0.376953125, + "learning_rate": 3e-05, + "loss": 1.6248, + "step": 514 + }, + { + "epoch": 0.030977443609022555, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.719, + "step": 515 + }, + { + "epoch": 0.031037593984962406, + "grad_norm": 0.396484375, + "learning_rate": 3e-05, + "loss": 1.6422, + "step": 516 + }, + { + "epoch": 0.031097744360902256, + "grad_norm": 0.328125, + "learning_rate": 3e-05, + "loss": 1.6281, + "step": 517 + }, + { + "epoch": 0.031157894736842107, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.5856, + "step": 518 + }, + { + "epoch": 0.031218045112781954, + "grad_norm": 0.396484375, + "learning_rate": 3e-05, + "loss": 1.6604, + "step": 519 + }, + { + "epoch": 0.03127819548872181, + "grad_norm": 0.384765625, + "learning_rate": 3e-05, + "loss": 1.6908, + "step": 520 + }, + { + "epoch": 0.03133834586466165, + "grad_norm": 0.37890625, + "learning_rate": 3e-05, + "loss": 1.6022, + "step": 521 + }, + { + "epoch": 0.0313984962406015, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.6319, + "step": 522 + }, + { + "epoch": 0.03145864661654135, + "grad_norm": 0.3515625, + "learning_rate": 3e-05, + "loss": 1.7282, + "step": 523 + }, + { + "epoch": 0.0315187969924812, + "grad_norm": 0.37109375, + "learning_rate": 3e-05, + "loss": 1.5997, + "step": 524 + }, + { + "epoch": 0.031578947368421054, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.6207, + "step": 525 + }, + { + "epoch": 0.031639097744360904, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.6152, + "step": 526 + }, + { + "epoch": 0.031699248120300755, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.6444, + "step": 527 + }, + { + "epoch": 0.0317593984962406, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.63, + "step": 528 + }, + { + "epoch": 0.03181954887218045, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.6547, + "step": 529 + }, + { + "epoch": 0.0318796992481203, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.628, + "step": 530 + }, + { + "epoch": 0.03193984962406015, + "grad_norm": 0.349609375, + "learning_rate": 3e-05, + "loss": 1.6469, + "step": 531 + }, + { + "epoch": 0.032, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.6612, + "step": 532 + }, + { + "epoch": 0.03206015037593985, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.6525, + "step": 533 + }, + { + "epoch": 0.0321203007518797, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.6262, + "step": 534 + }, + { + "epoch": 0.032180451127819545, + "grad_norm": 0.384765625, + "learning_rate": 3e-05, + "loss": 1.582, + "step": 535 + }, + { + "epoch": 0.032240601503759396, + "grad_norm": 0.3515625, + "learning_rate": 3e-05, + "loss": 1.6203, + "step": 536 + }, + { + "epoch": 0.03230075187969925, + "grad_norm": 0.390625, + "learning_rate": 3e-05, + "loss": 1.6597, + "step": 537 + }, + { + "epoch": 0.0323609022556391, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.4622, + "step": 538 + }, + { + "epoch": 0.03242105263157895, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.6001, + "step": 539 + }, + { + "epoch": 0.0324812030075188, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.615, + "step": 540 + }, + { + "epoch": 0.03254135338345865, + "grad_norm": 0.314453125, + "learning_rate": 3e-05, + "loss": 1.7292, + "step": 541 + }, + { + "epoch": 0.0326015037593985, + "grad_norm": 0.388671875, + "learning_rate": 3e-05, + "loss": 1.5689, + "step": 542 + }, + { + "epoch": 0.03266165413533834, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.6036, + "step": 543 + }, + { + "epoch": 0.032721804511278194, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.5784, + "step": 544 + }, + { + "epoch": 0.032781954887218044, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.6603, + "step": 545 + }, + { + "epoch": 0.032842105263157895, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.6956, + "step": 546 + }, + { + "epoch": 0.032902255639097745, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.5317, + "step": 547 + }, + { + "epoch": 0.032962406015037596, + "grad_norm": 0.3984375, + "learning_rate": 3e-05, + "loss": 1.6626, + "step": 548 + }, + { + "epoch": 0.033022556390977446, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.6168, + "step": 549 + }, + { + "epoch": 0.03308270676691729, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5427, + "step": 550 + }, + { + "epoch": 0.03314285714285714, + "grad_norm": 0.3359375, + "learning_rate": 3e-05, + "loss": 1.5969, + "step": 551 + }, + { + "epoch": 0.03320300751879699, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.6369, + "step": 552 + }, + { + "epoch": 0.03326315789473684, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.5902, + "step": 553 + }, + { + "epoch": 0.03332330827067669, + "grad_norm": 0.3984375, + "learning_rate": 3e-05, + "loss": 1.5936, + "step": 554 + }, + { + "epoch": 0.03338345864661654, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.623, + "step": 555 + }, + { + "epoch": 0.03344360902255639, + "grad_norm": 0.36328125, + "learning_rate": 3e-05, + "loss": 1.6279, + "step": 556 + }, + { + "epoch": 0.033503759398496244, + "grad_norm": 0.35546875, + "learning_rate": 3e-05, + "loss": 1.5906, + "step": 557 + }, + { + "epoch": 0.03356390977443609, + "grad_norm": 0.375, + "learning_rate": 3e-05, + "loss": 1.5765, + "step": 558 + }, + { + "epoch": 0.03362406015037594, + "grad_norm": 0.421875, + "learning_rate": 3e-05, + "loss": 1.6852, + "step": 559 + }, + { + "epoch": 0.03368421052631579, + "grad_norm": 0.4375, + "learning_rate": 3e-05, + "loss": 1.6368, + "step": 560 + }, + { + "epoch": 0.03374436090225564, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.6017, + "step": 561 + }, + { + "epoch": 0.03380451127819549, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.6204, + "step": 562 + }, + { + "epoch": 0.03386466165413534, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.617, + "step": 563 + }, + { + "epoch": 0.03392481203007519, + "grad_norm": 0.328125, + "learning_rate": 3e-05, + "loss": 1.6297, + "step": 564 + }, + { + "epoch": 0.033984962406015035, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.5375, + "step": 565 + }, + { + "epoch": 0.034045112781954885, + "grad_norm": 0.322265625, + "learning_rate": 3e-05, + "loss": 1.6506, + "step": 566 + }, + { + "epoch": 0.034105263157894736, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.65, + "step": 567 + }, + { + "epoch": 0.034165413533834586, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.6122, + "step": 568 + }, + { + "epoch": 0.03422556390977444, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.6834, + "step": 569 + }, + { + "epoch": 0.03428571428571429, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.5937, + "step": 570 + }, + { + "epoch": 0.03434586466165414, + "grad_norm": 0.33984375, + "learning_rate": 3e-05, + "loss": 1.6337, + "step": 571 + }, + { + "epoch": 0.03440601503759398, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.6422, + "step": 572 + }, + { + "epoch": 0.03446616541353383, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.6621, + "step": 573 + }, + { + "epoch": 0.03452631578947368, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.6646, + "step": 574 + }, + { + "epoch": 0.03458646616541353, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.67, + "step": 575 + }, + { + "epoch": 0.034646616541353384, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.6912, + "step": 576 + }, + { + "epoch": 0.034706766917293234, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.6151, + "step": 577 + }, + { + "epoch": 0.034766917293233085, + "grad_norm": 0.388671875, + "learning_rate": 3e-05, + "loss": 1.6044, + "step": 578 + }, + { + "epoch": 0.034827067669172936, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.5894, + "step": 579 + }, + { + "epoch": 0.03488721804511278, + "grad_norm": 0.384765625, + "learning_rate": 3e-05, + "loss": 1.6397, + "step": 580 + }, + { + "epoch": 0.03494736842105263, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.5841, + "step": 581 + }, + { + "epoch": 0.03500751879699248, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.5984, + "step": 582 + }, + { + "epoch": 0.03506766917293233, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4452, + "step": 583 + }, + { + "epoch": 0.03512781954887218, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5672, + "step": 584 + }, + { + "epoch": 0.03518796992481203, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.6049, + "step": 585 + }, + { + "epoch": 0.03524812030075188, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.5622, + "step": 586 + }, + { + "epoch": 0.035308270676691726, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5929, + "step": 587 + }, + { + "epoch": 0.03536842105263158, + "grad_norm": 0.357421875, + "learning_rate": 3e-05, + "loss": 1.5686, + "step": 588 + }, + { + "epoch": 0.03542857142857143, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.617, + "step": 589 + }, + { + "epoch": 0.03548872180451128, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.6373, + "step": 590 + }, + { + "epoch": 0.03554887218045113, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.5678, + "step": 591 + }, + { + "epoch": 0.03560902255639098, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.5805, + "step": 592 + }, + { + "epoch": 0.03566917293233083, + "grad_norm": 0.396484375, + "learning_rate": 3e-05, + "loss": 1.5789, + "step": 593 + }, + { + "epoch": 0.03572932330827068, + "grad_norm": 0.310546875, + "learning_rate": 3e-05, + "loss": 1.597, + "step": 594 + }, + { + "epoch": 0.035789473684210524, + "grad_norm": 0.3515625, + "learning_rate": 3e-05, + "loss": 1.5963, + "step": 595 + }, + { + "epoch": 0.035849624060150374, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.6732, + "step": 596 + }, + { + "epoch": 0.035909774436090225, + "grad_norm": 0.373046875, + "learning_rate": 3e-05, + "loss": 1.6617, + "step": 597 + }, + { + "epoch": 0.035969924812030075, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.6135, + "step": 598 + }, + { + "epoch": 0.036030075187969926, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.6713, + "step": 599 + }, + { + "epoch": 0.03609022556390978, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.6124, + "step": 600 + }, + { + "epoch": 0.03615037593984963, + "grad_norm": 0.33984375, + "learning_rate": 3e-05, + "loss": 1.674, + "step": 601 + }, + { + "epoch": 0.03621052631578947, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.59, + "step": 602 + }, + { + "epoch": 0.03627067669172932, + "grad_norm": 0.3359375, + "learning_rate": 3e-05, + "loss": 1.6028, + "step": 603 + }, + { + "epoch": 0.03633082706766917, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.6312, + "step": 604 + }, + { + "epoch": 0.03639097744360902, + "grad_norm": 0.38671875, + "learning_rate": 3e-05, + "loss": 1.5781, + "step": 605 + }, + { + "epoch": 0.03645112781954887, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.5715, + "step": 606 + }, + { + "epoch": 0.036511278195488724, + "grad_norm": 0.421875, + "learning_rate": 3e-05, + "loss": 1.6748, + "step": 607 + }, + { + "epoch": 0.036571428571428574, + "grad_norm": 0.421875, + "learning_rate": 3e-05, + "loss": 1.6109, + "step": 608 + }, + { + "epoch": 0.03663157894736842, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.5897, + "step": 609 + }, + { + "epoch": 0.03669172932330827, + "grad_norm": 0.27734375, + "learning_rate": 3e-05, + "loss": 1.6298, + "step": 610 + }, + { + "epoch": 0.03675187969924812, + "grad_norm": 0.392578125, + "learning_rate": 3e-05, + "loss": 1.6317, + "step": 611 + }, + { + "epoch": 0.03681203007518797, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.5982, + "step": 612 + }, + { + "epoch": 0.03687218045112782, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.6164, + "step": 613 + }, + { + "epoch": 0.03693233082706767, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.5303, + "step": 614 + }, + { + "epoch": 0.03699248120300752, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.6088, + "step": 615 + }, + { + "epoch": 0.03705263157894737, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.6343, + "step": 616 + }, + { + "epoch": 0.037112781954887215, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.6646, + "step": 617 + }, + { + "epoch": 0.037172932330827066, + "grad_norm": 0.3671875, + "learning_rate": 3e-05, + "loss": 1.6177, + "step": 618 + }, + { + "epoch": 0.037233082706766917, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.6438, + "step": 619 + }, + { + "epoch": 0.03729323308270677, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.52, + "step": 620 + }, + { + "epoch": 0.03735338345864662, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.6666, + "step": 621 + }, + { + "epoch": 0.03741353383458647, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.5993, + "step": 622 + }, + { + "epoch": 0.03747368421052632, + "grad_norm": 0.462890625, + "learning_rate": 3e-05, + "loss": 1.6169, + "step": 623 + }, + { + "epoch": 0.03753383458646616, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.5851, + "step": 624 + }, + { + "epoch": 0.03759398496240601, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.754, + "step": 625 + }, + { + "epoch": 0.037654135338345864, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.6065, + "step": 626 + }, + { + "epoch": 0.037714285714285714, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.5223, + "step": 627 + }, + { + "epoch": 0.037774436090225565, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.6798, + "step": 628 + }, + { + "epoch": 0.037834586466165415, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.604, + "step": 629 + }, + { + "epoch": 0.037894736842105266, + "grad_norm": 0.388671875, + "learning_rate": 3e-05, + "loss": 1.5695, + "step": 630 + }, + { + "epoch": 0.03795488721804511, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.5532, + "step": 631 + }, + { + "epoch": 0.03801503759398496, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.5828, + "step": 632 + }, + { + "epoch": 0.03807518796992481, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.6149, + "step": 633 + }, + { + "epoch": 0.03813533834586466, + "grad_norm": 0.369140625, + "learning_rate": 3e-05, + "loss": 1.5461, + "step": 634 + }, + { + "epoch": 0.03819548872180451, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.587, + "step": 635 + }, + { + "epoch": 0.03825563909774436, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.6225, + "step": 636 + }, + { + "epoch": 0.03831578947368421, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.5994, + "step": 637 + }, + { + "epoch": 0.03837593984962406, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.6033, + "step": 638 + }, + { + "epoch": 0.03843609022556391, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.6159, + "step": 639 + }, + { + "epoch": 0.03849624060150376, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5334, + "step": 640 + }, + { + "epoch": 0.03855639097744361, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.6326, + "step": 641 + }, + { + "epoch": 0.03861654135338346, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.6308, + "step": 642 + }, + { + "epoch": 0.03867669172932331, + "grad_norm": 0.4140625, + "learning_rate": 3e-05, + "loss": 1.5955, + "step": 643 + }, + { + "epoch": 0.03873684210526316, + "grad_norm": 0.3359375, + "learning_rate": 3e-05, + "loss": 1.6097, + "step": 644 + }, + { + "epoch": 0.03879699248120301, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.5355, + "step": 645 + }, + { + "epoch": 0.038857142857142854, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.5362, + "step": 646 + }, + { + "epoch": 0.038917293233082705, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.5472, + "step": 647 + }, + { + "epoch": 0.038977443609022555, + "grad_norm": 0.326171875, + "learning_rate": 3e-05, + "loss": 1.579, + "step": 648 + }, + { + "epoch": 0.039037593984962406, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.5727, + "step": 649 + }, + { + "epoch": 0.039097744360902256, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.5925, + "step": 650 + }, + { + "epoch": 0.03915789473684211, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.6364, + "step": 651 + }, + { + "epoch": 0.03921804511278196, + "grad_norm": 0.328125, + "learning_rate": 3e-05, + "loss": 1.6038, + "step": 652 + }, + { + "epoch": 0.03927819548872181, + "grad_norm": 0.4140625, + "learning_rate": 3e-05, + "loss": 1.5942, + "step": 653 + }, + { + "epoch": 0.03933834586466165, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.5818, + "step": 654 + }, + { + "epoch": 0.0393984962406015, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.6557, + "step": 655 + }, + { + "epoch": 0.03945864661654135, + "grad_norm": 0.37109375, + "learning_rate": 3e-05, + "loss": 1.5844, + "step": 656 + }, + { + "epoch": 0.0395187969924812, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5899, + "step": 657 + }, + { + "epoch": 0.039578947368421054, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.5905, + "step": 658 + }, + { + "epoch": 0.039639097744360904, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.6498, + "step": 659 + }, + { + "epoch": 0.039699248120300755, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.5842, + "step": 660 + }, + { + "epoch": 0.0397593984962406, + "grad_norm": 0.376953125, + "learning_rate": 3e-05, + "loss": 1.6203, + "step": 661 + }, + { + "epoch": 0.03981954887218045, + "grad_norm": 0.400390625, + "learning_rate": 3e-05, + "loss": 1.5531, + "step": 662 + }, + { + "epoch": 0.0398796992481203, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.6282, + "step": 663 + }, + { + "epoch": 0.03993984962406015, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.6372, + "step": 664 + }, + { + "epoch": 0.04, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.6309, + "step": 665 + }, + { + "epoch": 0.04006015037593985, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.7135, + "step": 666 + }, + { + "epoch": 0.0401203007518797, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.5931, + "step": 667 + }, + { + "epoch": 0.040180451127819546, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.6017, + "step": 668 + }, + { + "epoch": 0.040240601503759396, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.5746, + "step": 669 + }, + { + "epoch": 0.04030075187969925, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.5716, + "step": 670 + }, + { + "epoch": 0.0403609022556391, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.6567, + "step": 671 + }, + { + "epoch": 0.04042105263157895, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5524, + "step": 672 + }, + { + "epoch": 0.0404812030075188, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.6688, + "step": 673 + }, + { + "epoch": 0.04054135338345865, + "grad_norm": 0.4375, + "learning_rate": 3e-05, + "loss": 1.6313, + "step": 674 + }, + { + "epoch": 0.0406015037593985, + "grad_norm": 0.31640625, + "learning_rate": 3e-05, + "loss": 1.6071, + "step": 675 + }, + { + "epoch": 0.04066165413533834, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.6303, + "step": 676 + }, + { + "epoch": 0.040721804511278194, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.5662, + "step": 677 + }, + { + "epoch": 0.040781954887218044, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.609, + "step": 678 + }, + { + "epoch": 0.040842105263157895, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.5578, + "step": 679 + }, + { + "epoch": 0.040902255639097745, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.6281, + "step": 680 + }, + { + "epoch": 0.040962406015037596, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.545, + "step": 681 + }, + { + "epoch": 0.04102255639097745, + "grad_norm": 0.4140625, + "learning_rate": 3e-05, + "loss": 1.6213, + "step": 682 + }, + { + "epoch": 0.04108270676691729, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.5941, + "step": 683 + }, + { + "epoch": 0.04114285714285714, + "grad_norm": 0.42578125, + "learning_rate": 3e-05, + "loss": 1.5683, + "step": 684 + }, + { + "epoch": 0.04120300751879699, + "grad_norm": 0.38671875, + "learning_rate": 3e-05, + "loss": 1.5981, + "step": 685 + }, + { + "epoch": 0.04126315789473684, + "grad_norm": 0.40625, + "learning_rate": 3e-05, + "loss": 1.5771, + "step": 686 + }, + { + "epoch": 0.04132330827067669, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.6468, + "step": 687 + }, + { + "epoch": 0.04138345864661654, + "grad_norm": 0.359375, + "learning_rate": 3e-05, + "loss": 1.6092, + "step": 688 + }, + { + "epoch": 0.041443609022556394, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.549, + "step": 689 + }, + { + "epoch": 0.04150375939849624, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4956, + "step": 690 + }, + { + "epoch": 0.04156390977443609, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.6122, + "step": 691 + }, + { + "epoch": 0.04162406015037594, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.5438, + "step": 692 + }, + { + "epoch": 0.04168421052631579, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.5989, + "step": 693 + }, + { + "epoch": 0.04174436090225564, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.5737, + "step": 694 + }, + { + "epoch": 0.04180451127819549, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.7, + "step": 695 + }, + { + "epoch": 0.04186466165413534, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.6082, + "step": 696 + }, + { + "epoch": 0.04192481203007519, + "grad_norm": 0.400390625, + "learning_rate": 3e-05, + "loss": 1.5859, + "step": 697 + }, + { + "epoch": 0.041984962406015035, + "grad_norm": 0.34375, + "learning_rate": 3e-05, + "loss": 1.6139, + "step": 698 + }, + { + "epoch": 0.042045112781954885, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.6382, + "step": 699 + }, + { + "epoch": 0.042105263157894736, + "grad_norm": 0.3671875, + "learning_rate": 3e-05, + "loss": 1.572, + "step": 700 + }, + { + "epoch": 0.042165413533834586, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.6076, + "step": 701 + }, + { + "epoch": 0.04222556390977444, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.5642, + "step": 702 + }, + { + "epoch": 0.04228571428571429, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.5516, + "step": 703 + }, + { + "epoch": 0.04234586466165414, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.5742, + "step": 704 + }, + { + "epoch": 0.04240601503759398, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.6308, + "step": 705 + }, + { + "epoch": 0.04246616541353383, + "grad_norm": 0.37109375, + "learning_rate": 3e-05, + "loss": 1.5839, + "step": 706 + }, + { + "epoch": 0.04252631578947368, + "grad_norm": 0.42578125, + "learning_rate": 3e-05, + "loss": 1.545, + "step": 707 + }, + { + "epoch": 0.04258646616541353, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5554, + "step": 708 + }, + { + "epoch": 0.042646616541353384, + "grad_norm": 0.390625, + "learning_rate": 3e-05, + "loss": 1.6094, + "step": 709 + }, + { + "epoch": 0.042706766917293235, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.6047, + "step": 710 + }, + { + "epoch": 0.042766917293233085, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.5893, + "step": 711 + }, + { + "epoch": 0.042827067669172936, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.6179, + "step": 712 + }, + { + "epoch": 0.04288721804511278, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.5604, + "step": 713 + }, + { + "epoch": 0.04294736842105263, + "grad_norm": 0.37109375, + "learning_rate": 3e-05, + "loss": 1.5993, + "step": 714 + }, + { + "epoch": 0.04300751879699248, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.6103, + "step": 715 + }, + { + "epoch": 0.04306766917293233, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.6473, + "step": 716 + }, + { + "epoch": 0.04312781954887218, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.6031, + "step": 717 + }, + { + "epoch": 0.04318796992481203, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.5567, + "step": 718 + }, + { + "epoch": 0.04324812030075188, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.5641, + "step": 719 + }, + { + "epoch": 0.043308270676691726, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.5617, + "step": 720 + }, + { + "epoch": 0.04336842105263158, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5458, + "step": 721 + }, + { + "epoch": 0.04342857142857143, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.553, + "step": 722 + }, + { + "epoch": 0.04348872180451128, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.6354, + "step": 723 + }, + { + "epoch": 0.04354887218045113, + "grad_norm": 0.376953125, + "learning_rate": 3e-05, + "loss": 1.5776, + "step": 724 + }, + { + "epoch": 0.04360902255639098, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.506, + "step": 725 + }, + { + "epoch": 0.04366917293233083, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.6246, + "step": 726 + }, + { + "epoch": 0.04372932330827067, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5258, + "step": 727 + }, + { + "epoch": 0.043789473684210524, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5348, + "step": 728 + }, + { + "epoch": 0.043849624060150375, + "grad_norm": 0.3984375, + "learning_rate": 3e-05, + "loss": 1.6225, + "step": 729 + }, + { + "epoch": 0.043909774436090225, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.5909, + "step": 730 + }, + { + "epoch": 0.043969924812030076, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.6772, + "step": 731 + }, + { + "epoch": 0.044030075187969926, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.6616, + "step": 732 + }, + { + "epoch": 0.04409022556390978, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.5925, + "step": 733 + }, + { + "epoch": 0.04415037593984963, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.707, + "step": 734 + }, + { + "epoch": 0.04421052631578947, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.4326, + "step": 735 + }, + { + "epoch": 0.04427067669172932, + "grad_norm": 0.333984375, + "learning_rate": 3e-05, + "loss": 1.5877, + "step": 736 + }, + { + "epoch": 0.04433082706766917, + "grad_norm": 0.4140625, + "learning_rate": 3e-05, + "loss": 1.6351, + "step": 737 + }, + { + "epoch": 0.04439097744360902, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.5819, + "step": 738 + }, + { + "epoch": 0.04445112781954887, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.6027, + "step": 739 + }, + { + "epoch": 0.044511278195488724, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.6491, + "step": 740 + }, + { + "epoch": 0.044571428571428574, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.5739, + "step": 741 + }, + { + "epoch": 0.04463157894736842, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.5418, + "step": 742 + }, + { + "epoch": 0.04469172932330827, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.5716, + "step": 743 + }, + { + "epoch": 0.04475187969924812, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.5058, + "step": 744 + }, + { + "epoch": 0.04481203007518797, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5515, + "step": 745 + }, + { + "epoch": 0.04487218045112782, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4625, + "step": 746 + }, + { + "epoch": 0.04493233082706767, + "grad_norm": 0.3984375, + "learning_rate": 3e-05, + "loss": 1.6198, + "step": 747 + }, + { + "epoch": 0.04499248120300752, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.6018, + "step": 748 + }, + { + "epoch": 0.045052631578947365, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.5788, + "step": 749 + }, + { + "epoch": 0.045112781954887216, + "grad_norm": 0.40234375, + "learning_rate": 3e-05, + "loss": 1.5325, + "step": 750 + }, + { + "epoch": 0.045172932330827066, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4658, + "step": 751 + }, + { + "epoch": 0.04523308270676692, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.5718, + "step": 752 + }, + { + "epoch": 0.04529323308270677, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.517, + "step": 753 + }, + { + "epoch": 0.04535338345864662, + "grad_norm": 0.38671875, + "learning_rate": 3e-05, + "loss": 1.7609, + "step": 754 + }, + { + "epoch": 0.04541353383458647, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.6149, + "step": 755 + }, + { + "epoch": 0.04547368421052632, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5198, + "step": 756 + }, + { + "epoch": 0.04553383458646616, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.6265, + "step": 757 + }, + { + "epoch": 0.04559398496240601, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5763, + "step": 758 + }, + { + "epoch": 0.045654135338345864, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.5923, + "step": 759 + }, + { + "epoch": 0.045714285714285714, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.5372, + "step": 760 + }, + { + "epoch": 0.045774436090225565, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.6552, + "step": 761 + }, + { + "epoch": 0.045834586466165415, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.5814, + "step": 762 + }, + { + "epoch": 0.045894736842105266, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.5361, + "step": 763 + }, + { + "epoch": 0.04595488721804511, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.5784, + "step": 764 + }, + { + "epoch": 0.04601503759398496, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.5658, + "step": 765 + }, + { + "epoch": 0.04607518796992481, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.5892, + "step": 766 + }, + { + "epoch": 0.04613533834586466, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.5676, + "step": 767 + }, + { + "epoch": 0.04619548872180451, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.5926, + "step": 768 + }, + { + "epoch": 0.04625563909774436, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.6767, + "step": 769 + }, + { + "epoch": 0.04631578947368421, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.5838, + "step": 770 + }, + { + "epoch": 0.046375939849624064, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5916, + "step": 771 + }, + { + "epoch": 0.04643609022556391, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.5158, + "step": 772 + }, + { + "epoch": 0.04649624060150376, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.5851, + "step": 773 + }, + { + "epoch": 0.04655639097744361, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.5438, + "step": 774 + }, + { + "epoch": 0.04661654135338346, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.569, + "step": 775 + }, + { + "epoch": 0.04667669172932331, + "grad_norm": 0.32421875, + "learning_rate": 3e-05, + "loss": 1.5878, + "step": 776 + }, + { + "epoch": 0.04673684210526316, + "grad_norm": 0.42578125, + "learning_rate": 3e-05, + "loss": 1.5774, + "step": 777 + }, + { + "epoch": 0.04679699248120301, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.6132, + "step": 778 + }, + { + "epoch": 0.046857142857142854, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.5658, + "step": 779 + }, + { + "epoch": 0.046917293233082705, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.604, + "step": 780 + }, + { + "epoch": 0.046977443609022555, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.6235, + "step": 781 + }, + { + "epoch": 0.047037593984962406, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.6035, + "step": 782 + }, + { + "epoch": 0.047097744360902256, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.5155, + "step": 783 + }, + { + "epoch": 0.04715789473684211, + "grad_norm": 0.373046875, + "learning_rate": 3e-05, + "loss": 1.5602, + "step": 784 + }, + { + "epoch": 0.04721804511278196, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.5024, + "step": 785 + }, + { + "epoch": 0.0472781954887218, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.6589, + "step": 786 + }, + { + "epoch": 0.04733834586466165, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.4958, + "step": 787 + }, + { + "epoch": 0.0473984962406015, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.6447, + "step": 788 + }, + { + "epoch": 0.04745864661654135, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.5581, + "step": 789 + }, + { + "epoch": 0.0475187969924812, + "grad_norm": 0.361328125, + "learning_rate": 3e-05, + "loss": 1.5968, + "step": 790 + }, + { + "epoch": 0.047578947368421054, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.597, + "step": 791 + }, + { + "epoch": 0.047639097744360905, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.6568, + "step": 792 + }, + { + "epoch": 0.047699248120300755, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.5264, + "step": 793 + }, + { + "epoch": 0.0477593984962406, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.5252, + "step": 794 + }, + { + "epoch": 0.04781954887218045, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5486, + "step": 795 + }, + { + "epoch": 0.0478796992481203, + "grad_norm": 0.3984375, + "learning_rate": 3e-05, + "loss": 1.558, + "step": 796 + }, + { + "epoch": 0.04793984962406015, + "grad_norm": 0.392578125, + "learning_rate": 3e-05, + "loss": 1.5699, + "step": 797 + }, + { + "epoch": 0.048, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.5457, + "step": 798 + }, + { + "epoch": 0.04806015037593985, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.5671, + "step": 799 + }, + { + "epoch": 0.0481203007518797, + "grad_norm": 0.35546875, + "learning_rate": 3e-05, + "loss": 1.5365, + "step": 800 + }, + { + "epoch": 0.048180451127819546, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.6139, + "step": 801 + }, + { + "epoch": 0.048240601503759396, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.5638, + "step": 802 + }, + { + "epoch": 0.04830075187969925, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5136, + "step": 803 + }, + { + "epoch": 0.0483609022556391, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.5364, + "step": 804 + }, + { + "epoch": 0.04842105263157895, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.597, + "step": 805 + }, + { + "epoch": 0.0484812030075188, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4832, + "step": 806 + }, + { + "epoch": 0.04854135338345865, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5959, + "step": 807 + }, + { + "epoch": 0.0486015037593985, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.605, + "step": 808 + }, + { + "epoch": 0.04866165413533834, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.5783, + "step": 809 + }, + { + "epoch": 0.048721804511278194, + "grad_norm": 0.38671875, + "learning_rate": 3e-05, + "loss": 1.5807, + "step": 810 + }, + { + "epoch": 0.048781954887218044, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.547, + "step": 811 + }, + { + "epoch": 0.048842105263157895, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.5051, + "step": 812 + }, + { + "epoch": 0.048902255639097746, + "grad_norm": 0.40234375, + "learning_rate": 3e-05, + "loss": 1.6337, + "step": 813 + }, + { + "epoch": 0.048962406015037596, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.5895, + "step": 814 + }, + { + "epoch": 0.04902255639097745, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.573, + "step": 815 + }, + { + "epoch": 0.04908270676691729, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.553, + "step": 816 + }, + { + "epoch": 0.04914285714285714, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.5181, + "step": 817 + }, + { + "epoch": 0.04920300751879699, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.66, + "step": 818 + }, + { + "epoch": 0.04926315789473684, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.6194, + "step": 819 + }, + { + "epoch": 0.04932330827067669, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.5712, + "step": 820 + }, + { + "epoch": 0.04938345864661654, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.524, + "step": 821 + }, + { + "epoch": 0.049443609022556394, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.4761, + "step": 822 + }, + { + "epoch": 0.04950375939849624, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.4637, + "step": 823 + }, + { + "epoch": 0.04956390977443609, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.5775, + "step": 824 + }, + { + "epoch": 0.04962406015037594, + "grad_norm": 0.38671875, + "learning_rate": 3e-05, + "loss": 1.5741, + "step": 825 + }, + { + "epoch": 0.04968421052631579, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5447, + "step": 826 + }, + { + "epoch": 0.04974436090225564, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.6073, + "step": 827 + }, + { + "epoch": 0.04980451127819549, + "grad_norm": 0.3984375, + "learning_rate": 3e-05, + "loss": 1.5719, + "step": 828 + }, + { + "epoch": 0.04986466165413534, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5493, + "step": 829 + }, + { + "epoch": 0.04992481203007519, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.5911, + "step": 830 + }, + { + "epoch": 0.049984962406015035, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.583, + "step": 831 + }, + { + "epoch": 0.050045112781954885, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.5714, + "step": 832 + }, + { + "epoch": 0.050105263157894736, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.5607, + "step": 833 + }, + { + "epoch": 0.05016541353383459, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.6194, + "step": 834 + }, + { + "epoch": 0.05022556390977444, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.5624, + "step": 835 + }, + { + "epoch": 0.05028571428571429, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.6174, + "step": 836 + }, + { + "epoch": 0.05034586466165414, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.5078, + "step": 837 + }, + { + "epoch": 0.05040601503759398, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.5822, + "step": 838 + }, + { + "epoch": 0.05046616541353383, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.5215, + "step": 839 + }, + { + "epoch": 0.05052631578947368, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.6168, + "step": 840 + }, + { + "epoch": 0.050586466165413534, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.544, + "step": 841 + }, + { + "epoch": 0.050646616541353384, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.5481, + "step": 842 + }, + { + "epoch": 0.050706766917293235, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.6038, + "step": 843 + }, + { + "epoch": 0.050766917293233085, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.6222, + "step": 844 + }, + { + "epoch": 0.05082706766917293, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.5438, + "step": 845 + }, + { + "epoch": 0.05088721804511278, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.5801, + "step": 846 + }, + { + "epoch": 0.05094736842105263, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.5071, + "step": 847 + }, + { + "epoch": 0.05100751879699248, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.5772, + "step": 848 + }, + { + "epoch": 0.05106766917293233, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.5633, + "step": 849 + }, + { + "epoch": 0.05112781954887218, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.5525, + "step": 850 + }, + { + "epoch": 0.05118796992481203, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.5868, + "step": 851 + }, + { + "epoch": 0.05124812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5963, + "step": 852 + }, + { + "epoch": 0.05130827067669173, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.6157, + "step": 853 + }, + { + "epoch": 0.05136842105263158, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.6228, + "step": 854 + }, + { + "epoch": 0.05142857142857143, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.5751, + "step": 855 + }, + { + "epoch": 0.05148872180451128, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.5108, + "step": 856 + }, + { + "epoch": 0.05154887218045113, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4613, + "step": 857 + }, + { + "epoch": 0.05160902255639098, + "grad_norm": 0.4375, + "learning_rate": 3e-05, + "loss": 1.5465, + "step": 858 + }, + { + "epoch": 0.05166917293233083, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.5914, + "step": 859 + }, + { + "epoch": 0.051729323308270674, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.6246, + "step": 860 + }, + { + "epoch": 0.051789473684210524, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.5611, + "step": 861 + }, + { + "epoch": 0.051849624060150375, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5449, + "step": 862 + }, + { + "epoch": 0.051909774436090225, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.5867, + "step": 863 + }, + { + "epoch": 0.051969924812030076, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.5459, + "step": 864 + }, + { + "epoch": 0.052030075187969926, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.5839, + "step": 865 + }, + { + "epoch": 0.05209022556390978, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.509, + "step": 866 + }, + { + "epoch": 0.05215037593984963, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.5394, + "step": 867 + }, + { + "epoch": 0.05221052631578947, + "grad_norm": 0.40625, + "learning_rate": 3e-05, + "loss": 1.5337, + "step": 868 + }, + { + "epoch": 0.05227067669172932, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.5194, + "step": 869 + }, + { + "epoch": 0.05233082706766917, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5467, + "step": 870 + }, + { + "epoch": 0.05239097744360902, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.5897, + "step": 871 + }, + { + "epoch": 0.05245112781954887, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.5688, + "step": 872 + }, + { + "epoch": 0.052511278195488724, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.5384, + "step": 873 + }, + { + "epoch": 0.052571428571428575, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.5851, + "step": 874 + }, + { + "epoch": 0.05263157894736842, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.5064, + "step": 875 + }, + { + "epoch": 0.05269172932330827, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.7505, + "step": 876 + }, + { + "epoch": 0.05275187969924812, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.5325, + "step": 877 + }, + { + "epoch": 0.05281203007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.5577, + "step": 878 + }, + { + "epoch": 0.05287218045112782, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5315, + "step": 879 + }, + { + "epoch": 0.05293233082706767, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.6023, + "step": 880 + }, + { + "epoch": 0.05299248120300752, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.6366, + "step": 881 + }, + { + "epoch": 0.053052631578947365, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.5252, + "step": 882 + }, + { + "epoch": 0.053112781954887216, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.5395, + "step": 883 + }, + { + "epoch": 0.053172932330827066, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.5865, + "step": 884 + }, + { + "epoch": 0.05323308270676692, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.5765, + "step": 885 + }, + { + "epoch": 0.05329323308270677, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.5874, + "step": 886 + }, + { + "epoch": 0.05335338345864662, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.5948, + "step": 887 + }, + { + "epoch": 0.05341353383458647, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.6613, + "step": 888 + }, + { + "epoch": 0.05347368421052632, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.5649, + "step": 889 + }, + { + "epoch": 0.05353383458646616, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.5389, + "step": 890 + }, + { + "epoch": 0.05359398496240601, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.5559, + "step": 891 + }, + { + "epoch": 0.053654135338345864, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.6374, + "step": 892 + }, + { + "epoch": 0.053714285714285714, + "grad_norm": 0.37890625, + "learning_rate": 3e-05, + "loss": 1.5908, + "step": 893 + }, + { + "epoch": 0.053774436090225565, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.5482, + "step": 894 + }, + { + "epoch": 0.053834586466165416, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.6233, + "step": 895 + }, + { + "epoch": 0.053894736842105266, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.5417, + "step": 896 + }, + { + "epoch": 0.05395488721804511, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5227, + "step": 897 + }, + { + "epoch": 0.05401503759398496, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.52, + "step": 898 + }, + { + "epoch": 0.05407518796992481, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.5493, + "step": 899 + }, + { + "epoch": 0.05413533834586466, + "grad_norm": 0.462890625, + "learning_rate": 3e-05, + "loss": 1.6009, + "step": 900 + }, + { + "epoch": 0.05419548872180451, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.5475, + "step": 901 + }, + { + "epoch": 0.05425563909774436, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5445, + "step": 902 + }, + { + "epoch": 0.05431578947368421, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4889, + "step": 903 + }, + { + "epoch": 0.05437593984962406, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5657, + "step": 904 + }, + { + "epoch": 0.05443609022556391, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.5547, + "step": 905 + }, + { + "epoch": 0.05449624060150376, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.5648, + "step": 906 + }, + { + "epoch": 0.05455639097744361, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4779, + "step": 907 + }, + { + "epoch": 0.05461654135338346, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.5551, + "step": 908 + }, + { + "epoch": 0.05467669172932331, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5358, + "step": 909 + }, + { + "epoch": 0.05473684210526316, + "grad_norm": 0.421875, + "learning_rate": 3e-05, + "loss": 1.6016, + "step": 910 + }, + { + "epoch": 0.05479699248120301, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.5603, + "step": 911 + }, + { + "epoch": 0.054857142857142854, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.5331, + "step": 912 + }, + { + "epoch": 0.054917293233082705, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.5951, + "step": 913 + }, + { + "epoch": 0.054977443609022555, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4936, + "step": 914 + }, + { + "epoch": 0.055037593984962406, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.5377, + "step": 915 + }, + { + "epoch": 0.05509774436090226, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.482, + "step": 916 + }, + { + "epoch": 0.05515789473684211, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4746, + "step": 917 + }, + { + "epoch": 0.05521804511278196, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.4804, + "step": 918 + }, + { + "epoch": 0.0552781954887218, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.5483, + "step": 919 + }, + { + "epoch": 0.05533834586466165, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.536, + "step": 920 + }, + { + "epoch": 0.0553984962406015, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4883, + "step": 921 + }, + { + "epoch": 0.05545864661654135, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.5833, + "step": 922 + }, + { + "epoch": 0.055518796992481204, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.599, + "step": 923 + }, + { + "epoch": 0.055578947368421054, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.617, + "step": 924 + }, + { + "epoch": 0.055639097744360905, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.524, + "step": 925 + }, + { + "epoch": 0.055699248120300755, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5581, + "step": 926 + }, + { + "epoch": 0.0557593984962406, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.5218, + "step": 927 + }, + { + "epoch": 0.05581954887218045, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.5479, + "step": 928 + }, + { + "epoch": 0.0558796992481203, + "grad_norm": 0.396484375, + "learning_rate": 3e-05, + "loss": 1.5586, + "step": 929 + }, + { + "epoch": 0.05593984962406015, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.6424, + "step": 930 + }, + { + "epoch": 0.056, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.5769, + "step": 931 + }, + { + "epoch": 0.05606015037593985, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.511, + "step": 932 + }, + { + "epoch": 0.0561203007518797, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.5565, + "step": 933 + }, + { + "epoch": 0.056180451127819546, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5138, + "step": 934 + }, + { + "epoch": 0.056240601503759396, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4898, + "step": 935 + }, + { + "epoch": 0.05630075187969925, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5482, + "step": 936 + }, + { + "epoch": 0.0563609022556391, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5737, + "step": 937 + }, + { + "epoch": 0.05642105263157895, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.5959, + "step": 938 + }, + { + "epoch": 0.0564812030075188, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.5565, + "step": 939 + }, + { + "epoch": 0.05654135338345865, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.5869, + "step": 940 + }, + { + "epoch": 0.05660150375939849, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4814, + "step": 941 + }, + { + "epoch": 0.056661654135338343, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.5298, + "step": 942 + }, + { + "epoch": 0.056721804511278194, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.5262, + "step": 943 + }, + { + "epoch": 0.056781954887218045, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.5585, + "step": 944 + }, + { + "epoch": 0.056842105263157895, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.5673, + "step": 945 + }, + { + "epoch": 0.056902255639097746, + "grad_norm": 0.462890625, + "learning_rate": 3e-05, + "loss": 1.5227, + "step": 946 + }, + { + "epoch": 0.056962406015037596, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.7153, + "step": 947 + }, + { + "epoch": 0.05702255639097745, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.5247, + "step": 948 + }, + { + "epoch": 0.05708270676691729, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.6487, + "step": 949 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.5624, + "step": 950 + }, + { + "epoch": 0.05720300751879699, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.5355, + "step": 951 + }, + { + "epoch": 0.05726315789473684, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5393, + "step": 952 + }, + { + "epoch": 0.05732330827067669, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5835, + "step": 953 + }, + { + "epoch": 0.05738345864661654, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.5173, + "step": 954 + }, + { + "epoch": 0.057443609022556394, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.578, + "step": 955 + }, + { + "epoch": 0.05750375939849624, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4903, + "step": 956 + }, + { + "epoch": 0.05756390977443609, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.5227, + "step": 957 + }, + { + "epoch": 0.05762406015037594, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.52, + "step": 958 + }, + { + "epoch": 0.05768421052631579, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4561, + "step": 959 + }, + { + "epoch": 0.05774436090225564, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5509, + "step": 960 + }, + { + "epoch": 0.05780451127819549, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.5184, + "step": 961 + }, + { + "epoch": 0.05786466165413534, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.6052, + "step": 962 + }, + { + "epoch": 0.057924812030075185, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.5692, + "step": 963 + }, + { + "epoch": 0.057984962406015035, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.5049, + "step": 964 + }, + { + "epoch": 0.058045112781954886, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.5079, + "step": 965 + }, + { + "epoch": 0.058105263157894736, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.5893, + "step": 966 + }, + { + "epoch": 0.05816541353383459, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.5093, + "step": 967 + }, + { + "epoch": 0.05822556390977444, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.5555, + "step": 968 + }, + { + "epoch": 0.05828571428571429, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.5841, + "step": 969 + }, + { + "epoch": 0.05834586466165414, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.5382, + "step": 970 + }, + { + "epoch": 0.05840601503759398, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4447, + "step": 971 + }, + { + "epoch": 0.05846616541353383, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.4989, + "step": 972 + }, + { + "epoch": 0.05852631578947368, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.5402, + "step": 973 + }, + { + "epoch": 0.058586466165413534, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4726, + "step": 974 + }, + { + "epoch": 0.058646616541353384, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.5161, + "step": 975 + }, + { + "epoch": 0.058706766917293235, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4837, + "step": 976 + }, + { + "epoch": 0.058766917293233085, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.5027, + "step": 977 + }, + { + "epoch": 0.05882706766917293, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.616, + "step": 978 + }, + { + "epoch": 0.05888721804511278, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4989, + "step": 979 + }, + { + "epoch": 0.05894736842105263, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5584, + "step": 980 + }, + { + "epoch": 0.05900751879699248, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.5899, + "step": 981 + }, + { + "epoch": 0.05906766917293233, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.5253, + "step": 982 + }, + { + "epoch": 0.05912781954887218, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.5444, + "step": 983 + }, + { + "epoch": 0.05918796992481203, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.5302, + "step": 984 + }, + { + "epoch": 0.05924812030075188, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.5013, + "step": 985 + }, + { + "epoch": 0.05930827067669173, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.5304, + "step": 986 + }, + { + "epoch": 0.05936842105263158, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5735, + "step": 987 + }, + { + "epoch": 0.05942857142857143, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.5834, + "step": 988 + }, + { + "epoch": 0.05948872180451128, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4555, + "step": 989 + }, + { + "epoch": 0.05954887218045113, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.5081, + "step": 990 + }, + { + "epoch": 0.05960902255639098, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5394, + "step": 991 + }, + { + "epoch": 0.05966917293233083, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4906, + "step": 992 + }, + { + "epoch": 0.059729323308270674, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.5442, + "step": 993 + }, + { + "epoch": 0.059789473684210524, + "grad_norm": 0.400390625, + "learning_rate": 3e-05, + "loss": 1.524, + "step": 994 + }, + { + "epoch": 0.059849624060150375, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.5745, + "step": 995 + }, + { + "epoch": 0.059909774436090225, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.4913, + "step": 996 + }, + { + "epoch": 0.059969924812030076, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.514, + "step": 997 + }, + { + "epoch": 0.06003007518796993, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.5075, + "step": 998 + }, + { + "epoch": 0.06009022556390978, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5588, + "step": 999 + }, + { + "epoch": 0.06015037593984962, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.5427, + "step": 1000 + }, + { + "epoch": 0.06021052631578947, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.5188, + "step": 1001 + }, + { + "epoch": 0.06027067669172932, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.481, + "step": 1002 + }, + { + "epoch": 0.06033082706766917, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.4748, + "step": 1003 + }, + { + "epoch": 0.06039097744360902, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.5544, + "step": 1004 + }, + { + "epoch": 0.060451127819548874, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.5326, + "step": 1005 + }, + { + "epoch": 0.060511278195488724, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4882, + "step": 1006 + }, + { + "epoch": 0.060571428571428575, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.5961, + "step": 1007 + }, + { + "epoch": 0.06063157894736842, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4205, + "step": 1008 + }, + { + "epoch": 0.06069172932330827, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.5111, + "step": 1009 + }, + { + "epoch": 0.06075187969924812, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4931, + "step": 1010 + }, + { + "epoch": 0.06081203007518797, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.5053, + "step": 1011 + }, + { + "epoch": 0.06087218045112782, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.5349, + "step": 1012 + }, + { + "epoch": 0.06093233082706767, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.536, + "step": 1013 + }, + { + "epoch": 0.06099248120300752, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.5284, + "step": 1014 + }, + { + "epoch": 0.061052631578947365, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.5219, + "step": 1015 + }, + { + "epoch": 0.061112781954887216, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.5032, + "step": 1016 + }, + { + "epoch": 0.061172932330827066, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5642, + "step": 1017 + }, + { + "epoch": 0.06123308270676692, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.5595, + "step": 1018 + }, + { + "epoch": 0.06129323308270677, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.5658, + "step": 1019 + }, + { + "epoch": 0.06135338345864662, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.5524, + "step": 1020 + }, + { + "epoch": 0.06141353383458647, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.4697, + "step": 1021 + }, + { + "epoch": 0.06147368421052631, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5012, + "step": 1022 + }, + { + "epoch": 0.06153383458646616, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.5572, + "step": 1023 + }, + { + "epoch": 0.06159398496240601, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4781, + "step": 1024 + }, + { + "epoch": 0.061654135338345864, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4169, + "step": 1025 + }, + { + "epoch": 0.061714285714285715, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.5556, + "step": 1026 + }, + { + "epoch": 0.061774436090225565, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.5682, + "step": 1027 + }, + { + "epoch": 0.061834586466165416, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.4449, + "step": 1028 + }, + { + "epoch": 0.061894736842105266, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.5749, + "step": 1029 + }, + { + "epoch": 0.06195488721804511, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5066, + "step": 1030 + }, + { + "epoch": 0.06201503759398496, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.5416, + "step": 1031 + }, + { + "epoch": 0.06207518796992481, + "grad_norm": 0.462890625, + "learning_rate": 3e-05, + "loss": 1.5207, + "step": 1032 + }, + { + "epoch": 0.06213533834586466, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.5323, + "step": 1033 + }, + { + "epoch": 0.06219548872180451, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.5629, + "step": 1034 + }, + { + "epoch": 0.06225563909774436, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.5456, + "step": 1035 + }, + { + "epoch": 0.06231578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.589, + "step": 1036 + }, + { + "epoch": 0.06237593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.5218, + "step": 1037 + }, + { + "epoch": 0.06243609022556391, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.504, + "step": 1038 + }, + { + "epoch": 0.06249624060150376, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.602, + "step": 1039 + }, + { + "epoch": 0.06255639097744362, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.5125, + "step": 1040 + }, + { + "epoch": 0.06261654135338346, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.4919, + "step": 1041 + }, + { + "epoch": 0.0626766917293233, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.552, + "step": 1042 + }, + { + "epoch": 0.06273684210526316, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4929, + "step": 1043 + }, + { + "epoch": 0.062796992481203, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.4626, + "step": 1044 + }, + { + "epoch": 0.06285714285714286, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.5476, + "step": 1045 + }, + { + "epoch": 0.0629172932330827, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.5191, + "step": 1046 + }, + { + "epoch": 0.06297744360902256, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.4619, + "step": 1047 + }, + { + "epoch": 0.0630375939849624, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.5826, + "step": 1048 + }, + { + "epoch": 0.06309774436090225, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.5049, + "step": 1049 + }, + { + "epoch": 0.06315789473684211, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.5032, + "step": 1050 + }, + { + "epoch": 0.06321804511278195, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5619, + "step": 1051 + }, + { + "epoch": 0.06327819548872181, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5897, + "step": 1052 + }, + { + "epoch": 0.06333834586466165, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5712, + "step": 1053 + }, + { + "epoch": 0.06339849624060151, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.5135, + "step": 1054 + }, + { + "epoch": 0.06345864661654135, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.5405, + "step": 1055 + }, + { + "epoch": 0.0635187969924812, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.5469, + "step": 1056 + }, + { + "epoch": 0.06357894736842105, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.5713, + "step": 1057 + }, + { + "epoch": 0.0636390977443609, + "grad_norm": 0.40625, + "learning_rate": 3e-05, + "loss": 1.5251, + "step": 1058 + }, + { + "epoch": 0.06369924812030076, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.5002, + "step": 1059 + }, + { + "epoch": 0.0637593984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.5026, + "step": 1060 + }, + { + "epoch": 0.06381954887218046, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.5129, + "step": 1061 + }, + { + "epoch": 0.0638796992481203, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.5454, + "step": 1062 + }, + { + "epoch": 0.06393984962406014, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.5881, + "step": 1063 + }, + { + "epoch": 0.064, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4953, + "step": 1064 + }, + { + "epoch": 0.06406015037593984, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.5381, + "step": 1065 + }, + { + "epoch": 0.0641203007518797, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.4735, + "step": 1066 + }, + { + "epoch": 0.06418045112781955, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.5493, + "step": 1067 + }, + { + "epoch": 0.0642406015037594, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.6153, + "step": 1068 + }, + { + "epoch": 0.06430075187969925, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4901, + "step": 1069 + }, + { + "epoch": 0.06436090225563909, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.5221, + "step": 1070 + }, + { + "epoch": 0.06442105263157895, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.5678, + "step": 1071 + }, + { + "epoch": 0.06448120300751879, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.4468, + "step": 1072 + }, + { + "epoch": 0.06454135338345865, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.4904, + "step": 1073 + }, + { + "epoch": 0.0646015037593985, + "grad_norm": 0.462890625, + "learning_rate": 3e-05, + "loss": 1.5019, + "step": 1074 + }, + { + "epoch": 0.06466165413533835, + "grad_norm": 0.4140625, + "learning_rate": 3e-05, + "loss": 1.6227, + "step": 1075 + }, + { + "epoch": 0.0647218045112782, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4585, + "step": 1076 + }, + { + "epoch": 0.06478195488721805, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3889, + "step": 1077 + }, + { + "epoch": 0.0648421052631579, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5406, + "step": 1078 + }, + { + "epoch": 0.06490225563909774, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.4859, + "step": 1079 + }, + { + "epoch": 0.0649624060150376, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.5055, + "step": 1080 + }, + { + "epoch": 0.06502255639097744, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4392, + "step": 1081 + }, + { + "epoch": 0.0650827067669173, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.4997, + "step": 1082 + }, + { + "epoch": 0.06514285714285714, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.5165, + "step": 1083 + }, + { + "epoch": 0.065203007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.5803, + "step": 1084 + }, + { + "epoch": 0.06526315789473684, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.5219, + "step": 1085 + }, + { + "epoch": 0.06532330827067669, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.4636, + "step": 1086 + }, + { + "epoch": 0.06538345864661654, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.5818, + "step": 1087 + }, + { + "epoch": 0.06544360902255639, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.5395, + "step": 1088 + }, + { + "epoch": 0.06550375939849624, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4926, + "step": 1089 + }, + { + "epoch": 0.06556390977443609, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.4741, + "step": 1090 + }, + { + "epoch": 0.06562406015037595, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.5739, + "step": 1091 + }, + { + "epoch": 0.06568421052631579, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.5611, + "step": 1092 + }, + { + "epoch": 0.06574436090225563, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.6174, + "step": 1093 + }, + { + "epoch": 0.06580451127819549, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.5404, + "step": 1094 + }, + { + "epoch": 0.06586466165413533, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.5083, + "step": 1095 + }, + { + "epoch": 0.06592481203007519, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.5165, + "step": 1096 + }, + { + "epoch": 0.06598496240601504, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.4706, + "step": 1097 + }, + { + "epoch": 0.06604511278195489, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5577, + "step": 1098 + }, + { + "epoch": 0.06610526315789474, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.5297, + "step": 1099 + }, + { + "epoch": 0.06616541353383458, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.5141, + "step": 1100 + }, + { + "epoch": 0.06622556390977444, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.5244, + "step": 1101 + }, + { + "epoch": 0.06628571428571428, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.5166, + "step": 1102 + }, + { + "epoch": 0.06634586466165414, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.639, + "step": 1103 + }, + { + "epoch": 0.06640601503759398, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.5501, + "step": 1104 + }, + { + "epoch": 0.06646616541353384, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.5587, + "step": 1105 + }, + { + "epoch": 0.06652631578947368, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4964, + "step": 1106 + }, + { + "epoch": 0.06658646616541353, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.4914, + "step": 1107 + }, + { + "epoch": 0.06664661654135338, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.5448, + "step": 1108 + }, + { + "epoch": 0.06670676691729323, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.477, + "step": 1109 + }, + { + "epoch": 0.06676691729323309, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.5576, + "step": 1110 + }, + { + "epoch": 0.06682706766917293, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.5531, + "step": 1111 + }, + { + "epoch": 0.06688721804511279, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.4342, + "step": 1112 + }, + { + "epoch": 0.06694736842105263, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4421, + "step": 1113 + }, + { + "epoch": 0.06700751879699249, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.566, + "step": 1114 + }, + { + "epoch": 0.06706766917293233, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4677, + "step": 1115 + }, + { + "epoch": 0.06712781954887218, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5803, + "step": 1116 + }, + { + "epoch": 0.06718796992481203, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4973, + "step": 1117 + }, + { + "epoch": 0.06724812030075188, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.5116, + "step": 1118 + }, + { + "epoch": 0.06730827067669173, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.5356, + "step": 1119 + }, + { + "epoch": 0.06736842105263158, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.5703, + "step": 1120 + }, + { + "epoch": 0.06742857142857143, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.5425, + "step": 1121 + }, + { + "epoch": 0.06748872180451128, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.4901, + "step": 1122 + }, + { + "epoch": 0.06754887218045112, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.6184, + "step": 1123 + }, + { + "epoch": 0.06760902255639098, + "grad_norm": 0.4375, + "learning_rate": 3e-05, + "loss": 1.4349, + "step": 1124 + }, + { + "epoch": 0.06766917293233082, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.5333, + "step": 1125 + }, + { + "epoch": 0.06772932330827068, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.582, + "step": 1126 + }, + { + "epoch": 0.06778947368421052, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.524, + "step": 1127 + }, + { + "epoch": 0.06784962406015038, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.4902, + "step": 1128 + }, + { + "epoch": 0.06790977443609023, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.449, + "step": 1129 + }, + { + "epoch": 0.06796992481203007, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5204, + "step": 1130 + }, + { + "epoch": 0.06803007518796993, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.6725, + "step": 1131 + }, + { + "epoch": 0.06809022556390977, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4983, + "step": 1132 + }, + { + "epoch": 0.06815037593984963, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.4711, + "step": 1133 + }, + { + "epoch": 0.06821052631578947, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.5256, + "step": 1134 + }, + { + "epoch": 0.06827067669172933, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.5939, + "step": 1135 + }, + { + "epoch": 0.06833082706766917, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5091, + "step": 1136 + }, + { + "epoch": 0.06839097744360902, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.5373, + "step": 1137 + }, + { + "epoch": 0.06845112781954887, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4632, + "step": 1138 + }, + { + "epoch": 0.06851127819548872, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.5354, + "step": 1139 + }, + { + "epoch": 0.06857142857142857, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.5398, + "step": 1140 + }, + { + "epoch": 0.06863157894736842, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.5041, + "step": 1141 + }, + { + "epoch": 0.06869172932330828, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4954, + "step": 1142 + }, + { + "epoch": 0.06875187969924812, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.5664, + "step": 1143 + }, + { + "epoch": 0.06881203007518796, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.5304, + "step": 1144 + }, + { + "epoch": 0.06887218045112782, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.5417, + "step": 1145 + }, + { + "epoch": 0.06893233082706766, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.4884, + "step": 1146 + }, + { + "epoch": 0.06899248120300752, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.5037, + "step": 1147 + }, + { + "epoch": 0.06905263157894737, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.5493, + "step": 1148 + }, + { + "epoch": 0.06911278195488722, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.4559, + "step": 1149 + }, + { + "epoch": 0.06917293233082707, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.5361, + "step": 1150 + }, + { + "epoch": 0.06923308270676692, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.5084, + "step": 1151 + }, + { + "epoch": 0.06929323308270677, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.5755, + "step": 1152 + }, + { + "epoch": 0.06935338345864661, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.5416, + "step": 1153 + }, + { + "epoch": 0.06941353383458647, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4643, + "step": 1154 + }, + { + "epoch": 0.06947368421052631, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.6036, + "step": 1155 + }, + { + "epoch": 0.06953383458646617, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.4967, + "step": 1156 + }, + { + "epoch": 0.06959398496240601, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.5596, + "step": 1157 + }, + { + "epoch": 0.06965413533834587, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4784, + "step": 1158 + }, + { + "epoch": 0.06971428571428571, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.5564, + "step": 1159 + }, + { + "epoch": 0.06977443609022556, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.5673, + "step": 1160 + }, + { + "epoch": 0.06983458646616542, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4738, + "step": 1161 + }, + { + "epoch": 0.06989473684210526, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.5365, + "step": 1162 + }, + { + "epoch": 0.06995488721804512, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.5134, + "step": 1163 + }, + { + "epoch": 0.07001503759398496, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.5832, + "step": 1164 + }, + { + "epoch": 0.07007518796992482, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.5995, + "step": 1165 + }, + { + "epoch": 0.07013533834586466, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.4384, + "step": 1166 + }, + { + "epoch": 0.0701954887218045, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4868, + "step": 1167 + }, + { + "epoch": 0.07025563909774436, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4451, + "step": 1168 + }, + { + "epoch": 0.0703157894736842, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.5848, + "step": 1169 + }, + { + "epoch": 0.07037593984962406, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.457, + "step": 1170 + }, + { + "epoch": 0.07043609022556391, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.4152, + "step": 1171 + }, + { + "epoch": 0.07049624060150377, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4347, + "step": 1172 + }, + { + "epoch": 0.07055639097744361, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.4475, + "step": 1173 + }, + { + "epoch": 0.07061654135338345, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.5557, + "step": 1174 + }, + { + "epoch": 0.07067669172932331, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4904, + "step": 1175 + }, + { + "epoch": 0.07073684210526315, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.5091, + "step": 1176 + }, + { + "epoch": 0.07079699248120301, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.492, + "step": 1177 + }, + { + "epoch": 0.07085714285714285, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.5128, + "step": 1178 + }, + { + "epoch": 0.07091729323308271, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.5011, + "step": 1179 + }, + { + "epoch": 0.07097744360902256, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4253, + "step": 1180 + }, + { + "epoch": 0.0710375939849624, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.566, + "step": 1181 + }, + { + "epoch": 0.07109774436090226, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4836, + "step": 1182 + }, + { + "epoch": 0.0711578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5138, + "step": 1183 + }, + { + "epoch": 0.07121804511278196, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.4301, + "step": 1184 + }, + { + "epoch": 0.0712781954887218, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4675, + "step": 1185 + }, + { + "epoch": 0.07133834586466166, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.5089, + "step": 1186 + }, + { + "epoch": 0.0713984962406015, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.5564, + "step": 1187 + }, + { + "epoch": 0.07145864661654136, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.4983, + "step": 1188 + }, + { + "epoch": 0.0715187969924812, + "grad_norm": 0.4375, + "learning_rate": 3e-05, + "loss": 1.5204, + "step": 1189 + }, + { + "epoch": 0.07157894736842105, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.5422, + "step": 1190 + }, + { + "epoch": 0.0716390977443609, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4664, + "step": 1191 + }, + { + "epoch": 0.07169924812030075, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.5023, + "step": 1192 + }, + { + "epoch": 0.0717593984962406, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.508, + "step": 1193 + }, + { + "epoch": 0.07181954887218045, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.4671, + "step": 1194 + }, + { + "epoch": 0.07187969924812031, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.5435, + "step": 1195 + }, + { + "epoch": 0.07193984962406015, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.5483, + "step": 1196 + }, + { + "epoch": 0.072, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4468, + "step": 1197 + }, + { + "epoch": 0.07206015037593985, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5568, + "step": 1198 + }, + { + "epoch": 0.0721203007518797, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.5287, + "step": 1199 + }, + { + "epoch": 0.07218045112781955, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.5373, + "step": 1200 + }, + { + "epoch": 0.0722406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.456, + "step": 1201 + }, + { + "epoch": 0.07230075187969925, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.5966, + "step": 1202 + }, + { + "epoch": 0.0723609022556391, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5607, + "step": 1203 + }, + { + "epoch": 0.07242105263157894, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4664, + "step": 1204 + }, + { + "epoch": 0.0724812030075188, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4822, + "step": 1205 + }, + { + "epoch": 0.07254135338345864, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.5083, + "step": 1206 + }, + { + "epoch": 0.0726015037593985, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5205, + "step": 1207 + }, + { + "epoch": 0.07266165413533834, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.5625, + "step": 1208 + }, + { + "epoch": 0.0727218045112782, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.5686, + "step": 1209 + }, + { + "epoch": 0.07278195488721804, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.5375, + "step": 1210 + }, + { + "epoch": 0.07284210526315789, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.5736, + "step": 1211 + }, + { + "epoch": 0.07290225563909775, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.564, + "step": 1212 + }, + { + "epoch": 0.07296240601503759, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.5639, + "step": 1213 + }, + { + "epoch": 0.07302255639097745, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5053, + "step": 1214 + }, + { + "epoch": 0.07308270676691729, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.554, + "step": 1215 + }, + { + "epoch": 0.07314285714285715, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.4786, + "step": 1216 + }, + { + "epoch": 0.07320300751879699, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.4726, + "step": 1217 + }, + { + "epoch": 0.07326315789473684, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.5573, + "step": 1218 + }, + { + "epoch": 0.0733233082706767, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.5106, + "step": 1219 + }, + { + "epoch": 0.07338345864661654, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4343, + "step": 1220 + }, + { + "epoch": 0.0734436090225564, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4882, + "step": 1221 + }, + { + "epoch": 0.07350375939849624, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.5247, + "step": 1222 + }, + { + "epoch": 0.0735639097744361, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5035, + "step": 1223 + }, + { + "epoch": 0.07362406015037594, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5003, + "step": 1224 + }, + { + "epoch": 0.07368421052631578, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4837, + "step": 1225 + }, + { + "epoch": 0.07374436090225564, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4511, + "step": 1226 + }, + { + "epoch": 0.07380451127819548, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.4956, + "step": 1227 + }, + { + "epoch": 0.07386466165413534, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5744, + "step": 1228 + }, + { + "epoch": 0.07392481203007518, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.5658, + "step": 1229 + }, + { + "epoch": 0.07398496240601504, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.4871, + "step": 1230 + }, + { + "epoch": 0.07404511278195489, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.5131, + "step": 1231 + }, + { + "epoch": 0.07410526315789474, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.4873, + "step": 1232 + }, + { + "epoch": 0.07416541353383459, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.5107, + "step": 1233 + }, + { + "epoch": 0.07422556390977443, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.5062, + "step": 1234 + }, + { + "epoch": 0.07428571428571429, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5096, + "step": 1235 + }, + { + "epoch": 0.07434586466165413, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.5119, + "step": 1236 + }, + { + "epoch": 0.07440601503759399, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.51, + "step": 1237 + }, + { + "epoch": 0.07446616541353383, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.5053, + "step": 1238 + }, + { + "epoch": 0.07452631578947369, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3791, + "step": 1239 + }, + { + "epoch": 0.07458646616541353, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.4748, + "step": 1240 + }, + { + "epoch": 0.07464661654135338, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.5373, + "step": 1241 + }, + { + "epoch": 0.07470676691729324, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.5002, + "step": 1242 + }, + { + "epoch": 0.07476691729323308, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.4561, + "step": 1243 + }, + { + "epoch": 0.07482706766917294, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5366, + "step": 1244 + }, + { + "epoch": 0.07488721804511278, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.4937, + "step": 1245 + }, + { + "epoch": 0.07494736842105264, + "grad_norm": 0.408203125, + "learning_rate": 3e-05, + "loss": 1.5222, + "step": 1246 + }, + { + "epoch": 0.07500751879699248, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.5329, + "step": 1247 + }, + { + "epoch": 0.07506766917293232, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.4957, + "step": 1248 + }, + { + "epoch": 0.07512781954887218, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4381, + "step": 1249 + }, + { + "epoch": 0.07518796992481203, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.617, + "step": 1250 + }, + { + "epoch": 0.07524812030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.6078, + "step": 1251 + }, + { + "epoch": 0.07530827067669173, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.518, + "step": 1252 + }, + { + "epoch": 0.07536842105263158, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.5364, + "step": 1253 + }, + { + "epoch": 0.07542857142857143, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.5476, + "step": 1254 + }, + { + "epoch": 0.07548872180451127, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4301, + "step": 1255 + }, + { + "epoch": 0.07554887218045113, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5083, + "step": 1256 + }, + { + "epoch": 0.07560902255639097, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.5395, + "step": 1257 + }, + { + "epoch": 0.07566917293233083, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.5223, + "step": 1258 + }, + { + "epoch": 0.07572932330827067, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.6259, + "step": 1259 + }, + { + "epoch": 0.07578947368421053, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.4741, + "step": 1260 + }, + { + "epoch": 0.07584962406015038, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.647, + "step": 1261 + }, + { + "epoch": 0.07590977443609022, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.4431, + "step": 1262 + }, + { + "epoch": 0.07596992481203008, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4024, + "step": 1263 + }, + { + "epoch": 0.07603007518796992, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.5029, + "step": 1264 + }, + { + "epoch": 0.07609022556390978, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4499, + "step": 1265 + }, + { + "epoch": 0.07615037593984962, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.436, + "step": 1266 + }, + { + "epoch": 0.07621052631578948, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.4701, + "step": 1267 + }, + { + "epoch": 0.07627067669172932, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.488, + "step": 1268 + }, + { + "epoch": 0.07633082706766918, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.442, + "step": 1269 + }, + { + "epoch": 0.07639097744360902, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.5148, + "step": 1270 + }, + { + "epoch": 0.07645112781954887, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.5057, + "step": 1271 + }, + { + "epoch": 0.07651127819548872, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.5877, + "step": 1272 + }, + { + "epoch": 0.07657142857142857, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4242, + "step": 1273 + }, + { + "epoch": 0.07663157894736843, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4271, + "step": 1274 + }, + { + "epoch": 0.07669172932330827, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4535, + "step": 1275 + }, + { + "epoch": 0.07675187969924813, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5014, + "step": 1276 + }, + { + "epoch": 0.07681203007518797, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5151, + "step": 1277 + }, + { + "epoch": 0.07687218045112781, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.5178, + "step": 1278 + }, + { + "epoch": 0.07693233082706767, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.5551, + "step": 1279 + }, + { + "epoch": 0.07699248120300752, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4917, + "step": 1280 + }, + { + "epoch": 0.07705263157894737, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4337, + "step": 1281 + }, + { + "epoch": 0.07711278195488722, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.5299, + "step": 1282 + }, + { + "epoch": 0.07717293233082707, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.5116, + "step": 1283 + }, + { + "epoch": 0.07723308270676692, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.5269, + "step": 1284 + }, + { + "epoch": 0.07729323308270676, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4077, + "step": 1285 + }, + { + "epoch": 0.07735338345864662, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.6329, + "step": 1286 + }, + { + "epoch": 0.07741353383458646, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5611, + "step": 1287 + }, + { + "epoch": 0.07747368421052632, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.4598, + "step": 1288 + }, + { + "epoch": 0.07753383458646616, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.561, + "step": 1289 + }, + { + "epoch": 0.07759398496240602, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.5124, + "step": 1290 + }, + { + "epoch": 0.07765413533834586, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4624, + "step": 1291 + }, + { + "epoch": 0.07771428571428571, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.4794, + "step": 1292 + }, + { + "epoch": 0.07777443609022557, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4482, + "step": 1293 + }, + { + "epoch": 0.07783458646616541, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4592, + "step": 1294 + }, + { + "epoch": 0.07789473684210527, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4543, + "step": 1295 + }, + { + "epoch": 0.07795488721804511, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4776, + "step": 1296 + }, + { + "epoch": 0.07801503759398497, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.4465, + "step": 1297 + }, + { + "epoch": 0.07807518796992481, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.445, + "step": 1298 + }, + { + "epoch": 0.07813533834586465, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.5244, + "step": 1299 + }, + { + "epoch": 0.07819548872180451, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.5421, + "step": 1300 + }, + { + "epoch": 0.07825563909774436, + "grad_norm": 0.38671875, + "learning_rate": 3e-05, + "loss": 1.513, + "step": 1301 + }, + { + "epoch": 0.07831578947368421, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.44, + "step": 1302 + }, + { + "epoch": 0.07837593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4905, + "step": 1303 + }, + { + "epoch": 0.07843609022556391, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4574, + "step": 1304 + }, + { + "epoch": 0.07849624060150376, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4302, + "step": 1305 + }, + { + "epoch": 0.07855639097744362, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.5685, + "step": 1306 + }, + { + "epoch": 0.07861654135338346, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.4916, + "step": 1307 + }, + { + "epoch": 0.0786766917293233, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.4522, + "step": 1308 + }, + { + "epoch": 0.07873684210526316, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.5581, + "step": 1309 + }, + { + "epoch": 0.078796992481203, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4541, + "step": 1310 + }, + { + "epoch": 0.07885714285714286, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.4125, + "step": 1311 + }, + { + "epoch": 0.0789172932330827, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.5736, + "step": 1312 + }, + { + "epoch": 0.07897744360902256, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4963, + "step": 1313 + }, + { + "epoch": 0.0790375939849624, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4654, + "step": 1314 + }, + { + "epoch": 0.07909774436090225, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.5378, + "step": 1315 + }, + { + "epoch": 0.07915789473684211, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5105, + "step": 1316 + }, + { + "epoch": 0.07921804511278195, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.4902, + "step": 1317 + }, + { + "epoch": 0.07927819548872181, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.542, + "step": 1318 + }, + { + "epoch": 0.07933834586466165, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4887, + "step": 1319 + }, + { + "epoch": 0.07939849624060151, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.5123, + "step": 1320 + }, + { + "epoch": 0.07945864661654135, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.488, + "step": 1321 + }, + { + "epoch": 0.0795187969924812, + "grad_norm": 0.380859375, + "learning_rate": 3e-05, + "loss": 1.4973, + "step": 1322 + }, + { + "epoch": 0.07957894736842105, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.4419, + "step": 1323 + }, + { + "epoch": 0.0796390977443609, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4787, + "step": 1324 + }, + { + "epoch": 0.07969924812030076, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.4732, + "step": 1325 + }, + { + "epoch": 0.0797593984962406, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.503, + "step": 1326 + }, + { + "epoch": 0.07981954887218046, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4672, + "step": 1327 + }, + { + "epoch": 0.0798796992481203, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4833, + "step": 1328 + }, + { + "epoch": 0.07993984962406014, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.4085, + "step": 1329 + }, + { + "epoch": 0.08, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.5606, + "step": 1330 + }, + { + "epoch": 0.08006015037593985, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4983, + "step": 1331 + }, + { + "epoch": 0.0801203007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4931, + "step": 1332 + }, + { + "epoch": 0.08018045112781955, + "grad_norm": 0.390625, + "learning_rate": 3e-05, + "loss": 1.5081, + "step": 1333 + }, + { + "epoch": 0.0802406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4975, + "step": 1334 + }, + { + "epoch": 0.08030075187969925, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.4857, + "step": 1335 + }, + { + "epoch": 0.08036090225563909, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.6141, + "step": 1336 + }, + { + "epoch": 0.08042105263157895, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4762, + "step": 1337 + }, + { + "epoch": 0.08048120300751879, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.491, + "step": 1338 + }, + { + "epoch": 0.08054135338345865, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5563, + "step": 1339 + }, + { + "epoch": 0.0806015037593985, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.4664, + "step": 1340 + }, + { + "epoch": 0.08066165413533835, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.5618, + "step": 1341 + }, + { + "epoch": 0.0807218045112782, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4416, + "step": 1342 + }, + { + "epoch": 0.08078195488721804, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5089, + "step": 1343 + }, + { + "epoch": 0.0808421052631579, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5565, + "step": 1344 + }, + { + "epoch": 0.08090225563909774, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4982, + "step": 1345 + }, + { + "epoch": 0.0809624060150376, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5302, + "step": 1346 + }, + { + "epoch": 0.08102255639097744, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.4045, + "step": 1347 + }, + { + "epoch": 0.0810827067669173, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.5193, + "step": 1348 + }, + { + "epoch": 0.08114285714285714, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.4571, + "step": 1349 + }, + { + "epoch": 0.081203007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4345, + "step": 1350 + }, + { + "epoch": 0.08126315789473684, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.5443, + "step": 1351 + }, + { + "epoch": 0.08132330827067669, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4721, + "step": 1352 + }, + { + "epoch": 0.08138345864661654, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.5409, + "step": 1353 + }, + { + "epoch": 0.08144360902255639, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.5008, + "step": 1354 + }, + { + "epoch": 0.08150375939849624, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4909, + "step": 1355 + }, + { + "epoch": 0.08156390977443609, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.4827, + "step": 1356 + }, + { + "epoch": 0.08162406015037595, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.4982, + "step": 1357 + }, + { + "epoch": 0.08168421052631579, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.4576, + "step": 1358 + }, + { + "epoch": 0.08174436090225563, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4833, + "step": 1359 + }, + { + "epoch": 0.08180451127819549, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.5173, + "step": 1360 + }, + { + "epoch": 0.08186466165413533, + "grad_norm": 0.462890625, + "learning_rate": 3e-05, + "loss": 1.4845, + "step": 1361 + }, + { + "epoch": 0.08192481203007519, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.5045, + "step": 1362 + }, + { + "epoch": 0.08198496240601504, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.472, + "step": 1363 + }, + { + "epoch": 0.0820451127819549, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4854, + "step": 1364 + }, + { + "epoch": 0.08210526315789474, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.5342, + "step": 1365 + }, + { + "epoch": 0.08216541353383458, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.5186, + "step": 1366 + }, + { + "epoch": 0.08222556390977444, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.5401, + "step": 1367 + }, + { + "epoch": 0.08228571428571428, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.7527, + "step": 1368 + }, + { + "epoch": 0.08234586466165414, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.4655, + "step": 1369 + }, + { + "epoch": 0.08240601503759398, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.4556, + "step": 1370 + }, + { + "epoch": 0.08246616541353384, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4433, + "step": 1371 + }, + { + "epoch": 0.08252631578947368, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.4557, + "step": 1372 + }, + { + "epoch": 0.08258646616541353, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.5146, + "step": 1373 + }, + { + "epoch": 0.08264661654135338, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4968, + "step": 1374 + }, + { + "epoch": 0.08270676691729323, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.4813, + "step": 1375 + }, + { + "epoch": 0.08276691729323309, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.5551, + "step": 1376 + }, + { + "epoch": 0.08282706766917293, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.4718, + "step": 1377 + }, + { + "epoch": 0.08288721804511279, + "grad_norm": 0.40625, + "learning_rate": 3e-05, + "loss": 1.5044, + "step": 1378 + }, + { + "epoch": 0.08294736842105263, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.5596, + "step": 1379 + }, + { + "epoch": 0.08300751879699247, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.5253, + "step": 1380 + }, + { + "epoch": 0.08306766917293233, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.427, + "step": 1381 + }, + { + "epoch": 0.08312781954887218, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4921, + "step": 1382 + }, + { + "epoch": 0.08318796992481203, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.5417, + "step": 1383 + }, + { + "epoch": 0.08324812030075188, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.5213, + "step": 1384 + }, + { + "epoch": 0.08330827067669173, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4969, + "step": 1385 + }, + { + "epoch": 0.08336842105263158, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.5171, + "step": 1386 + }, + { + "epoch": 0.08342857142857144, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.5683, + "step": 1387 + }, + { + "epoch": 0.08348872180451128, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.5061, + "step": 1388 + }, + { + "epoch": 0.08354887218045112, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.4607, + "step": 1389 + }, + { + "epoch": 0.08360902255639098, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.5593, + "step": 1390 + }, + { + "epoch": 0.08366917293233082, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.485, + "step": 1391 + }, + { + "epoch": 0.08372932330827068, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.5293, + "step": 1392 + }, + { + "epoch": 0.08378947368421052, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.5272, + "step": 1393 + }, + { + "epoch": 0.08384962406015038, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.4929, + "step": 1394 + }, + { + "epoch": 0.08390977443609023, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.4794, + "step": 1395 + }, + { + "epoch": 0.08396992481203007, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.5016, + "step": 1396 + }, + { + "epoch": 0.08403007518796993, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5339, + "step": 1397 + }, + { + "epoch": 0.08409022556390977, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.5087, + "step": 1398 + }, + { + "epoch": 0.08415037593984963, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4347, + "step": 1399 + }, + { + "epoch": 0.08421052631578947, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4936, + "step": 1400 + }, + { + "epoch": 0.08427067669172933, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.4636, + "step": 1401 + }, + { + "epoch": 0.08433082706766917, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.5336, + "step": 1402 + }, + { + "epoch": 0.08439097744360902, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.6408, + "step": 1403 + }, + { + "epoch": 0.08445112781954887, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.442, + "step": 1404 + }, + { + "epoch": 0.08451127819548872, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4779, + "step": 1405 + }, + { + "epoch": 0.08457142857142858, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4542, + "step": 1406 + }, + { + "epoch": 0.08463157894736842, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.4423, + "step": 1407 + }, + { + "epoch": 0.08469172932330828, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4708, + "step": 1408 + }, + { + "epoch": 0.08475187969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.5849, + "step": 1409 + }, + { + "epoch": 0.08481203007518796, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.5928, + "step": 1410 + }, + { + "epoch": 0.08487218045112782, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.5917, + "step": 1411 + }, + { + "epoch": 0.08493233082706766, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.4514, + "step": 1412 + }, + { + "epoch": 0.08499248120300752, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4594, + "step": 1413 + }, + { + "epoch": 0.08505263157894737, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.4808, + "step": 1414 + }, + { + "epoch": 0.08511278195488722, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4936, + "step": 1415 + }, + { + "epoch": 0.08517293233082707, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.5362, + "step": 1416 + }, + { + "epoch": 0.08523308270676691, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.5485, + "step": 1417 + }, + { + "epoch": 0.08529323308270677, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4801, + "step": 1418 + }, + { + "epoch": 0.08535338345864661, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.4739, + "step": 1419 + }, + { + "epoch": 0.08541353383458647, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.5106, + "step": 1420 + }, + { + "epoch": 0.08547368421052631, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4381, + "step": 1421 + }, + { + "epoch": 0.08553383458646617, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4513, + "step": 1422 + }, + { + "epoch": 0.08559398496240601, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4154, + "step": 1423 + }, + { + "epoch": 0.08565413533834587, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4983, + "step": 1424 + }, + { + "epoch": 0.08571428571428572, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5089, + "step": 1425 + }, + { + "epoch": 0.08577443609022556, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.5577, + "step": 1426 + }, + { + "epoch": 0.08583458646616542, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.5162, + "step": 1427 + }, + { + "epoch": 0.08589473684210526, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.4365, + "step": 1428 + }, + { + "epoch": 0.08595488721804512, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4833, + "step": 1429 + }, + { + "epoch": 0.08601503759398496, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.5454, + "step": 1430 + }, + { + "epoch": 0.08607518796992482, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4578, + "step": 1431 + }, + { + "epoch": 0.08613533834586466, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5135, + "step": 1432 + }, + { + "epoch": 0.0861954887218045, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.5651, + "step": 1433 + }, + { + "epoch": 0.08625563909774436, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4783, + "step": 1434 + }, + { + "epoch": 0.0863157894736842, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.4249, + "step": 1435 + }, + { + "epoch": 0.08637593984962406, + "grad_norm": 0.462890625, + "learning_rate": 3e-05, + "loss": 1.5111, + "step": 1436 + }, + { + "epoch": 0.08643609022556391, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.4941, + "step": 1437 + }, + { + "epoch": 0.08649624060150377, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.4959, + "step": 1438 + }, + { + "epoch": 0.08655639097744361, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4407, + "step": 1439 + }, + { + "epoch": 0.08661654135338345, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5289, + "step": 1440 + }, + { + "epoch": 0.08667669172932331, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.5833, + "step": 1441 + }, + { + "epoch": 0.08673684210526315, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.5436, + "step": 1442 + }, + { + "epoch": 0.08679699248120301, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.4399, + "step": 1443 + }, + { + "epoch": 0.08685714285714285, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.4761, + "step": 1444 + }, + { + "epoch": 0.08691729323308271, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5045, + "step": 1445 + }, + { + "epoch": 0.08697744360902256, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.5007, + "step": 1446 + }, + { + "epoch": 0.0870375939849624, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4343, + "step": 1447 + }, + { + "epoch": 0.08709774436090226, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4978, + "step": 1448 + }, + { + "epoch": 0.0871578947368421, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.453, + "step": 1449 + }, + { + "epoch": 0.08721804511278196, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4446, + "step": 1450 + }, + { + "epoch": 0.0872781954887218, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5842, + "step": 1451 + }, + { + "epoch": 0.08733834586466166, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5641, + "step": 1452 + }, + { + "epoch": 0.0873984962406015, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4825, + "step": 1453 + }, + { + "epoch": 0.08745864661654135, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4411, + "step": 1454 + }, + { + "epoch": 0.0875187969924812, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4921, + "step": 1455 + }, + { + "epoch": 0.08757894736842105, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.5643, + "step": 1456 + }, + { + "epoch": 0.0876390977443609, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.3939, + "step": 1457 + }, + { + "epoch": 0.08769924812030075, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.47, + "step": 1458 + }, + { + "epoch": 0.0877593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.5307, + "step": 1459 + }, + { + "epoch": 0.08781954887218045, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.4743, + "step": 1460 + }, + { + "epoch": 0.08787969924812031, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4938, + "step": 1461 + }, + { + "epoch": 0.08793984962406015, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.5196, + "step": 1462 + }, + { + "epoch": 0.088, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.5018, + "step": 1463 + }, + { + "epoch": 0.08806015037593985, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.53, + "step": 1464 + }, + { + "epoch": 0.0881203007518797, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.4403, + "step": 1465 + }, + { + "epoch": 0.08818045112781955, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5726, + "step": 1466 + }, + { + "epoch": 0.0882406015037594, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4661, + "step": 1467 + }, + { + "epoch": 0.08830075187969925, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5036, + "step": 1468 + }, + { + "epoch": 0.0883609022556391, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5018, + "step": 1469 + }, + { + "epoch": 0.08842105263157894, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.486, + "step": 1470 + }, + { + "epoch": 0.0884812030075188, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4495, + "step": 1471 + }, + { + "epoch": 0.08854135338345864, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.4305, + "step": 1472 + }, + { + "epoch": 0.0886015037593985, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4524, + "step": 1473 + }, + { + "epoch": 0.08866165413533834, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.5037, + "step": 1474 + }, + { + "epoch": 0.0887218045112782, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4995, + "step": 1475 + }, + { + "epoch": 0.08878195488721805, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4728, + "step": 1476 + }, + { + "epoch": 0.08884210526315789, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.499, + "step": 1477 + }, + { + "epoch": 0.08890225563909775, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5138, + "step": 1478 + }, + { + "epoch": 0.08896240601503759, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4548, + "step": 1479 + }, + { + "epoch": 0.08902255639097745, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.4827, + "step": 1480 + }, + { + "epoch": 0.08908270676691729, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4056, + "step": 1481 + }, + { + "epoch": 0.08914285714285715, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.5121, + "step": 1482 + }, + { + "epoch": 0.08920300751879699, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4347, + "step": 1483 + }, + { + "epoch": 0.08926315789473684, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4649, + "step": 1484 + }, + { + "epoch": 0.0893233082706767, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4637, + "step": 1485 + }, + { + "epoch": 0.08938345864661654, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.4743, + "step": 1486 + }, + { + "epoch": 0.0894436090225564, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.4768, + "step": 1487 + }, + { + "epoch": 0.08950375939849624, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.399, + "step": 1488 + }, + { + "epoch": 0.0895639097744361, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4636, + "step": 1489 + }, + { + "epoch": 0.08962406015037594, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4938, + "step": 1490 + }, + { + "epoch": 0.08968421052631578, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.5205, + "step": 1491 + }, + { + "epoch": 0.08974436090225564, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.5066, + "step": 1492 + }, + { + "epoch": 0.08980451127819548, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.449, + "step": 1493 + }, + { + "epoch": 0.08986466165413534, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4888, + "step": 1494 + }, + { + "epoch": 0.08992481203007519, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4613, + "step": 1495 + }, + { + "epoch": 0.08998496240601504, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.4429, + "step": 1496 + }, + { + "epoch": 0.09004511278195489, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.4452, + "step": 1497 + }, + { + "epoch": 0.09010526315789473, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.4378, + "step": 1498 + }, + { + "epoch": 0.09016541353383459, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4448, + "step": 1499 + }, + { + "epoch": 0.09022556390977443, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.5188, + "step": 1500 + }, + { + "epoch": 0.09028571428571429, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.4279, + "step": 1501 + }, + { + "epoch": 0.09034586466165413, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.565, + "step": 1502 + }, + { + "epoch": 0.09040601503759399, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4803, + "step": 1503 + }, + { + "epoch": 0.09046616541353383, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4766, + "step": 1504 + }, + { + "epoch": 0.09052631578947369, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.5102, + "step": 1505 + }, + { + "epoch": 0.09058646616541353, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.5055, + "step": 1506 + }, + { + "epoch": 0.09064661654135338, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.5354, + "step": 1507 + }, + { + "epoch": 0.09070676691729324, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3914, + "step": 1508 + }, + { + "epoch": 0.09076691729323308, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5089, + "step": 1509 + }, + { + "epoch": 0.09082706766917294, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4403, + "step": 1510 + }, + { + "epoch": 0.09088721804511278, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.401, + "step": 1511 + }, + { + "epoch": 0.09094736842105264, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.6383, + "step": 1512 + }, + { + "epoch": 0.09100751879699248, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.5288, + "step": 1513 + }, + { + "epoch": 0.09106766917293233, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5018, + "step": 1514 + }, + { + "epoch": 0.09112781954887218, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.4609, + "step": 1515 + }, + { + "epoch": 0.09118796992481203, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4668, + "step": 1516 + }, + { + "epoch": 0.09124812030075188, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5011, + "step": 1517 + }, + { + "epoch": 0.09130827067669173, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.5174, + "step": 1518 + }, + { + "epoch": 0.09136842105263158, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4115, + "step": 1519 + }, + { + "epoch": 0.09142857142857143, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4914, + "step": 1520 + }, + { + "epoch": 0.09148872180451127, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4867, + "step": 1521 + }, + { + "epoch": 0.09154887218045113, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4182, + "step": 1522 + }, + { + "epoch": 0.09160902255639097, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.5123, + "step": 1523 + }, + { + "epoch": 0.09166917293233083, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.481, + "step": 1524 + }, + { + "epoch": 0.09172932330827067, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4404, + "step": 1525 + }, + { + "epoch": 0.09178947368421053, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5188, + "step": 1526 + }, + { + "epoch": 0.09184962406015038, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.5154, + "step": 1527 + }, + { + "epoch": 0.09190977443609022, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4809, + "step": 1528 + }, + { + "epoch": 0.09196992481203008, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3615, + "step": 1529 + }, + { + "epoch": 0.09203007518796992, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3749, + "step": 1530 + }, + { + "epoch": 0.09209022556390978, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4882, + "step": 1531 + }, + { + "epoch": 0.09215037593984962, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.5013, + "step": 1532 + }, + { + "epoch": 0.09221052631578948, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.487, + "step": 1533 + }, + { + "epoch": 0.09227067669172932, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5089, + "step": 1534 + }, + { + "epoch": 0.09233082706766917, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.4662, + "step": 1535 + }, + { + "epoch": 0.09239097744360902, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5058, + "step": 1536 + }, + { + "epoch": 0.09245112781954887, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.5274, + "step": 1537 + }, + { + "epoch": 0.09251127819548872, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.4804, + "step": 1538 + }, + { + "epoch": 0.09257142857142857, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.4598, + "step": 1539 + }, + { + "epoch": 0.09263157894736843, + "grad_norm": 0.42578125, + "learning_rate": 3e-05, + "loss": 1.4942, + "step": 1540 + }, + { + "epoch": 0.09269172932330827, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4405, + "step": 1541 + }, + { + "epoch": 0.09275187969924813, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.4724, + "step": 1542 + }, + { + "epoch": 0.09281203007518797, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4433, + "step": 1543 + }, + { + "epoch": 0.09287218045112781, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4174, + "step": 1544 + }, + { + "epoch": 0.09293233082706767, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.46, + "step": 1545 + }, + { + "epoch": 0.09299248120300752, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.5034, + "step": 1546 + }, + { + "epoch": 0.09305263157894737, + "grad_norm": 0.42578125, + "learning_rate": 3e-05, + "loss": 1.4328, + "step": 1547 + }, + { + "epoch": 0.09311278195488722, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.5077, + "step": 1548 + }, + { + "epoch": 0.09317293233082707, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.5005, + "step": 1549 + }, + { + "epoch": 0.09323308270676692, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.4515, + "step": 1550 + }, + { + "epoch": 0.09329323308270676, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4654, + "step": 1551 + }, + { + "epoch": 0.09335338345864662, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4295, + "step": 1552 + }, + { + "epoch": 0.09341353383458646, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5325, + "step": 1553 + }, + { + "epoch": 0.09347368421052632, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.4625, + "step": 1554 + }, + { + "epoch": 0.09353383458646616, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4743, + "step": 1555 + }, + { + "epoch": 0.09359398496240602, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4718, + "step": 1556 + }, + { + "epoch": 0.09365413533834586, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.625, + "step": 1557 + }, + { + "epoch": 0.09371428571428571, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.4273, + "step": 1558 + }, + { + "epoch": 0.09377443609022557, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4172, + "step": 1559 + }, + { + "epoch": 0.09383458646616541, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4752, + "step": 1560 + }, + { + "epoch": 0.09389473684210527, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4686, + "step": 1561 + }, + { + "epoch": 0.09395488721804511, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.4665, + "step": 1562 + }, + { + "epoch": 0.09401503759398497, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3837, + "step": 1563 + }, + { + "epoch": 0.09407518796992481, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.5503, + "step": 1564 + }, + { + "epoch": 0.09413533834586466, + "grad_norm": 0.462890625, + "learning_rate": 3e-05, + "loss": 1.4857, + "step": 1565 + }, + { + "epoch": 0.09419548872180451, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.467, + "step": 1566 + }, + { + "epoch": 0.09425563909774436, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4678, + "step": 1567 + }, + { + "epoch": 0.09431578947368421, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4774, + "step": 1568 + }, + { + "epoch": 0.09437593984962406, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4713, + "step": 1569 + }, + { + "epoch": 0.09443609022556392, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3835, + "step": 1570 + }, + { + "epoch": 0.09449624060150376, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.3738, + "step": 1571 + }, + { + "epoch": 0.0945563909774436, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.5321, + "step": 1572 + }, + { + "epoch": 0.09461654135338346, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4497, + "step": 1573 + }, + { + "epoch": 0.0946766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.5092, + "step": 1574 + }, + { + "epoch": 0.09473684210526316, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4421, + "step": 1575 + }, + { + "epoch": 0.094796992481203, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.5055, + "step": 1576 + }, + { + "epoch": 0.09485714285714286, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.4896, + "step": 1577 + }, + { + "epoch": 0.0949172932330827, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4702, + "step": 1578 + }, + { + "epoch": 0.09497744360902256, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.5427, + "step": 1579 + }, + { + "epoch": 0.0950375939849624, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5296, + "step": 1580 + }, + { + "epoch": 0.09509774436090225, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.5238, + "step": 1581 + }, + { + "epoch": 0.09515789473684211, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.4204, + "step": 1582 + }, + { + "epoch": 0.09521804511278195, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.5507, + "step": 1583 + }, + { + "epoch": 0.09527819548872181, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3981, + "step": 1584 + }, + { + "epoch": 0.09533834586466165, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4463, + "step": 1585 + }, + { + "epoch": 0.09539849624060151, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5377, + "step": 1586 + }, + { + "epoch": 0.09545864661654135, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.5506, + "step": 1587 + }, + { + "epoch": 0.0955187969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.5451, + "step": 1588 + }, + { + "epoch": 0.09557894736842105, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4474, + "step": 1589 + }, + { + "epoch": 0.0956390977443609, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4518, + "step": 1590 + }, + { + "epoch": 0.09569924812030076, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.4995, + "step": 1591 + }, + { + "epoch": 0.0957593984962406, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.5241, + "step": 1592 + }, + { + "epoch": 0.09581954887218046, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.512, + "step": 1593 + }, + { + "epoch": 0.0958796992481203, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.441, + "step": 1594 + }, + { + "epoch": 0.09593984962406014, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.4982, + "step": 1595 + }, + { + "epoch": 0.096, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4864, + "step": 1596 + }, + { + "epoch": 0.09606015037593985, + "grad_norm": 0.423828125, + "learning_rate": 3e-05, + "loss": 1.4628, + "step": 1597 + }, + { + "epoch": 0.0961203007518797, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4686, + "step": 1598 + }, + { + "epoch": 0.09618045112781955, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.4733, + "step": 1599 + }, + { + "epoch": 0.0962406015037594, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.5022, + "step": 1600 + }, + { + "epoch": 0.09630075187969925, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4299, + "step": 1601 + }, + { + "epoch": 0.09636090225563909, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.4551, + "step": 1602 + }, + { + "epoch": 0.09642105263157895, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.5001, + "step": 1603 + }, + { + "epoch": 0.09648120300751879, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.4885, + "step": 1604 + }, + { + "epoch": 0.09654135338345865, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.4658, + "step": 1605 + }, + { + "epoch": 0.0966015037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4037, + "step": 1606 + }, + { + "epoch": 0.09666165413533835, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4747, + "step": 1607 + }, + { + "epoch": 0.0967218045112782, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4167, + "step": 1608 + }, + { + "epoch": 0.09678195488721804, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.441, + "step": 1609 + }, + { + "epoch": 0.0968421052631579, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3697, + "step": 1610 + }, + { + "epoch": 0.09690225563909774, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.4751, + "step": 1611 + }, + { + "epoch": 0.0969624060150376, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4206, + "step": 1612 + }, + { + "epoch": 0.09702255639097744, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4532, + "step": 1613 + }, + { + "epoch": 0.0970827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.4859, + "step": 1614 + }, + { + "epoch": 0.09714285714285714, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.5496, + "step": 1615 + }, + { + "epoch": 0.097203007518797, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.5534, + "step": 1616 + }, + { + "epoch": 0.09726315789473684, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.6745, + "step": 1617 + }, + { + "epoch": 0.09732330827067669, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.4551, + "step": 1618 + }, + { + "epoch": 0.09738345864661654, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.4832, + "step": 1619 + }, + { + "epoch": 0.09744360902255639, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4567, + "step": 1620 + }, + { + "epoch": 0.09750375939849625, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3937, + "step": 1621 + }, + { + "epoch": 0.09756390977443609, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.4941, + "step": 1622 + }, + { + "epoch": 0.09762406015037595, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.4526, + "step": 1623 + }, + { + "epoch": 0.09768421052631579, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.4767, + "step": 1624 + }, + { + "epoch": 0.09774436090225563, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4072, + "step": 1625 + }, + { + "epoch": 0.09780451127819549, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5296, + "step": 1626 + }, + { + "epoch": 0.09786466165413533, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4468, + "step": 1627 + }, + { + "epoch": 0.09792481203007519, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.4093, + "step": 1628 + }, + { + "epoch": 0.09798496240601504, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.519, + "step": 1629 + }, + { + "epoch": 0.0980451127819549, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.5687, + "step": 1630 + }, + { + "epoch": 0.09810526315789474, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4188, + "step": 1631 + }, + { + "epoch": 0.09816541353383458, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4701, + "step": 1632 + }, + { + "epoch": 0.09822556390977444, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.5335, + "step": 1633 + }, + { + "epoch": 0.09828571428571428, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4246, + "step": 1634 + }, + { + "epoch": 0.09834586466165414, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4598, + "step": 1635 + }, + { + "epoch": 0.09840601503759398, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.5464, + "step": 1636 + }, + { + "epoch": 0.09846616541353384, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.6081, + "step": 1637 + }, + { + "epoch": 0.09852631578947368, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.4482, + "step": 1638 + }, + { + "epoch": 0.09858646616541353, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.4637, + "step": 1639 + }, + { + "epoch": 0.09864661654135339, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.4726, + "step": 1640 + }, + { + "epoch": 0.09870676691729323, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.4926, + "step": 1641 + }, + { + "epoch": 0.09876691729323309, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.5223, + "step": 1642 + }, + { + "epoch": 0.09882706766917293, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.4838, + "step": 1643 + }, + { + "epoch": 0.09888721804511279, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4991, + "step": 1644 + }, + { + "epoch": 0.09894736842105263, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4707, + "step": 1645 + }, + { + "epoch": 0.09900751879699247, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.5961, + "step": 1646 + }, + { + "epoch": 0.09906766917293233, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.7145, + "step": 1647 + }, + { + "epoch": 0.09912781954887218, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.4377, + "step": 1648 + }, + { + "epoch": 0.09918796992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.5083, + "step": 1649 + }, + { + "epoch": 0.09924812030075188, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3907, + "step": 1650 + }, + { + "epoch": 0.09930827067669173, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4794, + "step": 1651 + }, + { + "epoch": 0.09936842105263158, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.4816, + "step": 1652 + }, + { + "epoch": 0.09942857142857142, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.427, + "step": 1653 + }, + { + "epoch": 0.09948872180451128, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4302, + "step": 1654 + }, + { + "epoch": 0.09954887218045112, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.437, + "step": 1655 + }, + { + "epoch": 0.09960902255639098, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.438, + "step": 1656 + }, + { + "epoch": 0.09966917293233082, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.5222, + "step": 1657 + }, + { + "epoch": 0.09972932330827068, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4332, + "step": 1658 + }, + { + "epoch": 0.09978947368421053, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4984, + "step": 1659 + }, + { + "epoch": 0.09984962406015038, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3867, + "step": 1660 + }, + { + "epoch": 0.09990977443609023, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4626, + "step": 1661 + }, + { + "epoch": 0.09996992481203007, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.4591, + "step": 1662 + }, + { + "epoch": 0.10003007518796993, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.5077, + "step": 1663 + }, + { + "epoch": 0.10009022556390977, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4789, + "step": 1664 + }, + { + "epoch": 0.10015037593984963, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.4353, + "step": 1665 + }, + { + "epoch": 0.10021052631578947, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.4765, + "step": 1666 + }, + { + "epoch": 0.10027067669172933, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.4421, + "step": 1667 + }, + { + "epoch": 0.10033082706766917, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.5468, + "step": 1668 + }, + { + "epoch": 0.10039097744360902, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4274, + "step": 1669 + }, + { + "epoch": 0.10045112781954887, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4609, + "step": 1670 + }, + { + "epoch": 0.10051127819548872, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.5024, + "step": 1671 + }, + { + "epoch": 0.10057142857142858, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4436, + "step": 1672 + }, + { + "epoch": 0.10063157894736842, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3913, + "step": 1673 + }, + { + "epoch": 0.10069172932330828, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4552, + "step": 1674 + }, + { + "epoch": 0.10075187969924812, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.4358, + "step": 1675 + }, + { + "epoch": 0.10081203007518796, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3776, + "step": 1676 + }, + { + "epoch": 0.10087218045112782, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4799, + "step": 1677 + }, + { + "epoch": 0.10093233082706767, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5315, + "step": 1678 + }, + { + "epoch": 0.10099248120300752, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.4315, + "step": 1679 + }, + { + "epoch": 0.10105263157894737, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4686, + "step": 1680 + }, + { + "epoch": 0.10111278195488722, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4488, + "step": 1681 + }, + { + "epoch": 0.10117293233082707, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.4864, + "step": 1682 + }, + { + "epoch": 0.10123308270676691, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.5155, + "step": 1683 + }, + { + "epoch": 0.10129323308270677, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4516, + "step": 1684 + }, + { + "epoch": 0.10135338345864661, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.552, + "step": 1685 + }, + { + "epoch": 0.10141353383458647, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4435, + "step": 1686 + }, + { + "epoch": 0.10147368421052631, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4142, + "step": 1687 + }, + { + "epoch": 0.10153383458646617, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.4306, + "step": 1688 + }, + { + "epoch": 0.10159398496240601, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.451, + "step": 1689 + }, + { + "epoch": 0.10165413533834586, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4224, + "step": 1690 + }, + { + "epoch": 0.10171428571428572, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.394, + "step": 1691 + }, + { + "epoch": 0.10177443609022556, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4431, + "step": 1692 + }, + { + "epoch": 0.10183458646616542, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.4921, + "step": 1693 + }, + { + "epoch": 0.10189473684210526, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.501, + "step": 1694 + }, + { + "epoch": 0.10195488721804512, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.4086, + "step": 1695 + }, + { + "epoch": 0.10201503759398496, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.439, + "step": 1696 + }, + { + "epoch": 0.10207518796992482, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.5312, + "step": 1697 + }, + { + "epoch": 0.10213533834586466, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.4588, + "step": 1698 + }, + { + "epoch": 0.1021954887218045, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4778, + "step": 1699 + }, + { + "epoch": 0.10225563909774436, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.4027, + "step": 1700 + }, + { + "epoch": 0.10231578947368421, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4344, + "step": 1701 + }, + { + "epoch": 0.10237593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4667, + "step": 1702 + }, + { + "epoch": 0.10243609022556391, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4713, + "step": 1703 + }, + { + "epoch": 0.10249624060150377, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4518, + "step": 1704 + }, + { + "epoch": 0.10255639097744361, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.6017, + "step": 1705 + }, + { + "epoch": 0.10261654135338345, + "grad_norm": 0.392578125, + "learning_rate": 3e-05, + "loss": 1.4815, + "step": 1706 + }, + { + "epoch": 0.10267669172932331, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.4469, + "step": 1707 + }, + { + "epoch": 0.10273684210526315, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4113, + "step": 1708 + }, + { + "epoch": 0.10279699248120301, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.5186, + "step": 1709 + }, + { + "epoch": 0.10285714285714286, + "grad_norm": 0.412109375, + "learning_rate": 3e-05, + "loss": 1.4787, + "step": 1710 + }, + { + "epoch": 0.10291729323308271, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.5257, + "step": 1711 + }, + { + "epoch": 0.10297744360902256, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.5148, + "step": 1712 + }, + { + "epoch": 0.1030375939849624, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.4323, + "step": 1713 + }, + { + "epoch": 0.10309774436090226, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.6283, + "step": 1714 + }, + { + "epoch": 0.1031578947368421, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.498, + "step": 1715 + }, + { + "epoch": 0.10321804511278196, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4217, + "step": 1716 + }, + { + "epoch": 0.1032781954887218, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.7704, + "step": 1717 + }, + { + "epoch": 0.10333834586466166, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.46, + "step": 1718 + }, + { + "epoch": 0.1033984962406015, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.503, + "step": 1719 + }, + { + "epoch": 0.10345864661654135, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.5248, + "step": 1720 + }, + { + "epoch": 0.1035187969924812, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4379, + "step": 1721 + }, + { + "epoch": 0.10357894736842105, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4486, + "step": 1722 + }, + { + "epoch": 0.1036390977443609, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3777, + "step": 1723 + }, + { + "epoch": 0.10369924812030075, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.4189, + "step": 1724 + }, + { + "epoch": 0.1037593984962406, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.462, + "step": 1725 + }, + { + "epoch": 0.10381954887218045, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.5222, + "step": 1726 + }, + { + "epoch": 0.1038796992481203, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.4445, + "step": 1727 + }, + { + "epoch": 0.10393984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4028, + "step": 1728 + }, + { + "epoch": 0.104, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4767, + "step": 1729 + }, + { + "epoch": 0.10406015037593985, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.444, + "step": 1730 + }, + { + "epoch": 0.1041203007518797, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.4819, + "step": 1731 + }, + { + "epoch": 0.10418045112781955, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.42, + "step": 1732 + }, + { + "epoch": 0.1042406015037594, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.4882, + "step": 1733 + }, + { + "epoch": 0.10430075187969925, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.408, + "step": 1734 + }, + { + "epoch": 0.1043609022556391, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4339, + "step": 1735 + }, + { + "epoch": 0.10442105263157894, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.5082, + "step": 1736 + }, + { + "epoch": 0.1044812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.4552, + "step": 1737 + }, + { + "epoch": 0.10454135338345864, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4828, + "step": 1738 + }, + { + "epoch": 0.1046015037593985, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.4513, + "step": 1739 + }, + { + "epoch": 0.10466165413533834, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.4777, + "step": 1740 + }, + { + "epoch": 0.1047218045112782, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.4146, + "step": 1741 + }, + { + "epoch": 0.10478195488721805, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.5367, + "step": 1742 + }, + { + "epoch": 0.10484210526315789, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4367, + "step": 1743 + }, + { + "epoch": 0.10490225563909775, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4168, + "step": 1744 + }, + { + "epoch": 0.10496240601503759, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5301, + "step": 1745 + }, + { + "epoch": 0.10502255639097745, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.5045, + "step": 1746 + }, + { + "epoch": 0.10508270676691729, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5383, + "step": 1747 + }, + { + "epoch": 0.10514285714285715, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.4396, + "step": 1748 + }, + { + "epoch": 0.10520300751879699, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.4192, + "step": 1749 + }, + { + "epoch": 0.10526315789473684, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3811, + "step": 1750 + }, + { + "epoch": 0.1053233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4466, + "step": 1751 + }, + { + "epoch": 0.10538345864661654, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3898, + "step": 1752 + }, + { + "epoch": 0.1054436090225564, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.396, + "step": 1753 + }, + { + "epoch": 0.10550375939849624, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.4429, + "step": 1754 + }, + { + "epoch": 0.1055639097744361, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4596, + "step": 1755 + }, + { + "epoch": 0.10562406015037594, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4237, + "step": 1756 + }, + { + "epoch": 0.10568421052631578, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5037, + "step": 1757 + }, + { + "epoch": 0.10574436090225564, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4263, + "step": 1758 + }, + { + "epoch": 0.10580451127819548, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4269, + "step": 1759 + }, + { + "epoch": 0.10586466165413534, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4221, + "step": 1760 + }, + { + "epoch": 0.10592481203007519, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4414, + "step": 1761 + }, + { + "epoch": 0.10598496240601504, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.4537, + "step": 1762 + }, + { + "epoch": 0.10604511278195489, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.5012, + "step": 1763 + }, + { + "epoch": 0.10610526315789473, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.47, + "step": 1764 + }, + { + "epoch": 0.10616541353383459, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.5556, + "step": 1765 + }, + { + "epoch": 0.10622556390977443, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.499, + "step": 1766 + }, + { + "epoch": 0.10628571428571429, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4975, + "step": 1767 + }, + { + "epoch": 0.10634586466165413, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.504, + "step": 1768 + }, + { + "epoch": 0.10640601503759399, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3645, + "step": 1769 + }, + { + "epoch": 0.10646616541353383, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.4637, + "step": 1770 + }, + { + "epoch": 0.10652631578947368, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.5529, + "step": 1771 + }, + { + "epoch": 0.10658646616541353, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4733, + "step": 1772 + }, + { + "epoch": 0.10664661654135338, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.5521, + "step": 1773 + }, + { + "epoch": 0.10670676691729324, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4372, + "step": 1774 + }, + { + "epoch": 0.10676691729323308, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3928, + "step": 1775 + }, + { + "epoch": 0.10682706766917294, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4042, + "step": 1776 + }, + { + "epoch": 0.10688721804511278, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3757, + "step": 1777 + }, + { + "epoch": 0.10694736842105264, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4153, + "step": 1778 + }, + { + "epoch": 0.10700751879699248, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4105, + "step": 1779 + }, + { + "epoch": 0.10706766917293233, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.4217, + "step": 1780 + }, + { + "epoch": 0.10712781954887218, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3743, + "step": 1781 + }, + { + "epoch": 0.10718796992481203, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.651, + "step": 1782 + }, + { + "epoch": 0.10724812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4193, + "step": 1783 + }, + { + "epoch": 0.10730827067669173, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.475, + "step": 1784 + }, + { + "epoch": 0.10736842105263159, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.4699, + "step": 1785 + }, + { + "epoch": 0.10742857142857143, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4635, + "step": 1786 + }, + { + "epoch": 0.10748872180451127, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3699, + "step": 1787 + }, + { + "epoch": 0.10754887218045113, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.4664, + "step": 1788 + }, + { + "epoch": 0.10760902255639097, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.4791, + "step": 1789 + }, + { + "epoch": 0.10766917293233083, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.5336, + "step": 1790 + }, + { + "epoch": 0.10772932330827067, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.4819, + "step": 1791 + }, + { + "epoch": 0.10778947368421053, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.4202, + "step": 1792 + }, + { + "epoch": 0.10784962406015038, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.5256, + "step": 1793 + }, + { + "epoch": 0.10790977443609022, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4831, + "step": 1794 + }, + { + "epoch": 0.10796992481203008, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4602, + "step": 1795 + }, + { + "epoch": 0.10803007518796992, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4832, + "step": 1796 + }, + { + "epoch": 0.10809022556390978, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4552, + "step": 1797 + }, + { + "epoch": 0.10815037593984962, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.5052, + "step": 1798 + }, + { + "epoch": 0.10821052631578948, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4343, + "step": 1799 + }, + { + "epoch": 0.10827067669172932, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4222, + "step": 1800 + }, + { + "epoch": 0.10833082706766917, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.4388, + "step": 1801 + }, + { + "epoch": 0.10839097744360902, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4463, + "step": 1802 + }, + { + "epoch": 0.10845112781954887, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.5811, + "step": 1803 + }, + { + "epoch": 0.10851127819548873, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3773, + "step": 1804 + }, + { + "epoch": 0.10857142857142857, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.5497, + "step": 1805 + }, + { + "epoch": 0.10863157894736843, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.5009, + "step": 1806 + }, + { + "epoch": 0.10869172932330827, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.4353, + "step": 1807 + }, + { + "epoch": 0.10875187969924811, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.5412, + "step": 1808 + }, + { + "epoch": 0.10881203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.4486, + "step": 1809 + }, + { + "epoch": 0.10887218045112781, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3874, + "step": 1810 + }, + { + "epoch": 0.10893233082706767, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.4363, + "step": 1811 + }, + { + "epoch": 0.10899248120300752, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.5266, + "step": 1812 + }, + { + "epoch": 0.10905263157894737, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4408, + "step": 1813 + }, + { + "epoch": 0.10911278195488722, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.508, + "step": 1814 + }, + { + "epoch": 0.10917293233082707, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4643, + "step": 1815 + }, + { + "epoch": 0.10923308270676692, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4265, + "step": 1816 + }, + { + "epoch": 0.10929323308270676, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3897, + "step": 1817 + }, + { + "epoch": 0.10935338345864662, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.5092, + "step": 1818 + }, + { + "epoch": 0.10941353383458646, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3943, + "step": 1819 + }, + { + "epoch": 0.10947368421052632, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4444, + "step": 1820 + }, + { + "epoch": 0.10953383458646616, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.4195, + "step": 1821 + }, + { + "epoch": 0.10959398496240602, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.4702, + "step": 1822 + }, + { + "epoch": 0.10965413533834587, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.649, + "step": 1823 + }, + { + "epoch": 0.10971428571428571, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4235, + "step": 1824 + }, + { + "epoch": 0.10977443609022557, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4063, + "step": 1825 + }, + { + "epoch": 0.10983458646616541, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.5951, + "step": 1826 + }, + { + "epoch": 0.10989473684210527, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.4626, + "step": 1827 + }, + { + "epoch": 0.10995488721804511, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.4927, + "step": 1828 + }, + { + "epoch": 0.11001503759398497, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3916, + "step": 1829 + }, + { + "epoch": 0.11007518796992481, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4956, + "step": 1830 + }, + { + "epoch": 0.11013533834586466, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3981, + "step": 1831 + }, + { + "epoch": 0.11019548872180451, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.4466, + "step": 1832 + }, + { + "epoch": 0.11025563909774436, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4418, + "step": 1833 + }, + { + "epoch": 0.11031578947368421, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4436, + "step": 1834 + }, + { + "epoch": 0.11037593984962406, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.5247, + "step": 1835 + }, + { + "epoch": 0.11043609022556392, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.4443, + "step": 1836 + }, + { + "epoch": 0.11049624060150376, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.4959, + "step": 1837 + }, + { + "epoch": 0.1105563909774436, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.4243, + "step": 1838 + }, + { + "epoch": 0.11061654135338346, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3958, + "step": 1839 + }, + { + "epoch": 0.1106766917293233, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4122, + "step": 1840 + }, + { + "epoch": 0.11073684210526316, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.5346, + "step": 1841 + }, + { + "epoch": 0.110796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.4695, + "step": 1842 + }, + { + "epoch": 0.11085714285714286, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.4906, + "step": 1843 + }, + { + "epoch": 0.1109172932330827, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3816, + "step": 1844 + }, + { + "epoch": 0.11097744360902255, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.5307, + "step": 1845 + }, + { + "epoch": 0.11103759398496241, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3886, + "step": 1846 + }, + { + "epoch": 0.11109774436090225, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4617, + "step": 1847 + }, + { + "epoch": 0.11115789473684211, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.4269, + "step": 1848 + }, + { + "epoch": 0.11121804511278195, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4589, + "step": 1849 + }, + { + "epoch": 0.11127819548872181, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4826, + "step": 1850 + }, + { + "epoch": 0.11133834586466165, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4424, + "step": 1851 + }, + { + "epoch": 0.11139849624060151, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4597, + "step": 1852 + }, + { + "epoch": 0.11145864661654135, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.3091, + "step": 1853 + }, + { + "epoch": 0.1115187969924812, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4079, + "step": 1854 + }, + { + "epoch": 0.11157894736842106, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5101, + "step": 1855 + }, + { + "epoch": 0.1116390977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.5736, + "step": 1856 + }, + { + "epoch": 0.11169924812030076, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4223, + "step": 1857 + }, + { + "epoch": 0.1117593984962406, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4241, + "step": 1858 + }, + { + "epoch": 0.11181954887218046, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.5582, + "step": 1859 + }, + { + "epoch": 0.1118796992481203, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.546, + "step": 1860 + }, + { + "epoch": 0.11193984962406014, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4393, + "step": 1861 + }, + { + "epoch": 0.112, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.5101, + "step": 1862 + }, + { + "epoch": 0.11206015037593985, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4031, + "step": 1863 + }, + { + "epoch": 0.1121203007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.4457, + "step": 1864 + }, + { + "epoch": 0.11218045112781955, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4971, + "step": 1865 + }, + { + "epoch": 0.1122406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.4259, + "step": 1866 + }, + { + "epoch": 0.11230075187969925, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.4946, + "step": 1867 + }, + { + "epoch": 0.11236090225563909, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.5947, + "step": 1868 + }, + { + "epoch": 0.11242105263157895, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4836, + "step": 1869 + }, + { + "epoch": 0.11248120300751879, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.4822, + "step": 1870 + }, + { + "epoch": 0.11254135338345865, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4818, + "step": 1871 + }, + { + "epoch": 0.1126015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4329, + "step": 1872 + }, + { + "epoch": 0.11266165413533835, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4104, + "step": 1873 + }, + { + "epoch": 0.1127218045112782, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4089, + "step": 1874 + }, + { + "epoch": 0.11278195488721804, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.4328, + "step": 1875 + }, + { + "epoch": 0.1128421052631579, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.5181, + "step": 1876 + }, + { + "epoch": 0.11290225563909774, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.459, + "step": 1877 + }, + { + "epoch": 0.1129624060150376, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.504, + "step": 1878 + }, + { + "epoch": 0.11302255639097744, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.4051, + "step": 1879 + }, + { + "epoch": 0.1130827067669173, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3881, + "step": 1880 + }, + { + "epoch": 0.11314285714285714, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.396, + "step": 1881 + }, + { + "epoch": 0.11320300751879699, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3928, + "step": 1882 + }, + { + "epoch": 0.11326315789473684, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4611, + "step": 1883 + }, + { + "epoch": 0.11332330827067669, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.5285, + "step": 1884 + }, + { + "epoch": 0.11338345864661654, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3901, + "step": 1885 + }, + { + "epoch": 0.11344360902255639, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.5006, + "step": 1886 + }, + { + "epoch": 0.11350375939849625, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.4359, + "step": 1887 + }, + { + "epoch": 0.11356390977443609, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.3843, + "step": 1888 + }, + { + "epoch": 0.11362406015037595, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3191, + "step": 1889 + }, + { + "epoch": 0.11368421052631579, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4782, + "step": 1890 + }, + { + "epoch": 0.11374436090225563, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.5135, + "step": 1891 + }, + { + "epoch": 0.11380451127819549, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.492, + "step": 1892 + }, + { + "epoch": 0.11386466165413534, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.4227, + "step": 1893 + }, + { + "epoch": 0.11392481203007519, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.4297, + "step": 1894 + }, + { + "epoch": 0.11398496240601504, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3857, + "step": 1895 + }, + { + "epoch": 0.1140451127819549, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4526, + "step": 1896 + }, + { + "epoch": 0.11410526315789474, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.457, + "step": 1897 + }, + { + "epoch": 0.11416541353383458, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.4217, + "step": 1898 + }, + { + "epoch": 0.11422556390977444, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4046, + "step": 1899 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.4369, + "step": 1900 + }, + { + "epoch": 0.11434586466165414, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4947, + "step": 1901 + }, + { + "epoch": 0.11440601503759398, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.4167, + "step": 1902 + }, + { + "epoch": 0.11446616541353384, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4991, + "step": 1903 + }, + { + "epoch": 0.11452631578947368, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.6528, + "step": 1904 + }, + { + "epoch": 0.11458646616541353, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.4449, + "step": 1905 + }, + { + "epoch": 0.11464661654135339, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.4079, + "step": 1906 + }, + { + "epoch": 0.11470676691729323, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4285, + "step": 1907 + }, + { + "epoch": 0.11476691729323309, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3986, + "step": 1908 + }, + { + "epoch": 0.11482706766917293, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4578, + "step": 1909 + }, + { + "epoch": 0.11488721804511279, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4401, + "step": 1910 + }, + { + "epoch": 0.11494736842105263, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.472, + "step": 1911 + }, + { + "epoch": 0.11500751879699248, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4365, + "step": 1912 + }, + { + "epoch": 0.11506766917293233, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4131, + "step": 1913 + }, + { + "epoch": 0.11512781954887218, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4409, + "step": 1914 + }, + { + "epoch": 0.11518796992481203, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4181, + "step": 1915 + }, + { + "epoch": 0.11524812030075188, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3738, + "step": 1916 + }, + { + "epoch": 0.11530827067669173, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4521, + "step": 1917 + }, + { + "epoch": 0.11536842105263158, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.5469, + "step": 1918 + }, + { + "epoch": 0.11542857142857142, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.5491, + "step": 1919 + }, + { + "epoch": 0.11548872180451128, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4835, + "step": 1920 + }, + { + "epoch": 0.11554887218045112, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5069, + "step": 1921 + }, + { + "epoch": 0.11560902255639098, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4397, + "step": 1922 + }, + { + "epoch": 0.11566917293233082, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.5076, + "step": 1923 + }, + { + "epoch": 0.11572932330827068, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.376, + "step": 1924 + }, + { + "epoch": 0.11578947368421053, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4355, + "step": 1925 + }, + { + "epoch": 0.11584962406015037, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4003, + "step": 1926 + }, + { + "epoch": 0.11590977443609023, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.3939, + "step": 1927 + }, + { + "epoch": 0.11596992481203007, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.5135, + "step": 1928 + }, + { + "epoch": 0.11603007518796993, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4729, + "step": 1929 + }, + { + "epoch": 0.11609022556390977, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.5225, + "step": 1930 + }, + { + "epoch": 0.11615037593984963, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.5291, + "step": 1931 + }, + { + "epoch": 0.11621052631578947, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.5231, + "step": 1932 + }, + { + "epoch": 0.11627067669172933, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4595, + "step": 1933 + }, + { + "epoch": 0.11633082706766917, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4815, + "step": 1934 + }, + { + "epoch": 0.11639097744360902, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4183, + "step": 1935 + }, + { + "epoch": 0.11645112781954887, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4579, + "step": 1936 + }, + { + "epoch": 0.11651127819548872, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.5006, + "step": 1937 + }, + { + "epoch": 0.11657142857142858, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.6123, + "step": 1938 + }, + { + "epoch": 0.11663157894736842, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3887, + "step": 1939 + }, + { + "epoch": 0.11669172932330828, + "grad_norm": 0.4140625, + "learning_rate": 3e-05, + "loss": 1.4443, + "step": 1940 + }, + { + "epoch": 0.11675187969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4888, + "step": 1941 + }, + { + "epoch": 0.11681203007518796, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.4346, + "step": 1942 + }, + { + "epoch": 0.11687218045112782, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3974, + "step": 1943 + }, + { + "epoch": 0.11693233082706767, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4482, + "step": 1944 + }, + { + "epoch": 0.11699248120300752, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.4731, + "step": 1945 + }, + { + "epoch": 0.11705263157894737, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4948, + "step": 1946 + }, + { + "epoch": 0.11711278195488722, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4104, + "step": 1947 + }, + { + "epoch": 0.11717293233082707, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.4045, + "step": 1948 + }, + { + "epoch": 0.11723308270676691, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.4092, + "step": 1949 + }, + { + "epoch": 0.11729323308270677, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.4446, + "step": 1950 + }, + { + "epoch": 0.11735338345864661, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.4108, + "step": 1951 + }, + { + "epoch": 0.11741353383458647, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4332, + "step": 1952 + }, + { + "epoch": 0.11747368421052631, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.381, + "step": 1953 + }, + { + "epoch": 0.11753383458646617, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4284, + "step": 1954 + }, + { + "epoch": 0.11759398496240601, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.404, + "step": 1955 + }, + { + "epoch": 0.11765413533834586, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4489, + "step": 1956 + }, + { + "epoch": 0.11771428571428572, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3807, + "step": 1957 + }, + { + "epoch": 0.11777443609022556, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.448, + "step": 1958 + }, + { + "epoch": 0.11783458646616542, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4867, + "step": 1959 + }, + { + "epoch": 0.11789473684210526, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.5273, + "step": 1960 + }, + { + "epoch": 0.11795488721804512, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4829, + "step": 1961 + }, + { + "epoch": 0.11801503759398496, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.4165, + "step": 1962 + }, + { + "epoch": 0.1180751879699248, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.4884, + "step": 1963 + }, + { + "epoch": 0.11813533834586466, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.5027, + "step": 1964 + }, + { + "epoch": 0.1181954887218045, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.5004, + "step": 1965 + }, + { + "epoch": 0.11825563909774436, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3699, + "step": 1966 + }, + { + "epoch": 0.11831578947368421, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4151, + "step": 1967 + }, + { + "epoch": 0.11837593984962407, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.4244, + "step": 1968 + }, + { + "epoch": 0.11843609022556391, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.4354, + "step": 1969 + }, + { + "epoch": 0.11849624060150377, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.4138, + "step": 1970 + }, + { + "epoch": 0.11855639097744361, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4452, + "step": 1971 + }, + { + "epoch": 0.11861654135338345, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4042, + "step": 1972 + }, + { + "epoch": 0.11867669172932331, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3954, + "step": 1973 + }, + { + "epoch": 0.11873684210526315, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4506, + "step": 1974 + }, + { + "epoch": 0.11879699248120301, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.4436, + "step": 1975 + }, + { + "epoch": 0.11885714285714286, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.4132, + "step": 1976 + }, + { + "epoch": 0.11891729323308271, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4857, + "step": 1977 + }, + { + "epoch": 0.11897744360902256, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3986, + "step": 1978 + }, + { + "epoch": 0.1190375939849624, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4512, + "step": 1979 + }, + { + "epoch": 0.11909774436090226, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4517, + "step": 1980 + }, + { + "epoch": 0.1191578947368421, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3972, + "step": 1981 + }, + { + "epoch": 0.11921804511278196, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.4195, + "step": 1982 + }, + { + "epoch": 0.1192781954887218, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.4341, + "step": 1983 + }, + { + "epoch": 0.11933834586466166, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.516, + "step": 1984 + }, + { + "epoch": 0.1193984962406015, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.4606, + "step": 1985 + }, + { + "epoch": 0.11945864661654135, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4437, + "step": 1986 + }, + { + "epoch": 0.1195187969924812, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.2997, + "step": 1987 + }, + { + "epoch": 0.11957894736842105, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.4187, + "step": 1988 + }, + { + "epoch": 0.1196390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.4015, + "step": 1989 + }, + { + "epoch": 0.11969924812030075, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.5086, + "step": 1990 + }, + { + "epoch": 0.11975939849624061, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3901, + "step": 1991 + }, + { + "epoch": 0.11981954887218045, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.4715, + "step": 1992 + }, + { + "epoch": 0.1198796992481203, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.5141, + "step": 1993 + }, + { + "epoch": 0.11993984962406015, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4358, + "step": 1994 + }, + { + "epoch": 0.12, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4266, + "step": 1995 + }, + { + "epoch": 0.12006015037593985, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4258, + "step": 1996 + }, + { + "epoch": 0.1201203007518797, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.4859, + "step": 1997 + }, + { + "epoch": 0.12018045112781955, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.4692, + "step": 1998 + }, + { + "epoch": 0.1202406015037594, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4597, + "step": 1999 + }, + { + "epoch": 0.12030075187969924, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.5419, + "step": 2000 + }, + { + "epoch": 6.015037593984962e-05, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3841, + "step": 2001 + }, + { + "epoch": 0.00012030075187969925, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.397, + "step": 2002 + }, + { + "epoch": 0.00018045112781954887, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3504, + "step": 2003 + }, + { + "epoch": 0.0002406015037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4475, + "step": 2004 + }, + { + "epoch": 0.0003007518796992481, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4577, + "step": 2005 + }, + { + "epoch": 0.00036090225563909775, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4556, + "step": 2006 + }, + { + "epoch": 0.0004210526315789474, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.4177, + "step": 2007 + }, + { + "epoch": 0.000481203007518797, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4263, + "step": 2008 + }, + { + "epoch": 0.0005413533834586466, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.4416, + "step": 2009 + }, + { + "epoch": 0.0006015037593984962, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.4231, + "step": 2010 + }, + { + "epoch": 0.0006616541353383459, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.3546, + "step": 2011 + }, + { + "epoch": 0.0007218045112781955, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.5351, + "step": 2012 + }, + { + "epoch": 0.0007819548872180451, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4271, + "step": 2013 + }, + { + "epoch": 0.0008421052631578948, + "grad_norm": 0.421875, + "learning_rate": 3e-05, + "loss": 1.4462, + "step": 2014 + }, + { + "epoch": 0.0009022556390977444, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.4295, + "step": 2015 + }, + { + "epoch": 0.000962406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.5384, + "step": 2016 + }, + { + "epoch": 0.0010225563909774435, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3974, + "step": 2017 + }, + { + "epoch": 0.0010827067669172932, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4033, + "step": 2018 + }, + { + "epoch": 0.001142857142857143, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.336, + "step": 2019 + }, + { + "epoch": 0.0012030075187969924, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.4323, + "step": 2020 + }, + { + "epoch": 0.0012631578947368421, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.4604, + "step": 2021 + }, + { + "epoch": 0.0013233082706766918, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.5361, + "step": 2022 + }, + { + "epoch": 0.0013834586466165413, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.5023, + "step": 2023 + }, + { + "epoch": 0.001443609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3755, + "step": 2024 + }, + { + "epoch": 0.0015037593984962407, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.4452, + "step": 2025 + }, + { + "epoch": 0.0015639097744360902, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3984, + "step": 2026 + }, + { + "epoch": 0.0016240601503759399, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.4274, + "step": 2027 + }, + { + "epoch": 0.0016842105263157896, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.4935, + "step": 2028 + }, + { + "epoch": 0.001744360902255639, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3779, + "step": 2029 + }, + { + "epoch": 0.0018045112781954887, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.4175, + "step": 2030 + }, + { + "epoch": 0.0018646616541353384, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.3483, + "step": 2031 + }, + { + "epoch": 0.001924812030075188, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.4068, + "step": 2032 + }, + { + "epoch": 0.0019849624060150374, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4357, + "step": 2033 + }, + { + "epoch": 0.002045112781954887, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3937, + "step": 2034 + }, + { + "epoch": 0.002105263157894737, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4183, + "step": 2035 + }, + { + "epoch": 0.0021654135338345865, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4632, + "step": 2036 + }, + { + "epoch": 0.002225563909774436, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4078, + "step": 2037 + }, + { + "epoch": 0.002285714285714286, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.4411, + "step": 2038 + }, + { + "epoch": 0.002345864661654135, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.5261, + "step": 2039 + }, + { + "epoch": 0.002406015037593985, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.5289, + "step": 2040 + }, + { + "epoch": 0.0024661654135338345, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4763, + "step": 2041 + }, + { + "epoch": 0.0025263157894736842, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.4725, + "step": 2042 + }, + { + "epoch": 0.002586466165413534, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.397, + "step": 2043 + }, + { + "epoch": 0.0026466165413533836, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4433, + "step": 2044 + }, + { + "epoch": 0.002706766917293233, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.4984, + "step": 2045 + }, + { + "epoch": 0.0027669172932330826, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3822, + "step": 2046 + }, + { + "epoch": 0.0028270676691729323, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.5008, + "step": 2047 + }, + { + "epoch": 0.002887218045112782, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3866, + "step": 2048 + }, + { + "epoch": 0.0029473684210526317, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3757, + "step": 2049 + }, + { + "epoch": 0.0030075187969924814, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.4361, + "step": 2050 + }, + { + "epoch": 0.0030676691729323306, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.4425, + "step": 2051 + }, + { + "epoch": 0.0031278195488721803, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.475, + "step": 2052 + }, + { + "epoch": 0.00318796992481203, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.4419, + "step": 2053 + }, + { + "epoch": 0.0032481203007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4524, + "step": 2054 + }, + { + "epoch": 0.0033082706766917294, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.5413, + "step": 2055 + }, + { + "epoch": 0.003368421052631579, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3713, + "step": 2056 + }, + { + "epoch": 0.0034285714285714284, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4255, + "step": 2057 + }, + { + "epoch": 0.003488721804511278, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4328, + "step": 2058 + }, + { + "epoch": 0.003548872180451128, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.424, + "step": 2059 + }, + { + "epoch": 0.0036090225563909775, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.4586, + "step": 2060 + }, + { + "epoch": 0.003669172932330827, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.4274, + "step": 2061 + }, + { + "epoch": 0.003729323308270677, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.5086, + "step": 2062 + }, + { + "epoch": 0.003789473684210526, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4312, + "step": 2063 + }, + { + "epoch": 0.003849624060150376, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4836, + "step": 2064 + }, + { + "epoch": 0.003909774436090226, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3818, + "step": 2065 + }, + { + "epoch": 0.003969924812030075, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.5747, + "step": 2066 + }, + { + "epoch": 0.0040300751879699245, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3984, + "step": 2067 + }, + { + "epoch": 0.004090225563909774, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.404, + "step": 2068 + }, + { + "epoch": 0.004150375939849624, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.43, + "step": 2069 + }, + { + "epoch": 0.004210526315789474, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3871, + "step": 2070 + }, + { + "epoch": 0.004270676691729323, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3546, + "step": 2071 + }, + { + "epoch": 0.004330827067669173, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3172, + "step": 2072 + }, + { + "epoch": 0.004390977443609023, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.5418, + "step": 2073 + }, + { + "epoch": 0.004451127819548872, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3902, + "step": 2074 + }, + { + "epoch": 0.004511278195488722, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.3582, + "step": 2075 + }, + { + "epoch": 0.004571428571428572, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4256, + "step": 2076 + }, + { + "epoch": 0.0046315789473684215, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.5303, + "step": 2077 + }, + { + "epoch": 0.00469172932330827, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.4908, + "step": 2078 + }, + { + "epoch": 0.00475187969924812, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4148, + "step": 2079 + }, + { + "epoch": 0.00481203007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4356, + "step": 2080 + }, + { + "epoch": 0.004872180451127819, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3743, + "step": 2081 + }, + { + "epoch": 0.004932330827067669, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4295, + "step": 2082 + }, + { + "epoch": 0.004992481203007519, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3793, + "step": 2083 + }, + { + "epoch": 0.0050526315789473685, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3969, + "step": 2084 + }, + { + "epoch": 0.005112781954887218, + "grad_norm": 0.42578125, + "learning_rate": 3e-05, + "loss": 1.4566, + "step": 2085 + }, + { + "epoch": 0.005172932330827068, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4173, + "step": 2086 + }, + { + "epoch": 0.005233082706766918, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.3318, + "step": 2087 + }, + { + "epoch": 0.005293233082706767, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.468, + "step": 2088 + }, + { + "epoch": 0.005353383458646617, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4569, + "step": 2089 + }, + { + "epoch": 0.005413533834586466, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.432, + "step": 2090 + }, + { + "epoch": 0.0054736842105263155, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4382, + "step": 2091 + }, + { + "epoch": 0.005533834586466165, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4554, + "step": 2092 + }, + { + "epoch": 0.005593984962406015, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3987, + "step": 2093 + }, + { + "epoch": 0.005654135338345865, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4957, + "step": 2094 + }, + { + "epoch": 0.005714285714285714, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3517, + "step": 2095 + }, + { + "epoch": 0.005774436090225564, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4116, + "step": 2096 + }, + { + "epoch": 0.005834586466165414, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3863, + "step": 2097 + }, + { + "epoch": 0.005894736842105263, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.452, + "step": 2098 + }, + { + "epoch": 0.005954887218045113, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3831, + "step": 2099 + }, + { + "epoch": 0.006015037593984963, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.424, + "step": 2100 + }, + { + "epoch": 0.0060751879699248125, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3771, + "step": 2101 + }, + { + "epoch": 0.006135338345864661, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4374, + "step": 2102 + }, + { + "epoch": 0.006195488721804511, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3245, + "step": 2103 + }, + { + "epoch": 0.006255639097744361, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.4015, + "step": 2104 + }, + { + "epoch": 0.00631578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3726, + "step": 2105 + }, + { + "epoch": 0.00637593984962406, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4412, + "step": 2106 + }, + { + "epoch": 0.00643609022556391, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.3464, + "step": 2107 + }, + { + "epoch": 0.0064962406015037595, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.497, + "step": 2108 + }, + { + "epoch": 0.006556390977443609, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.4438, + "step": 2109 + }, + { + "epoch": 0.006616541353383459, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3856, + "step": 2110 + }, + { + "epoch": 0.0066766917293233086, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.48, + "step": 2111 + }, + { + "epoch": 0.006736842105263158, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4736, + "step": 2112 + }, + { + "epoch": 0.006796992481203007, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3854, + "step": 2113 + }, + { + "epoch": 0.006857142857142857, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.326, + "step": 2114 + }, + { + "epoch": 0.0069172932330827065, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.455, + "step": 2115 + }, + { + "epoch": 0.006977443609022556, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4125, + "step": 2116 + }, + { + "epoch": 0.007037593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.5764, + "step": 2117 + }, + { + "epoch": 0.007097744360902256, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3622, + "step": 2118 + }, + { + "epoch": 0.007157894736842105, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3541, + "step": 2119 + }, + { + "epoch": 0.007218045112781955, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4223, + "step": 2120 + }, + { + "epoch": 0.007278195488721805, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3358, + "step": 2121 + }, + { + "epoch": 0.007338345864661654, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.4015, + "step": 2122 + }, + { + "epoch": 0.007398496240601504, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4677, + "step": 2123 + }, + { + "epoch": 0.007458646616541354, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4168, + "step": 2124 + }, + { + "epoch": 0.007518796992481203, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4074, + "step": 2125 + }, + { + "epoch": 0.007578947368421052, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4063, + "step": 2126 + }, + { + "epoch": 0.007639097744360902, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.3816, + "step": 2127 + }, + { + "epoch": 0.007699248120300752, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3897, + "step": 2128 + }, + { + "epoch": 0.007759398496240601, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.348, + "step": 2129 + }, + { + "epoch": 0.007819548872180452, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4368, + "step": 2130 + }, + { + "epoch": 0.0078796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4535, + "step": 2131 + }, + { + "epoch": 0.00793984962406015, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.4733, + "step": 2132 + }, + { + "epoch": 0.008, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.4323, + "step": 2133 + }, + { + "epoch": 0.008060150375939849, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3878, + "step": 2134 + }, + { + "epoch": 0.0081203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.4321, + "step": 2135 + }, + { + "epoch": 0.008180451127819548, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3698, + "step": 2136 + }, + { + "epoch": 0.008240601503759399, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.434, + "step": 2137 + }, + { + "epoch": 0.008300751879699248, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4758, + "step": 2138 + }, + { + "epoch": 0.008360902255639098, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3885, + "step": 2139 + }, + { + "epoch": 0.008421052631578947, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4101, + "step": 2140 + }, + { + "epoch": 0.008481203007518798, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.5136, + "step": 2141 + }, + { + "epoch": 0.008541353383458647, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.4078, + "step": 2142 + }, + { + "epoch": 0.008601503759398495, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4936, + "step": 2143 + }, + { + "epoch": 0.008661654135338346, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.4199, + "step": 2144 + }, + { + "epoch": 0.008721804511278195, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4574, + "step": 2145 + }, + { + "epoch": 0.008781954887218045, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.4558, + "step": 2146 + }, + { + "epoch": 0.008842105263157894, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4318, + "step": 2147 + }, + { + "epoch": 0.008902255639097745, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3203, + "step": 2148 + }, + { + "epoch": 0.008962406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4272, + "step": 2149 + }, + { + "epoch": 0.009022556390977444, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.429, + "step": 2150 + }, + { + "epoch": 0.009082706766917293, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.44, + "step": 2151 + }, + { + "epoch": 0.009142857142857144, + "grad_norm": 0.3671875, + "learning_rate": 3e-05, + "loss": 1.459, + "step": 2152 + }, + { + "epoch": 0.009203007518796992, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3963, + "step": 2153 + }, + { + "epoch": 0.009263157894736843, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4566, + "step": 2154 + }, + { + "epoch": 0.009323308270676692, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.4641, + "step": 2155 + }, + { + "epoch": 0.00938345864661654, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4726, + "step": 2156 + }, + { + "epoch": 0.009443609022556391, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.41, + "step": 2157 + }, + { + "epoch": 0.00950375939849624, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.3667, + "step": 2158 + }, + { + "epoch": 0.00956390977443609, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4819, + "step": 2159 + }, + { + "epoch": 0.00962406015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.458, + "step": 2160 + }, + { + "epoch": 0.00968421052631579, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4492, + "step": 2161 + }, + { + "epoch": 0.009744360902255639, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.494, + "step": 2162 + }, + { + "epoch": 0.00980451127819549, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.481, + "step": 2163 + }, + { + "epoch": 0.009864661654135338, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.5054, + "step": 2164 + }, + { + "epoch": 0.009924812030075189, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.4446, + "step": 2165 + }, + { + "epoch": 0.009984962406015038, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.434, + "step": 2166 + }, + { + "epoch": 0.010045112781954886, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.4481, + "step": 2167 + }, + { + "epoch": 0.010105263157894737, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4533, + "step": 2168 + }, + { + "epoch": 0.010165413533834586, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4383, + "step": 2169 + }, + { + "epoch": 0.010225563909774436, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3884, + "step": 2170 + }, + { + "epoch": 0.010285714285714285, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4095, + "step": 2171 + }, + { + "epoch": 0.010345864661654136, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.438, + "step": 2172 + }, + { + "epoch": 0.010406015037593985, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3315, + "step": 2173 + }, + { + "epoch": 0.010466165413533835, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.4586, + "step": 2174 + }, + { + "epoch": 0.010526315789473684, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4165, + "step": 2175 + }, + { + "epoch": 0.010586466165413535, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.5058, + "step": 2176 + }, + { + "epoch": 0.010646616541353383, + "grad_norm": 0.4375, + "learning_rate": 3e-05, + "loss": 1.4194, + "step": 2177 + }, + { + "epoch": 0.010706766917293234, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3815, + "step": 2178 + }, + { + "epoch": 0.010766917293233083, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4716, + "step": 2179 + }, + { + "epoch": 0.010827067669172932, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3336, + "step": 2180 + }, + { + "epoch": 0.010887218045112782, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4227, + "step": 2181 + }, + { + "epoch": 0.010947368421052631, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3724, + "step": 2182 + }, + { + "epoch": 0.011007518796992482, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3609, + "step": 2183 + }, + { + "epoch": 0.01106766917293233, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3838, + "step": 2184 + }, + { + "epoch": 0.011127819548872181, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4535, + "step": 2185 + }, + { + "epoch": 0.01118796992481203, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4883, + "step": 2186 + }, + { + "epoch": 0.01124812030075188, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.39, + "step": 2187 + }, + { + "epoch": 0.01130827067669173, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3789, + "step": 2188 + }, + { + "epoch": 0.01136842105263158, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4139, + "step": 2189 + }, + { + "epoch": 0.011428571428571429, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3975, + "step": 2190 + }, + { + "epoch": 0.011488721804511277, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4169, + "step": 2191 + }, + { + "epoch": 0.011548872180451128, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3574, + "step": 2192 + }, + { + "epoch": 0.011609022556390977, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3843, + "step": 2193 + }, + { + "epoch": 0.011669172932330827, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3807, + "step": 2194 + }, + { + "epoch": 0.011729323308270676, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.3947, + "step": 2195 + }, + { + "epoch": 0.011789473684210527, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.3942, + "step": 2196 + }, + { + "epoch": 0.011849624060150376, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4819, + "step": 2197 + }, + { + "epoch": 0.011909774436090226, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.5153, + "step": 2198 + }, + { + "epoch": 0.011969924812030075, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3491, + "step": 2199 + }, + { + "epoch": 0.012030075187969926, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3469, + "step": 2200 + }, + { + "epoch": 0.012090225563909774, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4455, + "step": 2201 + }, + { + "epoch": 0.012150375939849625, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.4299, + "step": 2202 + }, + { + "epoch": 0.012210526315789474, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4086, + "step": 2203 + }, + { + "epoch": 0.012270676691729323, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3944, + "step": 2204 + }, + { + "epoch": 0.012330827067669173, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3114, + "step": 2205 + }, + { + "epoch": 0.012390977443609022, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.4032, + "step": 2206 + }, + { + "epoch": 0.012451127819548873, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.4064, + "step": 2207 + }, + { + "epoch": 0.012511278195488721, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.4342, + "step": 2208 + }, + { + "epoch": 0.012571428571428572, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.3753, + "step": 2209 + }, + { + "epoch": 0.01263157894736842, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3938, + "step": 2210 + }, + { + "epoch": 0.012691729323308271, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3724, + "step": 2211 + }, + { + "epoch": 0.01275187969924812, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.5018, + "step": 2212 + }, + { + "epoch": 0.01281203007518797, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2939, + "step": 2213 + }, + { + "epoch": 0.01287218045112782, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.5019, + "step": 2214 + }, + { + "epoch": 0.012932330827067668, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4115, + "step": 2215 + }, + { + "epoch": 0.012992481203007519, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.573, + "step": 2216 + }, + { + "epoch": 0.013052631578947368, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3512, + "step": 2217 + }, + { + "epoch": 0.013112781954887218, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3746, + "step": 2218 + }, + { + "epoch": 0.013172932330827067, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.472, + "step": 2219 + }, + { + "epoch": 0.013233082706766918, + "grad_norm": 0.40625, + "learning_rate": 3e-05, + "loss": 1.3873, + "step": 2220 + }, + { + "epoch": 0.013293233082706767, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.403, + "step": 2221 + }, + { + "epoch": 0.013353383458646617, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3952, + "step": 2222 + }, + { + "epoch": 0.013413533834586466, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3499, + "step": 2223 + }, + { + "epoch": 0.013473684210526317, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.389, + "step": 2224 + }, + { + "epoch": 0.013533834586466165, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.4308, + "step": 2225 + }, + { + "epoch": 0.013593984962406014, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3578, + "step": 2226 + }, + { + "epoch": 0.013654135338345865, + "grad_norm": 0.462890625, + "learning_rate": 3e-05, + "loss": 1.3362, + "step": 2227 + }, + { + "epoch": 0.013714285714285714, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.393, + "step": 2228 + }, + { + "epoch": 0.013774436090225564, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4454, + "step": 2229 + }, + { + "epoch": 0.013834586466165413, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4097, + "step": 2230 + }, + { + "epoch": 0.013894736842105264, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.4002, + "step": 2231 + }, + { + "epoch": 0.013954887218045112, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.4454, + "step": 2232 + }, + { + "epoch": 0.014015037593984963, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.412, + "step": 2233 + }, + { + "epoch": 0.014075187969924812, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.399, + "step": 2234 + }, + { + "epoch": 0.014135338345864662, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4299, + "step": 2235 + }, + { + "epoch": 0.014195488721804511, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3507, + "step": 2236 + }, + { + "epoch": 0.014255639097744362, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.4232, + "step": 2237 + }, + { + "epoch": 0.01431578947368421, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.423, + "step": 2238 + }, + { + "epoch": 0.01437593984962406, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4768, + "step": 2239 + }, + { + "epoch": 0.01443609022556391, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.4844, + "step": 2240 + }, + { + "epoch": 0.014496240601503759, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.4515, + "step": 2241 + }, + { + "epoch": 0.01455639097744361, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4869, + "step": 2242 + }, + { + "epoch": 0.014616541353383458, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3998, + "step": 2243 + }, + { + "epoch": 0.014676691729323309, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4114, + "step": 2244 + }, + { + "epoch": 0.014736842105263158, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3843, + "step": 2245 + }, + { + "epoch": 0.014796992481203008, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.4197, + "step": 2246 + }, + { + "epoch": 0.014857142857142857, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4781, + "step": 2247 + }, + { + "epoch": 0.014917293233082708, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3449, + "step": 2248 + }, + { + "epoch": 0.014977443609022556, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3701, + "step": 2249 + }, + { + "epoch": 0.015037593984962405, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3043, + "step": 2250 + }, + { + "epoch": 0.015097744360902256, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.4317, + "step": 2251 + }, + { + "epoch": 0.015157894736842105, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4475, + "step": 2252 + }, + { + "epoch": 0.015218045112781955, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3654, + "step": 2253 + }, + { + "epoch": 0.015278195488721804, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.4542, + "step": 2254 + }, + { + "epoch": 0.015338345864661655, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4219, + "step": 2255 + }, + { + "epoch": 0.015398496240601503, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3605, + "step": 2256 + }, + { + "epoch": 0.015458646616541354, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4525, + "step": 2257 + }, + { + "epoch": 0.015518796992481203, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4508, + "step": 2258 + }, + { + "epoch": 0.015578947368421053, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3575, + "step": 2259 + }, + { + "epoch": 0.015639097744360904, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4518, + "step": 2260 + }, + { + "epoch": 0.01569924812030075, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4497, + "step": 2261 + }, + { + "epoch": 0.0157593984962406, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4812, + "step": 2262 + }, + { + "epoch": 0.015819548872180452, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.4224, + "step": 2263 + }, + { + "epoch": 0.0158796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.4065, + "step": 2264 + }, + { + "epoch": 0.01593984962406015, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4066, + "step": 2265 + }, + { + "epoch": 0.016, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3778, + "step": 2266 + }, + { + "epoch": 0.01606015037593985, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.5032, + "step": 2267 + }, + { + "epoch": 0.016120300751879698, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.4829, + "step": 2268 + }, + { + "epoch": 0.01618045112781955, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3884, + "step": 2269 + }, + { + "epoch": 0.0162406015037594, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.3068, + "step": 2270 + }, + { + "epoch": 0.01630075187969925, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4406, + "step": 2271 + }, + { + "epoch": 0.016360902255639097, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.393, + "step": 2272 + }, + { + "epoch": 0.016421052631578947, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4103, + "step": 2273 + }, + { + "epoch": 0.016481203007518798, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3546, + "step": 2274 + }, + { + "epoch": 0.016541353383458645, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4124, + "step": 2275 + }, + { + "epoch": 0.016601503759398496, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.4244, + "step": 2276 + }, + { + "epoch": 0.016661654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3845, + "step": 2277 + }, + { + "epoch": 0.016721804511278197, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.417, + "step": 2278 + }, + { + "epoch": 0.016781954887218044, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3873, + "step": 2279 + }, + { + "epoch": 0.016842105263157894, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.4765, + "step": 2280 + }, + { + "epoch": 0.016902255639097745, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4802, + "step": 2281 + }, + { + "epoch": 0.016962406015037595, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3638, + "step": 2282 + }, + { + "epoch": 0.017022556390977443, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4387, + "step": 2283 + }, + { + "epoch": 0.017082706766917293, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.3287, + "step": 2284 + }, + { + "epoch": 0.017142857142857144, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4652, + "step": 2285 + }, + { + "epoch": 0.01720300751879699, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3987, + "step": 2286 + }, + { + "epoch": 0.01726315789473684, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3835, + "step": 2287 + }, + { + "epoch": 0.017323308270676692, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.3795, + "step": 2288 + }, + { + "epoch": 0.017383458646616542, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.4092, + "step": 2289 + }, + { + "epoch": 0.01744360902255639, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.4693, + "step": 2290 + }, + { + "epoch": 0.01750375939849624, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.419, + "step": 2291 + }, + { + "epoch": 0.01756390977443609, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.388, + "step": 2292 + }, + { + "epoch": 0.01762406015037594, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4695, + "step": 2293 + }, + { + "epoch": 0.01768421052631579, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3409, + "step": 2294 + }, + { + "epoch": 0.01774436090225564, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.4264, + "step": 2295 + }, + { + "epoch": 0.01780451127819549, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3883, + "step": 2296 + }, + { + "epoch": 0.01786466165413534, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3727, + "step": 2297 + }, + { + "epoch": 0.017924812030075187, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.4697, + "step": 2298 + }, + { + "epoch": 0.017984962406015038, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4559, + "step": 2299 + }, + { + "epoch": 0.01804511278195489, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3571, + "step": 2300 + }, + { + "epoch": 0.018105263157894735, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3693, + "step": 2301 + }, + { + "epoch": 0.018165413533834586, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4033, + "step": 2302 + }, + { + "epoch": 0.018225563909774437, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3868, + "step": 2303 + }, + { + "epoch": 0.018285714285714287, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3231, + "step": 2304 + }, + { + "epoch": 0.018345864661654134, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4835, + "step": 2305 + }, + { + "epoch": 0.018406015037593985, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3928, + "step": 2306 + }, + { + "epoch": 0.018466165413533835, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3661, + "step": 2307 + }, + { + "epoch": 0.018526315789473686, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.3747, + "step": 2308 + }, + { + "epoch": 0.018586466165413533, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.372, + "step": 2309 + }, + { + "epoch": 0.018646616541353384, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4221, + "step": 2310 + }, + { + "epoch": 0.018706766917293234, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.4095, + "step": 2311 + }, + { + "epoch": 0.01876691729323308, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.401, + "step": 2312 + }, + { + "epoch": 0.018827067669172932, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3673, + "step": 2313 + }, + { + "epoch": 0.018887218045112782, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4254, + "step": 2314 + }, + { + "epoch": 0.018947368421052633, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4332, + "step": 2315 + }, + { + "epoch": 0.01900751879699248, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.3741, + "step": 2316 + }, + { + "epoch": 0.01906766917293233, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3689, + "step": 2317 + }, + { + "epoch": 0.01912781954887218, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.4501, + "step": 2318 + }, + { + "epoch": 0.01918796992481203, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3984, + "step": 2319 + }, + { + "epoch": 0.01924812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4181, + "step": 2320 + }, + { + "epoch": 0.01930827067669173, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3776, + "step": 2321 + }, + { + "epoch": 0.01936842105263158, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4029, + "step": 2322 + }, + { + "epoch": 0.019428571428571427, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3875, + "step": 2323 + }, + { + "epoch": 0.019488721804511278, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.4143, + "step": 2324 + }, + { + "epoch": 0.019548872180451128, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.5071, + "step": 2325 + }, + { + "epoch": 0.01960902255639098, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3506, + "step": 2326 + }, + { + "epoch": 0.019669172932330826, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3618, + "step": 2327 + }, + { + "epoch": 0.019729323308270676, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3122, + "step": 2328 + }, + { + "epoch": 0.019789473684210527, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.4052, + "step": 2329 + }, + { + "epoch": 0.019849624060150377, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.4477, + "step": 2330 + }, + { + "epoch": 0.019909774436090225, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.3964, + "step": 2331 + }, + { + "epoch": 0.019969924812030075, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3516, + "step": 2332 + }, + { + "epoch": 0.020030075187969926, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4351, + "step": 2333 + }, + { + "epoch": 0.020090225563909773, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.4076, + "step": 2334 + }, + { + "epoch": 0.020150375939849623, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3463, + "step": 2335 + }, + { + "epoch": 0.020210526315789474, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3585, + "step": 2336 + }, + { + "epoch": 0.020270676691729324, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4127, + "step": 2337 + }, + { + "epoch": 0.02033082706766917, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.4255, + "step": 2338 + }, + { + "epoch": 0.020390977443609022, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.4537, + "step": 2339 + }, + { + "epoch": 0.020451127819548873, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.4565, + "step": 2340 + }, + { + "epoch": 0.020511278195488723, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3693, + "step": 2341 + }, + { + "epoch": 0.02057142857142857, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4012, + "step": 2342 + }, + { + "epoch": 0.02063157894736842, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4802, + "step": 2343 + }, + { + "epoch": 0.02069172932330827, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3632, + "step": 2344 + }, + { + "epoch": 0.02075187969924812, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.445, + "step": 2345 + }, + { + "epoch": 0.02081203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4398, + "step": 2346 + }, + { + "epoch": 0.02087218045112782, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3519, + "step": 2347 + }, + { + "epoch": 0.02093233082706767, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3845, + "step": 2348 + }, + { + "epoch": 0.020992481203007517, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4433, + "step": 2349 + }, + { + "epoch": 0.021052631578947368, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.4026, + "step": 2350 + }, + { + "epoch": 0.02111278195488722, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3379, + "step": 2351 + }, + { + "epoch": 0.02117293233082707, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.405, + "step": 2352 + }, + { + "epoch": 0.021233082706766916, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.484, + "step": 2353 + }, + { + "epoch": 0.021293233082706767, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4282, + "step": 2354 + }, + { + "epoch": 0.021353383458646617, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.343, + "step": 2355 + }, + { + "epoch": 0.021413533834586468, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.425, + "step": 2356 + }, + { + "epoch": 0.021473684210526315, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3903, + "step": 2357 + }, + { + "epoch": 0.021533834586466166, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.395, + "step": 2358 + }, + { + "epoch": 0.021593984962406016, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3897, + "step": 2359 + }, + { + "epoch": 0.021654135338345863, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4442, + "step": 2360 + }, + { + "epoch": 0.021714285714285714, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4294, + "step": 2361 + }, + { + "epoch": 0.021774436090225564, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4128, + "step": 2362 + }, + { + "epoch": 0.021834586466165415, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4028, + "step": 2363 + }, + { + "epoch": 0.021894736842105262, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3622, + "step": 2364 + }, + { + "epoch": 0.021954887218045113, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3564, + "step": 2365 + }, + { + "epoch": 0.022015037593984963, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.431, + "step": 2366 + }, + { + "epoch": 0.022075187969924814, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4735, + "step": 2367 + }, + { + "epoch": 0.02213533834586466, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.4707, + "step": 2368 + }, + { + "epoch": 0.02219548872180451, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3808, + "step": 2369 + }, + { + "epoch": 0.022255639097744362, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3938, + "step": 2370 + }, + { + "epoch": 0.02231578947368421, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2976, + "step": 2371 + }, + { + "epoch": 0.02237593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3382, + "step": 2372 + }, + { + "epoch": 0.02243609022556391, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.379, + "step": 2373 + }, + { + "epoch": 0.02249624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3869, + "step": 2374 + }, + { + "epoch": 0.022556390977443608, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3709, + "step": 2375 + }, + { + "epoch": 0.02261654135338346, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3878, + "step": 2376 + }, + { + "epoch": 0.02267669172932331, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3477, + "step": 2377 + }, + { + "epoch": 0.02273684210526316, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3674, + "step": 2378 + }, + { + "epoch": 0.022796992481203007, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.3974, + "step": 2379 + }, + { + "epoch": 0.022857142857142857, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.4088, + "step": 2380 + }, + { + "epoch": 0.022917293233082708, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3551, + "step": 2381 + }, + { + "epoch": 0.022977443609022555, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3855, + "step": 2382 + }, + { + "epoch": 0.023037593984962405, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4488, + "step": 2383 + }, + { + "epoch": 0.023097744360902256, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.382, + "step": 2384 + }, + { + "epoch": 0.023157894736842106, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.5112, + "step": 2385 + }, + { + "epoch": 0.023218045112781954, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3727, + "step": 2386 + }, + { + "epoch": 0.023278195488721804, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4118, + "step": 2387 + }, + { + "epoch": 0.023338345864661655, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3687, + "step": 2388 + }, + { + "epoch": 0.023398496240601505, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.4343, + "step": 2389 + }, + { + "epoch": 0.023458646616541352, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3734, + "step": 2390 + }, + { + "epoch": 0.023518796992481203, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3332, + "step": 2391 + }, + { + "epoch": 0.023578947368421053, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.335, + "step": 2392 + }, + { + "epoch": 0.0236390977443609, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.452, + "step": 2393 + }, + { + "epoch": 0.02369924812030075, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.4025, + "step": 2394 + }, + { + "epoch": 0.0237593984962406, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4288, + "step": 2395 + }, + { + "epoch": 0.023819548872180452, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4948, + "step": 2396 + }, + { + "epoch": 0.0238796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.4256, + "step": 2397 + }, + { + "epoch": 0.02393984962406015, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.378, + "step": 2398 + }, + { + "epoch": 0.024, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3707, + "step": 2399 + }, + { + "epoch": 0.02406015037593985, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4455, + "step": 2400 + }, + { + "epoch": 0.024120300751879698, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4088, + "step": 2401 + }, + { + "epoch": 0.02418045112781955, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3609, + "step": 2402 + }, + { + "epoch": 0.0242406015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.403, + "step": 2403 + }, + { + "epoch": 0.02430075187969925, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.4271, + "step": 2404 + }, + { + "epoch": 0.024360902255639097, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3563, + "step": 2405 + }, + { + "epoch": 0.024421052631578948, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4106, + "step": 2406 + }, + { + "epoch": 0.024481203007518798, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.4486, + "step": 2407 + }, + { + "epoch": 0.024541353383458645, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.3846, + "step": 2408 + }, + { + "epoch": 0.024601503759398496, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.393, + "step": 2409 + }, + { + "epoch": 0.024661654135338346, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.4084, + "step": 2410 + }, + { + "epoch": 0.024721804511278197, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4226, + "step": 2411 + }, + { + "epoch": 0.024781954887218044, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.4351, + "step": 2412 + }, + { + "epoch": 0.024842105263157895, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3565, + "step": 2413 + }, + { + "epoch": 0.024902255639097745, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3882, + "step": 2414 + }, + { + "epoch": 0.024962406015037596, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.5479, + "step": 2415 + }, + { + "epoch": 0.025022556390977443, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4413, + "step": 2416 + }, + { + "epoch": 0.025082706766917293, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.3994, + "step": 2417 + }, + { + "epoch": 0.025142857142857144, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4384, + "step": 2418 + }, + { + "epoch": 0.02520300751879699, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.4011, + "step": 2419 + }, + { + "epoch": 0.02526315789473684, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.4277, + "step": 2420 + }, + { + "epoch": 0.025323308270676692, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3038, + "step": 2421 + }, + { + "epoch": 0.025383458646616543, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3896, + "step": 2422 + }, + { + "epoch": 0.02544360902255639, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3816, + "step": 2423 + }, + { + "epoch": 0.02550375939849624, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.4199, + "step": 2424 + }, + { + "epoch": 0.02556390977443609, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.4298, + "step": 2425 + }, + { + "epoch": 0.02562406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.396, + "step": 2426 + }, + { + "epoch": 0.02568421052631579, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.428, + "step": 2427 + }, + { + "epoch": 0.02574436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.4054, + "step": 2428 + }, + { + "epoch": 0.02580451127819549, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.345, + "step": 2429 + }, + { + "epoch": 0.025864661654135337, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.4002, + "step": 2430 + }, + { + "epoch": 0.025924812030075187, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.3056, + "step": 2431 + }, + { + "epoch": 0.025984962406015038, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3954, + "step": 2432 + }, + { + "epoch": 0.02604511278195489, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.4206, + "step": 2433 + }, + { + "epoch": 0.026105263157894736, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3398, + "step": 2434 + }, + { + "epoch": 0.026165413533834586, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4739, + "step": 2435 + }, + { + "epoch": 0.026225563909774437, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3305, + "step": 2436 + }, + { + "epoch": 0.026285714285714287, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4022, + "step": 2437 + }, + { + "epoch": 0.026345864661654134, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3788, + "step": 2438 + }, + { + "epoch": 0.026406015037593985, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4074, + "step": 2439 + }, + { + "epoch": 0.026466165413533835, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4169, + "step": 2440 + }, + { + "epoch": 0.026526315789473683, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4134, + "step": 2441 + }, + { + "epoch": 0.026586466165413533, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4907, + "step": 2442 + }, + { + "epoch": 0.026646616541353384, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3623, + "step": 2443 + }, + { + "epoch": 0.026706766917293234, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4738, + "step": 2444 + }, + { + "epoch": 0.02676691729323308, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4004, + "step": 2445 + }, + { + "epoch": 0.026827067669172932, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3975, + "step": 2446 + }, + { + "epoch": 0.026887218045112782, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.3857, + "step": 2447 + }, + { + "epoch": 0.026947368421052633, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3885, + "step": 2448 + }, + { + "epoch": 0.02700751879699248, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2989, + "step": 2449 + }, + { + "epoch": 0.02706766917293233, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.5019, + "step": 2450 + }, + { + "epoch": 0.02712781954887218, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4286, + "step": 2451 + }, + { + "epoch": 0.02718796992481203, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.3891, + "step": 2452 + }, + { + "epoch": 0.02724812030075188, + "grad_norm": 0.41015625, + "learning_rate": 3e-05, + "loss": 1.3491, + "step": 2453 + }, + { + "epoch": 0.02730827067669173, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.3877, + "step": 2454 + }, + { + "epoch": 0.02736842105263158, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3806, + "step": 2455 + }, + { + "epoch": 0.027428571428571427, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3269, + "step": 2456 + }, + { + "epoch": 0.027488721804511278, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3526, + "step": 2457 + }, + { + "epoch": 0.02754887218045113, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3973, + "step": 2458 + }, + { + "epoch": 0.02760902255639098, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.4207, + "step": 2459 + }, + { + "epoch": 0.027669172932330826, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4263, + "step": 2460 + }, + { + "epoch": 0.027729323308270677, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3751, + "step": 2461 + }, + { + "epoch": 0.027789473684210527, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4429, + "step": 2462 + }, + { + "epoch": 0.027849624060150378, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.433, + "step": 2463 + }, + { + "epoch": 0.027909774436090225, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4265, + "step": 2464 + }, + { + "epoch": 0.027969924812030075, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3285, + "step": 2465 + }, + { + "epoch": 0.028030075187969926, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.5126, + "step": 2466 + }, + { + "epoch": 0.028090225563909773, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3652, + "step": 2467 + }, + { + "epoch": 0.028150375939849624, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3784, + "step": 2468 + }, + { + "epoch": 0.028210526315789474, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4471, + "step": 2469 + }, + { + "epoch": 0.028270676691729325, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4077, + "step": 2470 + }, + { + "epoch": 0.028330827067669172, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4036, + "step": 2471 + }, + { + "epoch": 0.028390977443609022, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3457, + "step": 2472 + }, + { + "epoch": 0.028451127819548873, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4037, + "step": 2473 + }, + { + "epoch": 0.028511278195488723, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3534, + "step": 2474 + }, + { + "epoch": 0.02857142857142857, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3874, + "step": 2475 + }, + { + "epoch": 0.02863157894736842, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.3669, + "step": 2476 + }, + { + "epoch": 0.02869172932330827, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.4573, + "step": 2477 + }, + { + "epoch": 0.02875187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.4818, + "step": 2478 + }, + { + "epoch": 0.02881203007518797, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3688, + "step": 2479 + }, + { + "epoch": 0.02887218045112782, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3766, + "step": 2480 + }, + { + "epoch": 0.02893233082706767, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4236, + "step": 2481 + }, + { + "epoch": 0.028992481203007518, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3923, + "step": 2482 + }, + { + "epoch": 0.029052631578947368, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3995, + "step": 2483 + }, + { + "epoch": 0.02911278195488722, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.4044, + "step": 2484 + }, + { + "epoch": 0.02917293233082707, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3724, + "step": 2485 + }, + { + "epoch": 0.029233082706766916, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.4661, + "step": 2486 + }, + { + "epoch": 0.029293233082706767, + "grad_norm": 0.40625, + "learning_rate": 3e-05, + "loss": 1.3538, + "step": 2487 + }, + { + "epoch": 0.029353383458646617, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.5083, + "step": 2488 + }, + { + "epoch": 0.029413533834586465, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4587, + "step": 2489 + }, + { + "epoch": 0.029473684210526315, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.36, + "step": 2490 + }, + { + "epoch": 0.029533834586466166, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3989, + "step": 2491 + }, + { + "epoch": 0.029593984962406016, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.333, + "step": 2492 + }, + { + "epoch": 0.029654135338345863, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3559, + "step": 2493 + }, + { + "epoch": 0.029714285714285714, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.4065, + "step": 2494 + }, + { + "epoch": 0.029774436090225564, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.4122, + "step": 2495 + }, + { + "epoch": 0.029834586466165415, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3239, + "step": 2496 + }, + { + "epoch": 0.029894736842105262, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4377, + "step": 2497 + }, + { + "epoch": 0.029954887218045113, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3431, + "step": 2498 + }, + { + "epoch": 0.030015037593984963, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4095, + "step": 2499 + }, + { + "epoch": 0.03007518796992481, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3789, + "step": 2500 + }, + { + "epoch": 0.03013533834586466, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3467, + "step": 2501 + }, + { + "epoch": 0.03019548872180451, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.4136, + "step": 2502 + }, + { + "epoch": 0.030255639097744362, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3116, + "step": 2503 + }, + { + "epoch": 0.03031578947368421, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3585, + "step": 2504 + }, + { + "epoch": 0.03037593984962406, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.4599, + "step": 2505 + }, + { + "epoch": 0.03043609022556391, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3193, + "step": 2506 + }, + { + "epoch": 0.03049624060150376, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.4251, + "step": 2507 + }, + { + "epoch": 0.030556390977443608, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.3712, + "step": 2508 + }, + { + "epoch": 0.03061654135338346, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3544, + "step": 2509 + }, + { + "epoch": 0.03067669172932331, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.5031, + "step": 2510 + }, + { + "epoch": 0.030736842105263156, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3538, + "step": 2511 + }, + { + "epoch": 0.030796992481203007, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.4149, + "step": 2512 + }, + { + "epoch": 0.030857142857142857, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3369, + "step": 2513 + }, + { + "epoch": 0.030917293233082708, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.4265, + "step": 2514 + }, + { + "epoch": 0.030977443609022555, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.4257, + "step": 2515 + }, + { + "epoch": 0.031037593984962406, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4122, + "step": 2516 + }, + { + "epoch": 0.031097744360902256, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4102, + "step": 2517 + }, + { + "epoch": 0.031157894736842107, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3361, + "step": 2518 + }, + { + "epoch": 0.031218045112781954, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4152, + "step": 2519 + }, + { + "epoch": 0.03127819548872181, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4947, + "step": 2520 + }, + { + "epoch": 0.03133834586466165, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3875, + "step": 2521 + }, + { + "epoch": 0.0313984962406015, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.356, + "step": 2522 + }, + { + "epoch": 0.03145864661654135, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.5039, + "step": 2523 + }, + { + "epoch": 0.0315187969924812, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3773, + "step": 2524 + }, + { + "epoch": 0.031578947368421054, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.399, + "step": 2525 + }, + { + "epoch": 0.031639097744360904, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3623, + "step": 2526 + }, + { + "epoch": 0.031699248120300755, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.4206, + "step": 2527 + }, + { + "epoch": 0.0317593984962406, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4087, + "step": 2528 + }, + { + "epoch": 0.03181954887218045, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.4315, + "step": 2529 + }, + { + "epoch": 0.0318796992481203, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3934, + "step": 2530 + }, + { + "epoch": 0.03193984962406015, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4118, + "step": 2531 + }, + { + "epoch": 0.032, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3956, + "step": 2532 + }, + { + "epoch": 0.03206015037593985, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.4568, + "step": 2533 + }, + { + "epoch": 0.0321203007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3658, + "step": 2534 + }, + { + "epoch": 0.032180451127819545, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3535, + "step": 2535 + }, + { + "epoch": 0.032240601503759396, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4024, + "step": 2536 + }, + { + "epoch": 0.03230075187969925, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4524, + "step": 2537 + }, + { + "epoch": 0.0323609022556391, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2458, + "step": 2538 + }, + { + "epoch": 0.03242105263157895, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3482, + "step": 2539 + }, + { + "epoch": 0.0324812030075188, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3819, + "step": 2540 + }, + { + "epoch": 0.03254135338345865, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.5518, + "step": 2541 + }, + { + "epoch": 0.0326015037593985, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3295, + "step": 2542 + }, + { + "epoch": 0.03266165413533834, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3595, + "step": 2543 + }, + { + "epoch": 0.032721804511278194, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3623, + "step": 2544 + }, + { + "epoch": 0.032781954887218044, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.475, + "step": 2545 + }, + { + "epoch": 0.032842105263157895, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4539, + "step": 2546 + }, + { + "epoch": 0.032902255639097745, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.298, + "step": 2547 + }, + { + "epoch": 0.032962406015037596, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.467, + "step": 2548 + }, + { + "epoch": 0.033022556390977446, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3594, + "step": 2549 + }, + { + "epoch": 0.03308270676691729, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3299, + "step": 2550 + }, + { + "epoch": 0.03314285714285714, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3639, + "step": 2551 + }, + { + "epoch": 0.03320300751879699, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3951, + "step": 2552 + }, + { + "epoch": 0.03326315789473684, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.344, + "step": 2553 + }, + { + "epoch": 0.03332330827067669, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3741, + "step": 2554 + }, + { + "epoch": 0.03338345864661654, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.388, + "step": 2555 + }, + { + "epoch": 0.03344360902255639, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.403, + "step": 2556 + }, + { + "epoch": 0.033503759398496244, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3823, + "step": 2557 + }, + { + "epoch": 0.03356390977443609, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3472, + "step": 2558 + }, + { + "epoch": 0.03362406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.4484, + "step": 2559 + }, + { + "epoch": 0.03368421052631579, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3891, + "step": 2560 + }, + { + "epoch": 0.03374436090225564, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.3656, + "step": 2561 + }, + { + "epoch": 0.03380451127819549, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.4298, + "step": 2562 + }, + { + "epoch": 0.03386466165413534, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.4039, + "step": 2563 + }, + { + "epoch": 0.03392481203007519, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4438, + "step": 2564 + }, + { + "epoch": 0.033984962406015035, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 2565 + }, + { + "epoch": 0.034045112781954885, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.459, + "step": 2566 + }, + { + "epoch": 0.034105263157894736, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.417, + "step": 2567 + }, + { + "epoch": 0.034165413533834586, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3823, + "step": 2568 + }, + { + "epoch": 0.03422556390977444, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.4957, + "step": 2569 + }, + { + "epoch": 0.03428571428571429, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.3653, + "step": 2570 + }, + { + "epoch": 0.03434586466165414, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4188, + "step": 2571 + }, + { + "epoch": 0.03440601503759398, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.4251, + "step": 2572 + }, + { + "epoch": 0.03446616541353383, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.4062, + "step": 2573 + }, + { + "epoch": 0.03452631578947368, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.4015, + "step": 2574 + }, + { + "epoch": 0.03458646616541353, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.4789, + "step": 2575 + }, + { + "epoch": 0.034646616541353384, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.4563, + "step": 2576 + }, + { + "epoch": 0.034706766917293234, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3613, + "step": 2577 + }, + { + "epoch": 0.034766917293233085, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.3813, + "step": 2578 + }, + { + "epoch": 0.034827067669172936, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3645, + "step": 2579 + }, + { + "epoch": 0.03488721804511278, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.4484, + "step": 2580 + }, + { + "epoch": 0.03494736842105263, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3596, + "step": 2581 + }, + { + "epoch": 0.03500751879699248, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3846, + "step": 2582 + }, + { + "epoch": 0.03506766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2383, + "step": 2583 + }, + { + "epoch": 0.03512781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3719, + "step": 2584 + }, + { + "epoch": 0.03518796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3676, + "step": 2585 + }, + { + "epoch": 0.03524812030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.331, + "step": 2586 + }, + { + "epoch": 0.035308270676691726, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3885, + "step": 2587 + }, + { + "epoch": 0.03536842105263158, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.353, + "step": 2588 + }, + { + "epoch": 0.03542857142857143, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3385, + "step": 2589 + }, + { + "epoch": 0.03548872180451128, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4337, + "step": 2590 + }, + { + "epoch": 0.03554887218045113, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3328, + "step": 2591 + }, + { + "epoch": 0.03560902255639098, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3562, + "step": 2592 + }, + { + "epoch": 0.03566917293233083, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3202, + "step": 2593 + }, + { + "epoch": 0.03572932330827068, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3858, + "step": 2594 + }, + { + "epoch": 0.035789473684210524, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.3414, + "step": 2595 + }, + { + "epoch": 0.035849624060150374, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.4325, + "step": 2596 + }, + { + "epoch": 0.035909774436090225, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.4504, + "step": 2597 + }, + { + "epoch": 0.035969924812030075, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3919, + "step": 2598 + }, + { + "epoch": 0.036030075187969926, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.4377, + "step": 2599 + }, + { + "epoch": 0.03609022556390978, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3934, + "step": 2600 + }, + { + "epoch": 0.03615037593984963, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4811, + "step": 2601 + }, + { + "epoch": 0.03621052631578947, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3678, + "step": 2602 + }, + { + "epoch": 0.03627067669172932, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.3986, + "step": 2603 + }, + { + "epoch": 0.03633082706766917, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.41, + "step": 2604 + }, + { + "epoch": 0.03639097744360902, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3706, + "step": 2605 + }, + { + "epoch": 0.03645112781954887, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3382, + "step": 2606 + }, + { + "epoch": 0.036511278195488724, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.471, + "step": 2607 + }, + { + "epoch": 0.036571428571428574, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4089, + "step": 2608 + }, + { + "epoch": 0.03663157894736842, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3675, + "step": 2609 + }, + { + "epoch": 0.03669172932330827, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.4449, + "step": 2610 + }, + { + "epoch": 0.03675187969924812, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4381, + "step": 2611 + }, + { + "epoch": 0.03681203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.393, + "step": 2612 + }, + { + "epoch": 0.03687218045112782, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4188, + "step": 2613 + }, + { + "epoch": 0.03693233082706767, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3084, + "step": 2614 + }, + { + "epoch": 0.03699248120300752, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.409, + "step": 2615 + }, + { + "epoch": 0.03705263157894737, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4628, + "step": 2616 + }, + { + "epoch": 0.037112781954887215, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4318, + "step": 2617 + }, + { + "epoch": 0.037172932330827066, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4375, + "step": 2618 + }, + { + "epoch": 0.037233082706766917, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.4394, + "step": 2619 + }, + { + "epoch": 0.03729323308270677, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2718, + "step": 2620 + }, + { + "epoch": 0.03735338345864662, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.4716, + "step": 2621 + }, + { + "epoch": 0.03741353383458647, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.3876, + "step": 2622 + }, + { + "epoch": 0.03747368421052632, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3844, + "step": 2623 + }, + { + "epoch": 0.03753383458646616, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3862, + "step": 2624 + }, + { + "epoch": 0.03759398496240601, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.5689, + "step": 2625 + }, + { + "epoch": 0.037654135338345864, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3755, + "step": 2626 + }, + { + "epoch": 0.037714285714285714, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2843, + "step": 2627 + }, + { + "epoch": 0.037774436090225565, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4758, + "step": 2628 + }, + { + "epoch": 0.037834586466165415, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4068, + "step": 2629 + }, + { + "epoch": 0.037894736842105266, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3537, + "step": 2630 + }, + { + "epoch": 0.03795488721804511, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3118, + "step": 2631 + }, + { + "epoch": 0.03801503759398496, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.343, + "step": 2632 + }, + { + "epoch": 0.03807518796992481, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4137, + "step": 2633 + }, + { + "epoch": 0.03813533834586466, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3357, + "step": 2634 + }, + { + "epoch": 0.03819548872180451, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3722, + "step": 2635 + }, + { + "epoch": 0.03825563909774436, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.4304, + "step": 2636 + }, + { + "epoch": 0.03831578947368421, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3824, + "step": 2637 + }, + { + "epoch": 0.03837593984962406, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3995, + "step": 2638 + }, + { + "epoch": 0.03843609022556391, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3701, + "step": 2639 + }, + { + "epoch": 0.03849624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3448, + "step": 2640 + }, + { + "epoch": 0.03855639097744361, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.4392, + "step": 2641 + }, + { + "epoch": 0.03861654135338346, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.417, + "step": 2642 + }, + { + "epoch": 0.03867669172932331, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.366, + "step": 2643 + }, + { + "epoch": 0.03873684210526316, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.4041, + "step": 2644 + }, + { + "epoch": 0.03879699248120301, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3222, + "step": 2645 + }, + { + "epoch": 0.038857142857142854, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3305, + "step": 2646 + }, + { + "epoch": 0.038917293233082705, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3146, + "step": 2647 + }, + { + "epoch": 0.038977443609022555, + "grad_norm": 0.39453125, + "learning_rate": 3e-05, + "loss": 1.3817, + "step": 2648 + }, + { + "epoch": 0.039037593984962406, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3468, + "step": 2649 + }, + { + "epoch": 0.039097744360902256, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3608, + "step": 2650 + }, + { + "epoch": 0.03915789473684211, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4665, + "step": 2651 + }, + { + "epoch": 0.03921804511278196, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.409, + "step": 2652 + }, + { + "epoch": 0.03927819548872181, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3761, + "step": 2653 + }, + { + "epoch": 0.03933834586466165, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.4002, + "step": 2654 + }, + { + "epoch": 0.0393984962406015, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.4406, + "step": 2655 + }, + { + "epoch": 0.03945864661654135, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3934, + "step": 2656 + }, + { + "epoch": 0.0395187969924812, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3716, + "step": 2657 + }, + { + "epoch": 0.039578947368421054, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3589, + "step": 2658 + }, + { + "epoch": 0.039639097744360904, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.4565, + "step": 2659 + }, + { + "epoch": 0.039699248120300755, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3682, + "step": 2660 + }, + { + "epoch": 0.0397593984962406, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4377, + "step": 2661 + }, + { + "epoch": 0.03981954887218045, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.337, + "step": 2662 + }, + { + "epoch": 0.0398796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4482, + "step": 2663 + }, + { + "epoch": 0.03993984962406015, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4024, + "step": 2664 + }, + { + "epoch": 0.04, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4502, + "step": 2665 + }, + { + "epoch": 0.04006015037593985, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.5103, + "step": 2666 + }, + { + "epoch": 0.0401203007518797, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3455, + "step": 2667 + }, + { + "epoch": 0.040180451127819546, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4011, + "step": 2668 + }, + { + "epoch": 0.040240601503759396, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3602, + "step": 2669 + }, + { + "epoch": 0.04030075187969925, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3464, + "step": 2670 + }, + { + "epoch": 0.0403609022556391, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.4333, + "step": 2671 + }, + { + "epoch": 0.04042105263157895, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3386, + "step": 2672 + }, + { + "epoch": 0.0404812030075188, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.4867, + "step": 2673 + }, + { + "epoch": 0.04054135338345865, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.436, + "step": 2674 + }, + { + "epoch": 0.0406015037593985, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.3981, + "step": 2675 + }, + { + "epoch": 0.04066165413533834, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4565, + "step": 2676 + }, + { + "epoch": 0.040721804511278194, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3414, + "step": 2677 + }, + { + "epoch": 0.040781954887218044, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.4053, + "step": 2678 + }, + { + "epoch": 0.040842105263157895, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3339, + "step": 2679 + }, + { + "epoch": 0.040902255639097745, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.432, + "step": 2680 + }, + { + "epoch": 0.040962406015037596, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3083, + "step": 2681 + }, + { + "epoch": 0.04102255639097745, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4143, + "step": 2682 + }, + { + "epoch": 0.04108270676691729, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3815, + "step": 2683 + }, + { + "epoch": 0.04114285714285714, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3482, + "step": 2684 + }, + { + "epoch": 0.04120300751879699, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4085, + "step": 2685 + }, + { + "epoch": 0.04126315789473684, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3745, + "step": 2686 + }, + { + "epoch": 0.04132330827067669, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4622, + "step": 2687 + }, + { + "epoch": 0.04138345864661654, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3984, + "step": 2688 + }, + { + "epoch": 0.041443609022556394, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3386, + "step": 2689 + }, + { + "epoch": 0.04150375939849624, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2849, + "step": 2690 + }, + { + "epoch": 0.04156390977443609, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3986, + "step": 2691 + }, + { + "epoch": 0.04162406015037594, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3111, + "step": 2692 + }, + { + "epoch": 0.04168421052631579, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4263, + "step": 2693 + }, + { + "epoch": 0.04174436090225564, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3843, + "step": 2694 + }, + { + "epoch": 0.04180451127819549, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.5083, + "step": 2695 + }, + { + "epoch": 0.04186466165413534, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.4128, + "step": 2696 + }, + { + "epoch": 0.04192481203007519, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3865, + "step": 2697 + }, + { + "epoch": 0.041984962406015035, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4329, + "step": 2698 + }, + { + "epoch": 0.042045112781954885, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.459, + "step": 2699 + }, + { + "epoch": 0.042105263157894736, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3809, + "step": 2700 + }, + { + "epoch": 0.042165413533834586, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.4148, + "step": 2701 + }, + { + "epoch": 0.04222556390977444, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3498, + "step": 2702 + }, + { + "epoch": 0.04228571428571429, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3469, + "step": 2703 + }, + { + "epoch": 0.04234586466165414, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.353, + "step": 2704 + }, + { + "epoch": 0.04240601503759398, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4456, + "step": 2705 + }, + { + "epoch": 0.04246616541353383, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4085, + "step": 2706 + }, + { + "epoch": 0.04252631578947368, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3368, + "step": 2707 + }, + { + "epoch": 0.04258646616541353, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.36, + "step": 2708 + }, + { + "epoch": 0.042646616541353384, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4276, + "step": 2709 + }, + { + "epoch": 0.042706766917293235, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.4052, + "step": 2710 + }, + { + "epoch": 0.042766917293233085, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3852, + "step": 2711 + }, + { + "epoch": 0.042827067669172936, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.4207, + "step": 2712 + }, + { + "epoch": 0.04288721804511278, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3663, + "step": 2713 + }, + { + "epoch": 0.04294736842105263, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.4079, + "step": 2714 + }, + { + "epoch": 0.04300751879699248, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4303, + "step": 2715 + }, + { + "epoch": 0.04306766917293233, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.4537, + "step": 2716 + }, + { + "epoch": 0.04312781954887218, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4129, + "step": 2717 + }, + { + "epoch": 0.04318796992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3497, + "step": 2718 + }, + { + "epoch": 0.04324812030075188, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3347, + "step": 2719 + }, + { + "epoch": 0.043308270676691726, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3477, + "step": 2720 + }, + { + "epoch": 0.04336842105263158, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.349, + "step": 2721 + }, + { + "epoch": 0.04342857142857143, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3698, + "step": 2722 + }, + { + "epoch": 0.04348872180451128, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4348, + "step": 2723 + }, + { + "epoch": 0.04354887218045113, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.382, + "step": 2724 + }, + { + "epoch": 0.04360902255639098, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2822, + "step": 2725 + }, + { + "epoch": 0.04366917293233083, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.4346, + "step": 2726 + }, + { + "epoch": 0.04372932330827067, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3363, + "step": 2727 + }, + { + "epoch": 0.043789473684210524, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3259, + "step": 2728 + }, + { + "epoch": 0.043849624060150375, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4175, + "step": 2729 + }, + { + "epoch": 0.043909774436090225, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.396, + "step": 2730 + }, + { + "epoch": 0.043969924812030076, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4865, + "step": 2731 + }, + { + "epoch": 0.044030075187969926, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.4429, + "step": 2732 + }, + { + "epoch": 0.04409022556390978, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3759, + "step": 2733 + }, + { + "epoch": 0.04415037593984963, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.5197, + "step": 2734 + }, + { + "epoch": 0.04421052631578947, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2223, + "step": 2735 + }, + { + "epoch": 0.04427067669172932, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.4021, + "step": 2736 + }, + { + "epoch": 0.04433082706766917, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4625, + "step": 2737 + }, + { + "epoch": 0.04439097744360902, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.38, + "step": 2738 + }, + { + "epoch": 0.04445112781954887, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.4238, + "step": 2739 + }, + { + "epoch": 0.044511278195488724, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4777, + "step": 2740 + }, + { + "epoch": 0.044571428571428574, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3889, + "step": 2741 + }, + { + "epoch": 0.04463157894736842, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3397, + "step": 2742 + }, + { + "epoch": 0.04469172932330827, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3958, + "step": 2743 + }, + { + "epoch": 0.04475187969924812, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2707, + "step": 2744 + }, + { + "epoch": 0.04481203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3057, + "step": 2745 + }, + { + "epoch": 0.04487218045112782, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2476, + "step": 2746 + }, + { + "epoch": 0.04493233082706767, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4545, + "step": 2747 + }, + { + "epoch": 0.04499248120300752, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.3806, + "step": 2748 + }, + { + "epoch": 0.045052631578947365, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3871, + "step": 2749 + }, + { + "epoch": 0.045112781954887216, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.3236, + "step": 2750 + }, + { + "epoch": 0.045172932330827066, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2544, + "step": 2751 + }, + { + "epoch": 0.04523308270676692, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3753, + "step": 2752 + }, + { + "epoch": 0.04529323308270677, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.324, + "step": 2753 + }, + { + "epoch": 0.04535338345864662, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.6025, + "step": 2754 + }, + { + "epoch": 0.04541353383458647, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4202, + "step": 2755 + }, + { + "epoch": 0.04547368421052632, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3086, + "step": 2756 + }, + { + "epoch": 0.04553383458646616, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.4122, + "step": 2757 + }, + { + "epoch": 0.04559398496240601, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3904, + "step": 2758 + }, + { + "epoch": 0.045654135338345864, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3776, + "step": 2759 + }, + { + "epoch": 0.045714285714285714, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3395, + "step": 2760 + }, + { + "epoch": 0.045774436090225565, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.4424, + "step": 2761 + }, + { + "epoch": 0.045834586466165415, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3726, + "step": 2762 + }, + { + "epoch": 0.045894736842105266, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.331, + "step": 2763 + }, + { + "epoch": 0.04595488721804511, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3656, + "step": 2764 + }, + { + "epoch": 0.04601503759398496, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3731, + "step": 2765 + }, + { + "epoch": 0.04607518796992481, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.382, + "step": 2766 + }, + { + "epoch": 0.04613533834586466, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3918, + "step": 2767 + }, + { + "epoch": 0.04619548872180451, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4392, + "step": 2768 + }, + { + "epoch": 0.04625563909774436, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4758, + "step": 2769 + }, + { + "epoch": 0.04631578947368421, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.3974, + "step": 2770 + }, + { + "epoch": 0.046375939849624064, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3962, + "step": 2771 + }, + { + "epoch": 0.04643609022556391, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3226, + "step": 2772 + }, + { + "epoch": 0.04649624060150376, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3715, + "step": 2773 + }, + { + "epoch": 0.04655639097744361, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3641, + "step": 2774 + }, + { + "epoch": 0.04661654135338346, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3886, + "step": 2775 + }, + { + "epoch": 0.04667669172932331, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.406, + "step": 2776 + }, + { + "epoch": 0.04673684210526316, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3919, + "step": 2777 + }, + { + "epoch": 0.04679699248120301, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3968, + "step": 2778 + }, + { + "epoch": 0.046857142857142854, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3838, + "step": 2779 + }, + { + "epoch": 0.046917293233082705, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3926, + "step": 2780 + }, + { + "epoch": 0.046977443609022555, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4421, + "step": 2781 + }, + { + "epoch": 0.047037593984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4282, + "step": 2782 + }, + { + "epoch": 0.047097744360902256, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3128, + "step": 2783 + }, + { + "epoch": 0.04715789473684211, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.3892, + "step": 2784 + }, + { + "epoch": 0.04721804511278196, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3113, + "step": 2785 + }, + { + "epoch": 0.0472781954887218, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4827, + "step": 2786 + }, + { + "epoch": 0.04733834586466165, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3026, + "step": 2787 + }, + { + "epoch": 0.0473984962406015, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.4811, + "step": 2788 + }, + { + "epoch": 0.04745864661654135, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3827, + "step": 2789 + }, + { + "epoch": 0.0475187969924812, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4259, + "step": 2790 + }, + { + "epoch": 0.047578947368421054, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3905, + "step": 2791 + }, + { + "epoch": 0.047639097744360905, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.488, + "step": 2792 + }, + { + "epoch": 0.047699248120300755, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.3328, + "step": 2793 + }, + { + "epoch": 0.0477593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3062, + "step": 2794 + }, + { + "epoch": 0.04781954887218045, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3578, + "step": 2795 + }, + { + "epoch": 0.0478796992481203, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.3634, + "step": 2796 + }, + { + "epoch": 0.04793984962406015, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3839, + "step": 2797 + }, + { + "epoch": 0.048, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.319, + "step": 2798 + }, + { + "epoch": 0.04806015037593985, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3749, + "step": 2799 + }, + { + "epoch": 0.0481203007518797, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.3448, + "step": 2800 + }, + { + "epoch": 0.048180451127819546, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4303, + "step": 2801 + }, + { + "epoch": 0.048240601503759396, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.391, + "step": 2802 + }, + { + "epoch": 0.04830075187969925, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3089, + "step": 2803 + }, + { + "epoch": 0.0483609022556391, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.344, + "step": 2804 + }, + { + "epoch": 0.04842105263157895, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4008, + "step": 2805 + }, + { + "epoch": 0.0484812030075188, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2719, + "step": 2806 + }, + { + "epoch": 0.04854135338345865, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4444, + "step": 2807 + }, + { + "epoch": 0.0486015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.4072, + "step": 2808 + }, + { + "epoch": 0.04866165413533834, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4062, + "step": 2809 + }, + { + "epoch": 0.048721804511278194, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.3944, + "step": 2810 + }, + { + "epoch": 0.048781954887218044, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3535, + "step": 2811 + }, + { + "epoch": 0.048842105263157895, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2963, + "step": 2812 + }, + { + "epoch": 0.048902255639097746, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4621, + "step": 2813 + }, + { + "epoch": 0.048962406015037596, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3976, + "step": 2814 + }, + { + "epoch": 0.04902255639097745, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3788, + "step": 2815 + }, + { + "epoch": 0.04908270676691729, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3538, + "step": 2816 + }, + { + "epoch": 0.04914285714285714, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3088, + "step": 2817 + }, + { + "epoch": 0.04920300751879699, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.4569, + "step": 2818 + }, + { + "epoch": 0.04926315789473684, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.4337, + "step": 2819 + }, + { + "epoch": 0.04932330827067669, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3795, + "step": 2820 + }, + { + "epoch": 0.04938345864661654, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.3256, + "step": 2821 + }, + { + "epoch": 0.049443609022556394, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 2822 + }, + { + "epoch": 0.04950375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2714, + "step": 2823 + }, + { + "epoch": 0.04956390977443609, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.3771, + "step": 2824 + }, + { + "epoch": 0.04962406015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3918, + "step": 2825 + }, + { + "epoch": 0.04968421052631579, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3289, + "step": 2826 + }, + { + "epoch": 0.04974436090225564, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.431, + "step": 2827 + }, + { + "epoch": 0.04980451127819549, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3903, + "step": 2828 + }, + { + "epoch": 0.04986466165413534, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3576, + "step": 2829 + }, + { + "epoch": 0.04992481203007519, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3893, + "step": 2830 + }, + { + "epoch": 0.049984962406015035, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3997, + "step": 2831 + }, + { + "epoch": 0.050045112781954885, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3836, + "step": 2832 + }, + { + "epoch": 0.050105263157894736, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3746, + "step": 2833 + }, + { + "epoch": 0.05016541353383459, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4394, + "step": 2834 + }, + { + "epoch": 0.05022556390977444, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3902, + "step": 2835 + }, + { + "epoch": 0.05028571428571429, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.4303, + "step": 2836 + }, + { + "epoch": 0.05034586466165414, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3075, + "step": 2837 + }, + { + "epoch": 0.05040601503759398, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.396, + "step": 2838 + }, + { + "epoch": 0.05046616541353383, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3355, + "step": 2839 + }, + { + "epoch": 0.05052631578947368, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4591, + "step": 2840 + }, + { + "epoch": 0.050586466165413534, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3457, + "step": 2841 + }, + { + "epoch": 0.050646616541353384, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3421, + "step": 2842 + }, + { + "epoch": 0.050706766917293235, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4267, + "step": 2843 + }, + { + "epoch": 0.050766917293233085, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4485, + "step": 2844 + }, + { + "epoch": 0.05082706766917293, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3502, + "step": 2845 + }, + { + "epoch": 0.05088721804511278, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4176, + "step": 2846 + }, + { + "epoch": 0.05094736842105263, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3092, + "step": 2847 + }, + { + "epoch": 0.05100751879699248, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4221, + "step": 2848 + }, + { + "epoch": 0.05106766917293233, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3873, + "step": 2849 + }, + { + "epoch": 0.05112781954887218, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.37, + "step": 2850 + }, + { + "epoch": 0.05118796992481203, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4211, + "step": 2851 + }, + { + "epoch": 0.05124812030075188, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3974, + "step": 2852 + }, + { + "epoch": 0.05130827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.4375, + "step": 2853 + }, + { + "epoch": 0.05136842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.4716, + "step": 2854 + }, + { + "epoch": 0.05142857142857143, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3901, + "step": 2855 + }, + { + "epoch": 0.05148872180451128, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3427, + "step": 2856 + }, + { + "epoch": 0.05154887218045113, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2554, + "step": 2857 + }, + { + "epoch": 0.05160902255639098, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3737, + "step": 2858 + }, + { + "epoch": 0.05166917293233083, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.4154, + "step": 2859 + }, + { + "epoch": 0.051729323308270674, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.4498, + "step": 2860 + }, + { + "epoch": 0.051789473684210524, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3885, + "step": 2861 + }, + { + "epoch": 0.051849624060150375, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.356, + "step": 2862 + }, + { + "epoch": 0.051909774436090225, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3811, + "step": 2863 + }, + { + "epoch": 0.051969924812030076, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3695, + "step": 2864 + }, + { + "epoch": 0.052030075187969926, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4199, + "step": 2865 + }, + { + "epoch": 0.05209022556390978, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3337, + "step": 2866 + }, + { + "epoch": 0.05215037593984963, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3572, + "step": 2867 + }, + { + "epoch": 0.05221052631578947, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.344, + "step": 2868 + }, + { + "epoch": 0.05227067669172932, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3371, + "step": 2869 + }, + { + "epoch": 0.05233082706766917, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3718, + "step": 2870 + }, + { + "epoch": 0.05239097744360902, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4133, + "step": 2871 + }, + { + "epoch": 0.05245112781954887, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3688, + "step": 2872 + }, + { + "epoch": 0.052511278195488724, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3573, + "step": 2873 + }, + { + "epoch": 0.052571428571428575, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3976, + "step": 2874 + }, + { + "epoch": 0.05263157894736842, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.3241, + "step": 2875 + }, + { + "epoch": 0.05269172932330827, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.5645, + "step": 2876 + }, + { + "epoch": 0.05275187969924812, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3465, + "step": 2877 + }, + { + "epoch": 0.05281203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3583, + "step": 2878 + }, + { + "epoch": 0.05287218045112782, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3531, + "step": 2879 + }, + { + "epoch": 0.05293233082706767, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4231, + "step": 2880 + }, + { + "epoch": 0.05299248120300752, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4864, + "step": 2881 + }, + { + "epoch": 0.053052631578947365, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3352, + "step": 2882 + }, + { + "epoch": 0.053112781954887216, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3528, + "step": 2883 + }, + { + "epoch": 0.053172932330827066, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.4293, + "step": 2884 + }, + { + "epoch": 0.05323308270676692, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.3932, + "step": 2885 + }, + { + "epoch": 0.05329323308270677, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4177, + "step": 2886 + }, + { + "epoch": 0.05335338345864662, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3902, + "step": 2887 + }, + { + "epoch": 0.05341353383458647, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.5134, + "step": 2888 + }, + { + "epoch": 0.05347368421052632, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.385, + "step": 2889 + }, + { + "epoch": 0.05353383458646616, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3645, + "step": 2890 + }, + { + "epoch": 0.05359398496240601, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3919, + "step": 2891 + }, + { + "epoch": 0.053654135338345864, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4745, + "step": 2892 + }, + { + "epoch": 0.053714285714285714, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.4313, + "step": 2893 + }, + { + "epoch": 0.053774436090225565, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3397, + "step": 2894 + }, + { + "epoch": 0.053834586466165416, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4607, + "step": 2895 + }, + { + "epoch": 0.053894736842105266, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3602, + "step": 2896 + }, + { + "epoch": 0.05395488721804511, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.332, + "step": 2897 + }, + { + "epoch": 0.05401503759398496, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3183, + "step": 2898 + }, + { + "epoch": 0.05407518796992481, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3673, + "step": 2899 + }, + { + "epoch": 0.05413533834586466, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4283, + "step": 2900 + }, + { + "epoch": 0.05419548872180451, + "grad_norm": 0.41796875, + "learning_rate": 3e-05, + "loss": 1.375, + "step": 2901 + }, + { + "epoch": 0.05425563909774436, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3563, + "step": 2902 + }, + { + "epoch": 0.05431578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3004, + "step": 2903 + }, + { + "epoch": 0.05437593984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3867, + "step": 2904 + }, + { + "epoch": 0.05443609022556391, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3771, + "step": 2905 + }, + { + "epoch": 0.05449624060150376, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3888, + "step": 2906 + }, + { + "epoch": 0.05455639097744361, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2814, + "step": 2907 + }, + { + "epoch": 0.05461654135338346, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3787, + "step": 2908 + }, + { + "epoch": 0.05467669172932331, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3573, + "step": 2909 + }, + { + "epoch": 0.05473684210526316, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.4312, + "step": 2910 + }, + { + "epoch": 0.05479699248120301, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3759, + "step": 2911 + }, + { + "epoch": 0.054857142857142854, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3429, + "step": 2912 + }, + { + "epoch": 0.054917293233082705, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4048, + "step": 2913 + }, + { + "epoch": 0.054977443609022555, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3148, + "step": 2914 + }, + { + "epoch": 0.055037593984962406, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3498, + "step": 2915 + }, + { + "epoch": 0.05509774436090226, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2852, + "step": 2916 + }, + { + "epoch": 0.05515789473684211, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2704, + "step": 2917 + }, + { + "epoch": 0.05521804511278196, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.299, + "step": 2918 + }, + { + "epoch": 0.0552781954887218, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.386, + "step": 2919 + }, + { + "epoch": 0.05533834586466165, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3562, + "step": 2920 + }, + { + "epoch": 0.0553984962406015, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3072, + "step": 2921 + }, + { + "epoch": 0.05545864661654135, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.404, + "step": 2922 + }, + { + "epoch": 0.055518796992481204, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4323, + "step": 2923 + }, + { + "epoch": 0.055578947368421054, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.4432, + "step": 2924 + }, + { + "epoch": 0.055639097744360905, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3259, + "step": 2925 + }, + { + "epoch": 0.055699248120300755, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.4015, + "step": 2926 + }, + { + "epoch": 0.0557593984962406, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3495, + "step": 2927 + }, + { + "epoch": 0.05581954887218045, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.373, + "step": 2928 + }, + { + "epoch": 0.0558796992481203, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4039, + "step": 2929 + }, + { + "epoch": 0.05593984962406015, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.4924, + "step": 2930 + }, + { + "epoch": 0.056, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3936, + "step": 2931 + }, + { + "epoch": 0.05606015037593985, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3359, + "step": 2932 + }, + { + "epoch": 0.0561203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3912, + "step": 2933 + }, + { + "epoch": 0.056180451127819546, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3422, + "step": 2934 + }, + { + "epoch": 0.056240601503759396, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3071, + "step": 2935 + }, + { + "epoch": 0.05630075187969925, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3722, + "step": 2936 + }, + { + "epoch": 0.0563609022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.4032, + "step": 2937 + }, + { + "epoch": 0.05642105263157895, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.4068, + "step": 2938 + }, + { + "epoch": 0.0564812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3784, + "step": 2939 + }, + { + "epoch": 0.05654135338345865, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.433, + "step": 2940 + }, + { + "epoch": 0.05660150375939849, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.2826, + "step": 2941 + }, + { + "epoch": 0.056661654135338343, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3554, + "step": 2942 + }, + { + "epoch": 0.056721804511278194, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3515, + "step": 2943 + }, + { + "epoch": 0.056781954887218045, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3959, + "step": 2944 + }, + { + "epoch": 0.056842105263157895, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.4009, + "step": 2945 + }, + { + "epoch": 0.056902255639097746, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3367, + "step": 2946 + }, + { + "epoch": 0.056962406015037596, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.5321, + "step": 2947 + }, + { + "epoch": 0.05702255639097745, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3266, + "step": 2948 + }, + { + "epoch": 0.05708270676691729, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.4827, + "step": 2949 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3931, + "step": 2950 + }, + { + "epoch": 0.05720300751879699, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3931, + "step": 2951 + }, + { + "epoch": 0.05726315789473684, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3494, + "step": 2952 + }, + { + "epoch": 0.05732330827067669, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4111, + "step": 2953 + }, + { + "epoch": 0.05738345864661654, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3249, + "step": 2954 + }, + { + "epoch": 0.057443609022556394, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3577, + "step": 2955 + }, + { + "epoch": 0.05750375939849624, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3078, + "step": 2956 + }, + { + "epoch": 0.05756390977443609, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3608, + "step": 2957 + }, + { + "epoch": 0.05762406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3529, + "step": 2958 + }, + { + "epoch": 0.05768421052631579, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2858, + "step": 2959 + }, + { + "epoch": 0.05774436090225564, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3901, + "step": 2960 + }, + { + "epoch": 0.05780451127819549, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3409, + "step": 2961 + }, + { + "epoch": 0.05786466165413534, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.4464, + "step": 2962 + }, + { + "epoch": 0.057924812030075185, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.394, + "step": 2963 + }, + { + "epoch": 0.057984962406015035, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3218, + "step": 2964 + }, + { + "epoch": 0.058045112781954886, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3237, + "step": 2965 + }, + { + "epoch": 0.058105263157894736, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4259, + "step": 2966 + }, + { + "epoch": 0.05816541353383459, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3342, + "step": 2967 + }, + { + "epoch": 0.05822556390977444, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3855, + "step": 2968 + }, + { + "epoch": 0.05828571428571429, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4319, + "step": 2969 + }, + { + "epoch": 0.05834586466165414, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3601, + "step": 2970 + }, + { + "epoch": 0.05840601503759398, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2682, + "step": 2971 + }, + { + "epoch": 0.05846616541353383, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.3141, + "step": 2972 + }, + { + "epoch": 0.05852631578947368, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.3581, + "step": 2973 + }, + { + "epoch": 0.058586466165413534, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2966, + "step": 2974 + }, + { + "epoch": 0.058646616541353384, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.3055, + "step": 2975 + }, + { + "epoch": 0.058706766917293235, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 2976 + }, + { + "epoch": 0.058766917293233085, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3195, + "step": 2977 + }, + { + "epoch": 0.05882706766917293, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3565, + "step": 2978 + }, + { + "epoch": 0.05888721804511278, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3143, + "step": 2979 + }, + { + "epoch": 0.05894736842105263, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3895, + "step": 2980 + }, + { + "epoch": 0.05900751879699248, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4425, + "step": 2981 + }, + { + "epoch": 0.05906766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3461, + "step": 2982 + }, + { + "epoch": 0.05912781954887218, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3636, + "step": 2983 + }, + { + "epoch": 0.05918796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.334, + "step": 2984 + }, + { + "epoch": 0.05924812030075188, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3176, + "step": 2985 + }, + { + "epoch": 0.05930827067669173, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.361, + "step": 2986 + }, + { + "epoch": 0.05936842105263158, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3984, + "step": 2987 + }, + { + "epoch": 0.05942857142857143, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4132, + "step": 2988 + }, + { + "epoch": 0.05948872180451128, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3079, + "step": 2989 + }, + { + "epoch": 0.05954887218045113, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3202, + "step": 2990 + }, + { + "epoch": 0.05960902255639098, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3608, + "step": 2991 + }, + { + "epoch": 0.05966917293233083, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3277, + "step": 2992 + }, + { + "epoch": 0.059729323308270674, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3694, + "step": 2993 + }, + { + "epoch": 0.059789473684210524, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3384, + "step": 2994 + }, + { + "epoch": 0.059849624060150375, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3996, + "step": 2995 + }, + { + "epoch": 0.059909774436090225, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.337, + "step": 2996 + }, + { + "epoch": 0.059969924812030076, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3393, + "step": 2997 + }, + { + "epoch": 0.06003007518796993, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3138, + "step": 2998 + }, + { + "epoch": 0.06009022556390978, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3692, + "step": 2999 + }, + { + "epoch": 0.06015037593984962, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3772, + "step": 3000 + }, + { + "epoch": 0.06021052631578947, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.3594, + "step": 3001 + }, + { + "epoch": 0.06027067669172932, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2871, + "step": 3002 + }, + { + "epoch": 0.06033082706766917, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.3013, + "step": 3003 + }, + { + "epoch": 0.06039097744360902, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4027, + "step": 3004 + }, + { + "epoch": 0.060451127819548874, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3516, + "step": 3005 + }, + { + "epoch": 0.060511278195488724, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3051, + "step": 3006 + }, + { + "epoch": 0.060571428571428575, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.4098, + "step": 3007 + }, + { + "epoch": 0.06063157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2606, + "step": 3008 + }, + { + "epoch": 0.06069172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3578, + "step": 3009 + }, + { + "epoch": 0.06075187969924812, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3128, + "step": 3010 + }, + { + "epoch": 0.06081203007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3105, + "step": 3011 + }, + { + "epoch": 0.06087218045112782, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3437, + "step": 3012 + }, + { + "epoch": 0.06093233082706767, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3751, + "step": 3013 + }, + { + "epoch": 0.06099248120300752, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3667, + "step": 3014 + }, + { + "epoch": 0.061052631578947365, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.3498, + "step": 3015 + }, + { + "epoch": 0.061112781954887216, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.3362, + "step": 3016 + }, + { + "epoch": 0.061172932330827066, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4111, + "step": 3017 + }, + { + "epoch": 0.06123308270676692, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.4029, + "step": 3018 + }, + { + "epoch": 0.06129323308270677, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.4001, + "step": 3019 + }, + { + "epoch": 0.06135338345864662, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.403, + "step": 3020 + }, + { + "epoch": 0.06141353383458647, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.2934, + "step": 3021 + }, + { + "epoch": 0.06147368421052631, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.323, + "step": 3022 + }, + { + "epoch": 0.06153383458646616, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3787, + "step": 3023 + }, + { + "epoch": 0.06159398496240601, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2972, + "step": 3024 + }, + { + "epoch": 0.061654135338345864, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2467, + "step": 3025 + }, + { + "epoch": 0.061714285714285715, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4026, + "step": 3026 + }, + { + "epoch": 0.061774436090225565, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.4205, + "step": 3027 + }, + { + "epoch": 0.061834586466165416, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.2872, + "step": 3028 + }, + { + "epoch": 0.061894736842105266, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4161, + "step": 3029 + }, + { + "epoch": 0.06195488721804511, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3287, + "step": 3030 + }, + { + "epoch": 0.06201503759398496, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3716, + "step": 3031 + }, + { + "epoch": 0.06207518796992481, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3505, + "step": 3032 + }, + { + "epoch": 0.06213533834586466, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3545, + "step": 3033 + }, + { + "epoch": 0.06219548872180451, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.4091, + "step": 3034 + }, + { + "epoch": 0.06225563909774436, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3638, + "step": 3035 + }, + { + "epoch": 0.06231578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3926, + "step": 3036 + }, + { + "epoch": 0.06237593984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3517, + "step": 3037 + }, + { + "epoch": 0.06243609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3221, + "step": 3038 + }, + { + "epoch": 0.06249624060150376, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4537, + "step": 3039 + }, + { + "epoch": 0.06255639097744362, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3456, + "step": 3040 + }, + { + "epoch": 0.06261654135338346, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.3388, + "step": 3041 + }, + { + "epoch": 0.0626766917293233, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.3921, + "step": 3042 + }, + { + "epoch": 0.06273684210526316, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3192, + "step": 3043 + }, + { + "epoch": 0.062796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.297, + "step": 3044 + }, + { + "epoch": 0.06285714285714286, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3706, + "step": 3045 + }, + { + "epoch": 0.0629172932330827, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3562, + "step": 3046 + }, + { + "epoch": 0.06297744360902256, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2902, + "step": 3047 + }, + { + "epoch": 0.0630375939849624, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4389, + "step": 3048 + }, + { + "epoch": 0.06309774436090225, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3549, + "step": 3049 + }, + { + "epoch": 0.06315789473684211, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.3288, + "step": 3050 + }, + { + "epoch": 0.06321804511278195, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4132, + "step": 3051 + }, + { + "epoch": 0.06327819548872181, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.4433, + "step": 3052 + }, + { + "epoch": 0.06333834586466165, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3997, + "step": 3053 + }, + { + "epoch": 0.06339849624060151, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.35, + "step": 3054 + }, + { + "epoch": 0.06345864661654135, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3832, + "step": 3055 + }, + { + "epoch": 0.0635187969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3952, + "step": 3056 + }, + { + "epoch": 0.06357894736842105, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4111, + "step": 3057 + }, + { + "epoch": 0.0636390977443609, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3576, + "step": 3058 + }, + { + "epoch": 0.06369924812030076, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3216, + "step": 3059 + }, + { + "epoch": 0.0637593984962406, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3396, + "step": 3060 + }, + { + "epoch": 0.06381954887218046, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3561, + "step": 3061 + }, + { + "epoch": 0.0638796992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.367, + "step": 3062 + }, + { + "epoch": 0.06393984962406014, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4302, + "step": 3063 + }, + { + "epoch": 0.064, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.336, + "step": 3064 + }, + { + "epoch": 0.06406015037593984, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3658, + "step": 3065 + }, + { + "epoch": 0.0641203007518797, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.3253, + "step": 3066 + }, + { + "epoch": 0.06418045112781955, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3747, + "step": 3067 + }, + { + "epoch": 0.0642406015037594, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4729, + "step": 3068 + }, + { + "epoch": 0.06430075187969925, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3173, + "step": 3069 + }, + { + "epoch": 0.06436090225563909, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3487, + "step": 3070 + }, + { + "epoch": 0.06442105263157895, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3973, + "step": 3071 + }, + { + "epoch": 0.06448120300751879, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.2741, + "step": 3072 + }, + { + "epoch": 0.06454135338345865, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3201, + "step": 3073 + }, + { + "epoch": 0.0646015037593985, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3207, + "step": 3074 + }, + { + "epoch": 0.06466165413533835, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4863, + "step": 3075 + }, + { + "epoch": 0.0647218045112782, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2862, + "step": 3076 + }, + { + "epoch": 0.06478195488721805, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2284, + "step": 3077 + }, + { + "epoch": 0.0648421052631579, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3576, + "step": 3078 + }, + { + "epoch": 0.06490225563909774, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3188, + "step": 3079 + }, + { + "epoch": 0.0649624060150376, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3523, + "step": 3080 + }, + { + "epoch": 0.06502255639097744, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.266, + "step": 3081 + }, + { + "epoch": 0.0650827067669173, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3479, + "step": 3082 + }, + { + "epoch": 0.06514285714285714, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3501, + "step": 3083 + }, + { + "epoch": 0.065203007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.377, + "step": 3084 + }, + { + "epoch": 0.06526315789473684, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3716, + "step": 3085 + }, + { + "epoch": 0.06532330827067669, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3129, + "step": 3086 + }, + { + "epoch": 0.06538345864661654, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.4379, + "step": 3087 + }, + { + "epoch": 0.06544360902255639, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3625, + "step": 3088 + }, + { + "epoch": 0.06550375939849624, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.309, + "step": 3089 + }, + { + "epoch": 0.06556390977443609, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3261, + "step": 3090 + }, + { + "epoch": 0.06562406015037595, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.4357, + "step": 3091 + }, + { + "epoch": 0.06568421052631579, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3916, + "step": 3092 + }, + { + "epoch": 0.06574436090225563, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4499, + "step": 3093 + }, + { + "epoch": 0.06580451127819549, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3817, + "step": 3094 + }, + { + "epoch": 0.06586466165413533, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3507, + "step": 3095 + }, + { + "epoch": 0.06592481203007519, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3561, + "step": 3096 + }, + { + "epoch": 0.06598496240601504, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3026, + "step": 3097 + }, + { + "epoch": 0.06604511278195489, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4017, + "step": 3098 + }, + { + "epoch": 0.06610526315789474, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3712, + "step": 3099 + }, + { + "epoch": 0.06616541353383458, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3469, + "step": 3100 + }, + { + "epoch": 0.06622556390977444, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3732, + "step": 3101 + }, + { + "epoch": 0.06628571428571428, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3465, + "step": 3102 + }, + { + "epoch": 0.06634586466165414, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.4875, + "step": 3103 + }, + { + "epoch": 0.06640601503759398, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.3914, + "step": 3104 + }, + { + "epoch": 0.06646616541353384, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4254, + "step": 3105 + }, + { + "epoch": 0.06652631578947368, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3152, + "step": 3106 + }, + { + "epoch": 0.06658646616541353, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3177, + "step": 3107 + }, + { + "epoch": 0.06664661654135338, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3774, + "step": 3108 + }, + { + "epoch": 0.06670676691729323, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.3166, + "step": 3109 + }, + { + "epoch": 0.06676691729323309, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3884, + "step": 3110 + }, + { + "epoch": 0.06682706766917293, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3978, + "step": 3111 + }, + { + "epoch": 0.06688721804511279, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2602, + "step": 3112 + }, + { + "epoch": 0.06694736842105263, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2548, + "step": 3113 + }, + { + "epoch": 0.06700751879699249, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4304, + "step": 3114 + }, + { + "epoch": 0.06706766917293233, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.3081, + "step": 3115 + }, + { + "epoch": 0.06712781954887218, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.4191, + "step": 3116 + }, + { + "epoch": 0.06718796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3405, + "step": 3117 + }, + { + "epoch": 0.06724812030075188, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3562, + "step": 3118 + }, + { + "epoch": 0.06730827067669173, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3742, + "step": 3119 + }, + { + "epoch": 0.06736842105263158, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4296, + "step": 3120 + }, + { + "epoch": 0.06742857142857143, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.393, + "step": 3121 + }, + { + "epoch": 0.06748872180451128, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.3015, + "step": 3122 + }, + { + "epoch": 0.06754887218045112, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.4561, + "step": 3123 + }, + { + "epoch": 0.06760902255639098, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2761, + "step": 3124 + }, + { + "epoch": 0.06766917293233082, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.3544, + "step": 3125 + }, + { + "epoch": 0.06772932330827068, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.4446, + "step": 3126 + }, + { + "epoch": 0.06778947368421052, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3561, + "step": 3127 + }, + { + "epoch": 0.06784962406015038, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3433, + "step": 3128 + }, + { + "epoch": 0.06790977443609023, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2727, + "step": 3129 + }, + { + "epoch": 0.06796992481203007, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3574, + "step": 3130 + }, + { + "epoch": 0.06803007518796993, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.5345, + "step": 3131 + }, + { + "epoch": 0.06809022556390977, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3173, + "step": 3132 + }, + { + "epoch": 0.06815037593984963, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.315, + "step": 3133 + }, + { + "epoch": 0.06821052631578947, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3673, + "step": 3134 + }, + { + "epoch": 0.06827067669172933, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.4414, + "step": 3135 + }, + { + "epoch": 0.06833082706766917, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3544, + "step": 3136 + }, + { + "epoch": 0.06839097744360902, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3829, + "step": 3137 + }, + { + "epoch": 0.06845112781954887, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3008, + "step": 3138 + }, + { + "epoch": 0.06851127819548872, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3906, + "step": 3139 + }, + { + "epoch": 0.06857142857142857, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3872, + "step": 3140 + }, + { + "epoch": 0.06863157894736842, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3294, + "step": 3141 + }, + { + "epoch": 0.06869172932330828, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3305, + "step": 3142 + }, + { + "epoch": 0.06875187969924812, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4112, + "step": 3143 + }, + { + "epoch": 0.06881203007518796, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3747, + "step": 3144 + }, + { + "epoch": 0.06887218045112782, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.386, + "step": 3145 + }, + { + "epoch": 0.06893233082706766, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3252, + "step": 3146 + }, + { + "epoch": 0.06899248120300752, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.3591, + "step": 3147 + }, + { + "epoch": 0.06905263157894737, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3888, + "step": 3148 + }, + { + "epoch": 0.06911278195488722, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2921, + "step": 3149 + }, + { + "epoch": 0.06917293233082707, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3696, + "step": 3150 + }, + { + "epoch": 0.06923308270676692, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3487, + "step": 3151 + }, + { + "epoch": 0.06929323308270677, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.4095, + "step": 3152 + }, + { + "epoch": 0.06935338345864661, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4014, + "step": 3153 + }, + { + "epoch": 0.06941353383458647, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2842, + "step": 3154 + }, + { + "epoch": 0.06947368421052631, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.4365, + "step": 3155 + }, + { + "epoch": 0.06953383458646617, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.3227, + "step": 3156 + }, + { + "epoch": 0.06959398496240601, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3978, + "step": 3157 + }, + { + "epoch": 0.06965413533834587, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3066, + "step": 3158 + }, + { + "epoch": 0.06971428571428571, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.4055, + "step": 3159 + }, + { + "epoch": 0.06977443609022556, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4378, + "step": 3160 + }, + { + "epoch": 0.06983458646616542, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3317, + "step": 3161 + }, + { + "epoch": 0.06989473684210526, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3938, + "step": 3162 + }, + { + "epoch": 0.06995488721804512, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.3748, + "step": 3163 + }, + { + "epoch": 0.07001503759398496, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.3995, + "step": 3164 + }, + { + "epoch": 0.07007518796992482, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.4513, + "step": 3165 + }, + { + "epoch": 0.07013533834586466, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2737, + "step": 3166 + }, + { + "epoch": 0.0701954887218045, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3179, + "step": 3167 + }, + { + "epoch": 0.07025563909774436, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2778, + "step": 3168 + }, + { + "epoch": 0.0703157894736842, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.4467, + "step": 3169 + }, + { + "epoch": 0.07037593984962406, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.3068, + "step": 3170 + }, + { + "epoch": 0.07043609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2477, + "step": 3171 + }, + { + "epoch": 0.07049624060150377, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2594, + "step": 3172 + }, + { + "epoch": 0.07055639097744361, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.271, + "step": 3173 + }, + { + "epoch": 0.07061654135338345, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.4151, + "step": 3174 + }, + { + "epoch": 0.07067669172932331, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3341, + "step": 3175 + }, + { + "epoch": 0.07073684210526315, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3551, + "step": 3176 + }, + { + "epoch": 0.07079699248120301, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3354, + "step": 3177 + }, + { + "epoch": 0.07085714285714285, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3616, + "step": 3178 + }, + { + "epoch": 0.07091729323308271, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.347, + "step": 3179 + }, + { + "epoch": 0.07097744360902256, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 3180 + }, + { + "epoch": 0.0710375939849624, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.4154, + "step": 3181 + }, + { + "epoch": 0.07109774436090226, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3248, + "step": 3182 + }, + { + "epoch": 0.0711578947368421, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3456, + "step": 3183 + }, + { + "epoch": 0.07121804511278196, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2734, + "step": 3184 + }, + { + "epoch": 0.0712781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.304, + "step": 3185 + }, + { + "epoch": 0.07133834586466166, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.3544, + "step": 3186 + }, + { + "epoch": 0.0713984962406015, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.411, + "step": 3187 + }, + { + "epoch": 0.07145864661654136, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3442, + "step": 3188 + }, + { + "epoch": 0.0715187969924812, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3663, + "step": 3189 + }, + { + "epoch": 0.07157894736842105, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.4091, + "step": 3190 + }, + { + "epoch": 0.0716390977443609, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3015, + "step": 3191 + }, + { + "epoch": 0.07169924812030075, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3428, + "step": 3192 + }, + { + "epoch": 0.0717593984962406, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.3465, + "step": 3193 + }, + { + "epoch": 0.07181954887218045, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3199, + "step": 3194 + }, + { + "epoch": 0.07187969924812031, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.402, + "step": 3195 + }, + { + "epoch": 0.07193984962406015, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3906, + "step": 3196 + }, + { + "epoch": 0.072, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 3197 + }, + { + "epoch": 0.07206015037593985, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.4047, + "step": 3198 + }, + { + "epoch": 0.0721203007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3713, + "step": 3199 + }, + { + "epoch": 0.07218045112781955, + "grad_norm": 0.404296875, + "learning_rate": 3e-05, + "loss": 1.4014, + "step": 3200 + }, + { + "epoch": 0.0722406015037594, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2979, + "step": 3201 + }, + { + "epoch": 0.07230075187969925, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.4411, + "step": 3202 + }, + { + "epoch": 0.0723609022556391, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.4197, + "step": 3203 + }, + { + "epoch": 0.07242105263157894, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3032, + "step": 3204 + }, + { + "epoch": 0.0724812030075188, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3237, + "step": 3205 + }, + { + "epoch": 0.07254135338345864, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3676, + "step": 3206 + }, + { + "epoch": 0.0726015037593985, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.341, + "step": 3207 + }, + { + "epoch": 0.07266165413533834, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.4111, + "step": 3208 + }, + { + "epoch": 0.0727218045112782, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.4278, + "step": 3209 + }, + { + "epoch": 0.07278195488721804, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3908, + "step": 3210 + }, + { + "epoch": 0.07284210526315789, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.4382, + "step": 3211 + }, + { + "epoch": 0.07290225563909775, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.399, + "step": 3212 + }, + { + "epoch": 0.07296240601503759, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4229, + "step": 3213 + }, + { + "epoch": 0.07302255639097745, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.346, + "step": 3214 + }, + { + "epoch": 0.07308270676691729, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4177, + "step": 3215 + }, + { + "epoch": 0.07314285714285715, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3302, + "step": 3216 + }, + { + "epoch": 0.07320300751879699, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.309, + "step": 3217 + }, + { + "epoch": 0.07326315789473684, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4127, + "step": 3218 + }, + { + "epoch": 0.0733233082706767, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3661, + "step": 3219 + }, + { + "epoch": 0.07338345864661654, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2693, + "step": 3220 + }, + { + "epoch": 0.0734436090225564, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3237, + "step": 3221 + }, + { + "epoch": 0.07350375939849624, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3801, + "step": 3222 + }, + { + "epoch": 0.0735639097744361, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3623, + "step": 3223 + }, + { + "epoch": 0.07362406015037594, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3616, + "step": 3224 + }, + { + "epoch": 0.07368421052631578, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3184, + "step": 3225 + }, + { + "epoch": 0.07374436090225564, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2897, + "step": 3226 + }, + { + "epoch": 0.07380451127819548, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3456, + "step": 3227 + }, + { + "epoch": 0.07386466165413534, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4307, + "step": 3228 + }, + { + "epoch": 0.07392481203007518, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4248, + "step": 3229 + }, + { + "epoch": 0.07398496240601504, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3224, + "step": 3230 + }, + { + "epoch": 0.07404511278195489, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.3581, + "step": 3231 + }, + { + "epoch": 0.07410526315789474, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3313, + "step": 3232 + }, + { + "epoch": 0.07416541353383459, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.362, + "step": 3233 + }, + { + "epoch": 0.07422556390977443, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3523, + "step": 3234 + }, + { + "epoch": 0.07428571428571429, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3574, + "step": 3235 + }, + { + "epoch": 0.07434586466165413, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3668, + "step": 3236 + }, + { + "epoch": 0.07440601503759399, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3613, + "step": 3237 + }, + { + "epoch": 0.07446616541353383, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3543, + "step": 3238 + }, + { + "epoch": 0.07452631578947369, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 3239 + }, + { + "epoch": 0.07458646616541353, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.3223, + "step": 3240 + }, + { + "epoch": 0.07464661654135338, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3856, + "step": 3241 + }, + { + "epoch": 0.07470676691729324, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3488, + "step": 3242 + }, + { + "epoch": 0.07476691729323308, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2795, + "step": 3243 + }, + { + "epoch": 0.07482706766917294, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3764, + "step": 3244 + }, + { + "epoch": 0.07488721804511278, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3377, + "step": 3245 + }, + { + "epoch": 0.07494736842105264, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3753, + "step": 3246 + }, + { + "epoch": 0.07500751879699248, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3854, + "step": 3247 + }, + { + "epoch": 0.07506766917293232, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3446, + "step": 3248 + }, + { + "epoch": 0.07512781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2886, + "step": 3249 + }, + { + "epoch": 0.07518796992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.4842, + "step": 3250 + }, + { + "epoch": 0.07524812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.4709, + "step": 3251 + }, + { + "epoch": 0.07530827067669173, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3625, + "step": 3252 + }, + { + "epoch": 0.07536842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3797, + "step": 3253 + }, + { + "epoch": 0.07542857142857143, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.4119, + "step": 3254 + }, + { + "epoch": 0.07548872180451127, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.289, + "step": 3255 + }, + { + "epoch": 0.07554887218045113, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3523, + "step": 3256 + }, + { + "epoch": 0.07560902255639097, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3945, + "step": 3257 + }, + { + "epoch": 0.07566917293233083, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3535, + "step": 3258 + }, + { + "epoch": 0.07572932330827067, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.4757, + "step": 3259 + }, + { + "epoch": 0.07578947368421053, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3245, + "step": 3260 + }, + { + "epoch": 0.07584962406015038, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.4756, + "step": 3261 + }, + { + "epoch": 0.07590977443609022, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2894, + "step": 3262 + }, + { + "epoch": 0.07596992481203008, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2479, + "step": 3263 + }, + { + "epoch": 0.07603007518796992, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3469, + "step": 3264 + }, + { + "epoch": 0.07609022556390978, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2862, + "step": 3265 + }, + { + "epoch": 0.07615037593984962, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 3266 + }, + { + "epoch": 0.07621052631578948, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3299, + "step": 3267 + }, + { + "epoch": 0.07627067669172932, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3499, + "step": 3268 + }, + { + "epoch": 0.07633082706766918, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3017, + "step": 3269 + }, + { + "epoch": 0.07639097744360902, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3644, + "step": 3270 + }, + { + "epoch": 0.07645112781954887, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3606, + "step": 3271 + }, + { + "epoch": 0.07651127819548872, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.458, + "step": 3272 + }, + { + "epoch": 0.07657142857142857, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2659, + "step": 3273 + }, + { + "epoch": 0.07663157894736843, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2802, + "step": 3274 + }, + { + "epoch": 0.07669172932330827, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3047, + "step": 3275 + }, + { + "epoch": 0.07675187969924813, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3571, + "step": 3276 + }, + { + "epoch": 0.07681203007518797, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3775, + "step": 3277 + }, + { + "epoch": 0.07687218045112781, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3592, + "step": 3278 + }, + { + "epoch": 0.07693233082706767, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4099, + "step": 3279 + }, + { + "epoch": 0.07699248120300752, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3494, + "step": 3280 + }, + { + "epoch": 0.07705263157894737, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2731, + "step": 3281 + }, + { + "epoch": 0.07711278195488722, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3791, + "step": 3282 + }, + { + "epoch": 0.07717293233082707, + "grad_norm": 0.44140625, + "learning_rate": 3e-05, + "loss": 1.3721, + "step": 3283 + }, + { + "epoch": 0.07723308270676692, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3805, + "step": 3284 + }, + { + "epoch": 0.07729323308270676, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2778, + "step": 3285 + }, + { + "epoch": 0.07735338345864662, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.493, + "step": 3286 + }, + { + "epoch": 0.07741353383458646, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4172, + "step": 3287 + }, + { + "epoch": 0.07747368421052632, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3072, + "step": 3288 + }, + { + "epoch": 0.07753383458646616, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4207, + "step": 3289 + }, + { + "epoch": 0.07759398496240602, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.36, + "step": 3290 + }, + { + "epoch": 0.07765413533834586, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3137, + "step": 3291 + }, + { + "epoch": 0.07771428571428571, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3393, + "step": 3292 + }, + { + "epoch": 0.07777443609022557, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2921, + "step": 3293 + }, + { + "epoch": 0.07783458646616541, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3102, + "step": 3294 + }, + { + "epoch": 0.07789473684210527, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3057, + "step": 3295 + }, + { + "epoch": 0.07795488721804511, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.327, + "step": 3296 + }, + { + "epoch": 0.07801503759398497, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2905, + "step": 3297 + }, + { + "epoch": 0.07807518796992481, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.297, + "step": 3298 + }, + { + "epoch": 0.07813533834586465, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3627, + "step": 3299 + }, + { + "epoch": 0.07819548872180451, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4009, + "step": 3300 + }, + { + "epoch": 0.07825563909774436, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3791, + "step": 3301 + }, + { + "epoch": 0.07831578947368421, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2887, + "step": 3302 + }, + { + "epoch": 0.07837593984962406, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3319, + "step": 3303 + }, + { + "epoch": 0.07843609022556391, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3011, + "step": 3304 + }, + { + "epoch": 0.07849624060150376, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2801, + "step": 3305 + }, + { + "epoch": 0.07855639097744362, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.4239, + "step": 3306 + }, + { + "epoch": 0.07861654135338346, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3526, + "step": 3307 + }, + { + "epoch": 0.0786766917293233, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2915, + "step": 3308 + }, + { + "epoch": 0.07873684210526316, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.414, + "step": 3309 + }, + { + "epoch": 0.078796992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2922, + "step": 3310 + }, + { + "epoch": 0.07885714285714286, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2659, + "step": 3311 + }, + { + "epoch": 0.0789172932330827, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.4171, + "step": 3312 + }, + { + "epoch": 0.07897744360902256, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3596, + "step": 3313 + }, + { + "epoch": 0.0790375939849624, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2986, + "step": 3314 + }, + { + "epoch": 0.07909774436090225, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3986, + "step": 3315 + }, + { + "epoch": 0.07915789473684211, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3621, + "step": 3316 + }, + { + "epoch": 0.07921804511278195, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.3697, + "step": 3317 + }, + { + "epoch": 0.07927819548872181, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3784, + "step": 3318 + }, + { + "epoch": 0.07933834586466165, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3342, + "step": 3319 + }, + { + "epoch": 0.07939849624060151, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.36, + "step": 3320 + }, + { + "epoch": 0.07945864661654135, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3321, + "step": 3321 + }, + { + "epoch": 0.0795187969924812, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.35, + "step": 3322 + }, + { + "epoch": 0.07957894736842105, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.287, + "step": 3323 + }, + { + "epoch": 0.0796390977443609, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.3354, + "step": 3324 + }, + { + "epoch": 0.07969924812030076, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3285, + "step": 3325 + }, + { + "epoch": 0.0797593984962406, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3492, + "step": 3326 + }, + { + "epoch": 0.07981954887218046, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3128, + "step": 3327 + }, + { + "epoch": 0.0798796992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3208, + "step": 3328 + }, + { + "epoch": 0.07993984962406014, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2537, + "step": 3329 + }, + { + "epoch": 0.08, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.4124, + "step": 3330 + }, + { + "epoch": 0.08006015037593985, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.362, + "step": 3331 + }, + { + "epoch": 0.0801203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3547, + "step": 3332 + }, + { + "epoch": 0.08018045112781955, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.3695, + "step": 3333 + }, + { + "epoch": 0.0802406015037594, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3363, + "step": 3334 + }, + { + "epoch": 0.08030075187969925, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3389, + "step": 3335 + }, + { + "epoch": 0.08036090225563909, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.4521, + "step": 3336 + }, + { + "epoch": 0.08042105263157895, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3201, + "step": 3337 + }, + { + "epoch": 0.08048120300751879, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3428, + "step": 3338 + }, + { + "epoch": 0.08054135338345865, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4281, + "step": 3339 + }, + { + "epoch": 0.0806015037593985, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3144, + "step": 3340 + }, + { + "epoch": 0.08066165413533835, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.4295, + "step": 3341 + }, + { + "epoch": 0.0807218045112782, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2872, + "step": 3342 + }, + { + "epoch": 0.08078195488721804, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3656, + "step": 3343 + }, + { + "epoch": 0.0808421052631579, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4115, + "step": 3344 + }, + { + "epoch": 0.08090225563909774, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3617, + "step": 3345 + }, + { + "epoch": 0.0809624060150376, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3839, + "step": 3346 + }, + { + "epoch": 0.08102255639097744, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.2536, + "step": 3347 + }, + { + "epoch": 0.0810827067669173, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3749, + "step": 3348 + }, + { + "epoch": 0.08114285714285714, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3095, + "step": 3349 + }, + { + "epoch": 0.081203007518797, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2878, + "step": 3350 + }, + { + "epoch": 0.08126315789473684, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3848, + "step": 3351 + }, + { + "epoch": 0.08132330827067669, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3274, + "step": 3352 + }, + { + "epoch": 0.08138345864661654, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4005, + "step": 3353 + }, + { + "epoch": 0.08144360902255639, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3645, + "step": 3354 + }, + { + "epoch": 0.08150375939849624, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3526, + "step": 3355 + }, + { + "epoch": 0.08156390977443609, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.3265, + "step": 3356 + }, + { + "epoch": 0.08162406015037595, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.3635, + "step": 3357 + }, + { + "epoch": 0.08168421052631579, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3043, + "step": 3358 + }, + { + "epoch": 0.08174436090225563, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.334, + "step": 3359 + }, + { + "epoch": 0.08180451127819549, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3679, + "step": 3360 + }, + { + "epoch": 0.08186466165413533, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3418, + "step": 3361 + }, + { + "epoch": 0.08192481203007519, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.368, + "step": 3362 + }, + { + "epoch": 0.08198496240601504, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3399, + "step": 3363 + }, + { + "epoch": 0.0820451127819549, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3381, + "step": 3364 + }, + { + "epoch": 0.08210526315789474, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.398, + "step": 3365 + }, + { + "epoch": 0.08216541353383458, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3736, + "step": 3366 + }, + { + "epoch": 0.08222556390977444, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.3786, + "step": 3367 + }, + { + "epoch": 0.08228571428571428, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.6369, + "step": 3368 + }, + { + "epoch": 0.08234586466165414, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.3182, + "step": 3369 + }, + { + "epoch": 0.08240601503759398, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3055, + "step": 3370 + }, + { + "epoch": 0.08246616541353384, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2879, + "step": 3371 + }, + { + "epoch": 0.08252631578947368, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3165, + "step": 3372 + }, + { + "epoch": 0.08258646616541353, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3722, + "step": 3373 + }, + { + "epoch": 0.08264661654135338, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3631, + "step": 3374 + }, + { + "epoch": 0.08270676691729323, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3316, + "step": 3375 + }, + { + "epoch": 0.08276691729323309, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.4196, + "step": 3376 + }, + { + "epoch": 0.08282706766917293, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3423, + "step": 3377 + }, + { + "epoch": 0.08288721804511279, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.37, + "step": 3378 + }, + { + "epoch": 0.08294736842105263, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.4253, + "step": 3379 + }, + { + "epoch": 0.08300751879699247, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3796, + "step": 3380 + }, + { + "epoch": 0.08306766917293233, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2871, + "step": 3381 + }, + { + "epoch": 0.08312781954887218, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3492, + "step": 3382 + }, + { + "epoch": 0.08318796992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.402, + "step": 3383 + }, + { + "epoch": 0.08324812030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3785, + "step": 3384 + }, + { + "epoch": 0.08330827067669173, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3732, + "step": 3385 + }, + { + "epoch": 0.08336842105263158, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.3684, + "step": 3386 + }, + { + "epoch": 0.08342857142857144, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4149, + "step": 3387 + }, + { + "epoch": 0.08348872180451128, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3747, + "step": 3388 + }, + { + "epoch": 0.08354887218045112, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.3164, + "step": 3389 + }, + { + "epoch": 0.08360902255639098, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4269, + "step": 3390 + }, + { + "epoch": 0.08366917293233082, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3387, + "step": 3391 + }, + { + "epoch": 0.08372932330827068, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4009, + "step": 3392 + }, + { + "epoch": 0.08378947368421052, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3765, + "step": 3393 + }, + { + "epoch": 0.08384962406015038, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3477, + "step": 3394 + }, + { + "epoch": 0.08390977443609023, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3458, + "step": 3395 + }, + { + "epoch": 0.08396992481203007, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.3412, + "step": 3396 + }, + { + "epoch": 0.08403007518796993, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.3955, + "step": 3397 + }, + { + "epoch": 0.08409022556390977, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.372, + "step": 3398 + }, + { + "epoch": 0.08415037593984963, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2896, + "step": 3399 + }, + { + "epoch": 0.08421052631578947, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3652, + "step": 3400 + }, + { + "epoch": 0.08427067669172933, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3227, + "step": 3401 + }, + { + "epoch": 0.08433082706766917, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.3921, + "step": 3402 + }, + { + "epoch": 0.08439097744360902, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.5024, + "step": 3403 + }, + { + "epoch": 0.08445112781954887, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3075, + "step": 3404 + }, + { + "epoch": 0.08451127819548872, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3396, + "step": 3405 + }, + { + "epoch": 0.08457142857142858, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.31, + "step": 3406 + }, + { + "epoch": 0.08463157894736842, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2861, + "step": 3407 + }, + { + "epoch": 0.08469172932330828, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3231, + "step": 3408 + }, + { + "epoch": 0.08475187969924812, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.4479, + "step": 3409 + }, + { + "epoch": 0.08481203007518796, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.4631, + "step": 3410 + }, + { + "epoch": 0.08487218045112782, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.4565, + "step": 3411 + }, + { + "epoch": 0.08493233082706766, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2909, + "step": 3412 + }, + { + "epoch": 0.08499248120300752, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.323, + "step": 3413 + }, + { + "epoch": 0.08505263157894737, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.3271, + "step": 3414 + }, + { + "epoch": 0.08511278195488722, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3573, + "step": 3415 + }, + { + "epoch": 0.08517293233082707, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3917, + "step": 3416 + }, + { + "epoch": 0.08523308270676691, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4267, + "step": 3417 + }, + { + "epoch": 0.08529323308270677, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3369, + "step": 3418 + }, + { + "epoch": 0.08535338345864661, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3323, + "step": 3419 + }, + { + "epoch": 0.08541353383458647, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.3818, + "step": 3420 + }, + { + "epoch": 0.08547368421052631, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2834, + "step": 3421 + }, + { + "epoch": 0.08553383458646617, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3014, + "step": 3422 + }, + { + "epoch": 0.08559398496240601, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2565, + "step": 3423 + }, + { + "epoch": 0.08565413533834587, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3588, + "step": 3424 + }, + { + "epoch": 0.08571428571428572, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.371, + "step": 3425 + }, + { + "epoch": 0.08577443609022556, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.4287, + "step": 3426 + }, + { + "epoch": 0.08583458646616542, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3845, + "step": 3427 + }, + { + "epoch": 0.08589473684210526, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2889, + "step": 3428 + }, + { + "epoch": 0.08595488721804512, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.331, + "step": 3429 + }, + { + "epoch": 0.08601503759398496, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4102, + "step": 3430 + }, + { + "epoch": 0.08607518796992482, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3061, + "step": 3431 + }, + { + "epoch": 0.08613533834586466, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3775, + "step": 3432 + }, + { + "epoch": 0.0861954887218045, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.4402, + "step": 3433 + }, + { + "epoch": 0.08625563909774436, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3371, + "step": 3434 + }, + { + "epoch": 0.0863157894736842, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2818, + "step": 3435 + }, + { + "epoch": 0.08637593984962406, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3702, + "step": 3436 + }, + { + "epoch": 0.08643609022556391, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.3607, + "step": 3437 + }, + { + "epoch": 0.08649624060150377, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.3687, + "step": 3438 + }, + { + "epoch": 0.08655639097744361, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2854, + "step": 3439 + }, + { + "epoch": 0.08661654135338345, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3853, + "step": 3440 + }, + { + "epoch": 0.08667669172932331, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.4539, + "step": 3441 + }, + { + "epoch": 0.08673684210526315, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.3738, + "step": 3442 + }, + { + "epoch": 0.08679699248120301, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3031, + "step": 3443 + }, + { + "epoch": 0.08685714285714285, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3309, + "step": 3444 + }, + { + "epoch": 0.08691729323308271, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3716, + "step": 3445 + }, + { + "epoch": 0.08697744360902256, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3533, + "step": 3446 + }, + { + "epoch": 0.0870375939849624, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2966, + "step": 3447 + }, + { + "epoch": 0.08709774436090226, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3531, + "step": 3448 + }, + { + "epoch": 0.0871578947368421, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3383, + "step": 3449 + }, + { + "epoch": 0.08721804511278196, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3029, + "step": 3450 + }, + { + "epoch": 0.0872781954887218, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4629, + "step": 3451 + }, + { + "epoch": 0.08733834586466166, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4448, + "step": 3452 + }, + { + "epoch": 0.0873984962406015, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3388, + "step": 3453 + }, + { + "epoch": 0.08745864661654135, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2869, + "step": 3454 + }, + { + "epoch": 0.0875187969924812, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3418, + "step": 3455 + }, + { + "epoch": 0.08757894736842105, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.4446, + "step": 3456 + }, + { + "epoch": 0.0876390977443609, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.264, + "step": 3457 + }, + { + "epoch": 0.08769924812030075, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3304, + "step": 3458 + }, + { + "epoch": 0.0877593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3814, + "step": 3459 + }, + { + "epoch": 0.08781954887218045, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3273, + "step": 3460 + }, + { + "epoch": 0.08787969924812031, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3633, + "step": 3461 + }, + { + "epoch": 0.08793984962406015, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3908, + "step": 3462 + }, + { + "epoch": 0.088, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.3605, + "step": 3463 + }, + { + "epoch": 0.08806015037593985, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.394, + "step": 3464 + }, + { + "epoch": 0.0881203007518797, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3006, + "step": 3465 + }, + { + "epoch": 0.08818045112781955, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4363, + "step": 3466 + }, + { + "epoch": 0.0882406015037594, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3171, + "step": 3467 + }, + { + "epoch": 0.08830075187969925, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3648, + "step": 3468 + }, + { + "epoch": 0.0883609022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.366, + "step": 3469 + }, + { + "epoch": 0.08842105263157894, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3536, + "step": 3470 + }, + { + "epoch": 0.0884812030075188, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3027, + "step": 3471 + }, + { + "epoch": 0.08854135338345864, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.3031, + "step": 3472 + }, + { + "epoch": 0.0886015037593985, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2968, + "step": 3473 + }, + { + "epoch": 0.08866165413533834, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3603, + "step": 3474 + }, + { + "epoch": 0.0887218045112782, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3721, + "step": 3475 + }, + { + "epoch": 0.08878195488721805, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3345, + "step": 3476 + }, + { + "epoch": 0.08884210526315789, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.361, + "step": 3477 + }, + { + "epoch": 0.08890225563909775, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3909, + "step": 3478 + }, + { + "epoch": 0.08896240601503759, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3142, + "step": 3479 + }, + { + "epoch": 0.08902255639097745, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.3188, + "step": 3480 + }, + { + "epoch": 0.08908270676691729, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2593, + "step": 3481 + }, + { + "epoch": 0.08914285714285715, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3843, + "step": 3482 + }, + { + "epoch": 0.08920300751879699, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2939, + "step": 3483 + }, + { + "epoch": 0.08926315789473684, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3312, + "step": 3484 + }, + { + "epoch": 0.0893233082706767, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3184, + "step": 3485 + }, + { + "epoch": 0.08938345864661654, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3406, + "step": 3486 + }, + { + "epoch": 0.0894436090225564, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.3507, + "step": 3487 + }, + { + "epoch": 0.08950375939849624, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2451, + "step": 3488 + }, + { + "epoch": 0.0895639097744361, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3318, + "step": 3489 + }, + { + "epoch": 0.08962406015037594, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3512, + "step": 3490 + }, + { + "epoch": 0.08968421052631578, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.386, + "step": 3491 + }, + { + "epoch": 0.08974436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3706, + "step": 3492 + }, + { + "epoch": 0.08980451127819548, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3026, + "step": 3493 + }, + { + "epoch": 0.08986466165413534, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3638, + "step": 3494 + }, + { + "epoch": 0.08992481203007519, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3276, + "step": 3495 + }, + { + "epoch": 0.08998496240601504, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2916, + "step": 3496 + }, + { + "epoch": 0.09004511278195489, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3104, + "step": 3497 + }, + { + "epoch": 0.09010526315789473, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3065, + "step": 3498 + }, + { + "epoch": 0.09016541353383459, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3051, + "step": 3499 + }, + { + "epoch": 0.09022556390977443, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.3727, + "step": 3500 + }, + { + "epoch": 0.09028571428571429, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.2813, + "step": 3501 + }, + { + "epoch": 0.09034586466165413, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4439, + "step": 3502 + }, + { + "epoch": 0.09040601503759399, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3492, + "step": 3503 + }, + { + "epoch": 0.09046616541353383, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3388, + "step": 3504 + }, + { + "epoch": 0.09052631578947369, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.38, + "step": 3505 + }, + { + "epoch": 0.09058646616541353, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3869, + "step": 3506 + }, + { + "epoch": 0.09064661654135338, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.4134, + "step": 3507 + }, + { + "epoch": 0.09070676691729324, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 3508 + }, + { + "epoch": 0.09076691729323308, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3748, + "step": 3509 + }, + { + "epoch": 0.09082706766917294, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3091, + "step": 3510 + }, + { + "epoch": 0.09088721804511278, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2587, + "step": 3511 + }, + { + "epoch": 0.09094736842105264, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.5175, + "step": 3512 + }, + { + "epoch": 0.09100751879699248, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4128, + "step": 3513 + }, + { + "epoch": 0.09106766917293233, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3635, + "step": 3514 + }, + { + "epoch": 0.09112781954887218, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3296, + "step": 3515 + }, + { + "epoch": 0.09118796992481203, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3372, + "step": 3516 + }, + { + "epoch": 0.09124812030075188, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3677, + "step": 3517 + }, + { + "epoch": 0.09130827067669173, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3939, + "step": 3518 + }, + { + "epoch": 0.09136842105263158, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2641, + "step": 3519 + }, + { + "epoch": 0.09142857142857143, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3742, + "step": 3520 + }, + { + "epoch": 0.09148872180451127, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3543, + "step": 3521 + }, + { + "epoch": 0.09154887218045113, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 3522 + }, + { + "epoch": 0.09160902255639097, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.3548, + "step": 3523 + }, + { + "epoch": 0.09166917293233083, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3394, + "step": 3524 + }, + { + "epoch": 0.09172932330827067, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2957, + "step": 3525 + }, + { + "epoch": 0.09178947368421053, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3857, + "step": 3526 + }, + { + "epoch": 0.09184962406015038, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.384, + "step": 3527 + }, + { + "epoch": 0.09190977443609022, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3409, + "step": 3528 + }, + { + "epoch": 0.09196992481203008, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2254, + "step": 3529 + }, + { + "epoch": 0.09203007518796992, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 3530 + }, + { + "epoch": 0.09209022556390978, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3465, + "step": 3531 + }, + { + "epoch": 0.09215037593984962, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3689, + "step": 3532 + }, + { + "epoch": 0.09221052631578948, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3653, + "step": 3533 + }, + { + "epoch": 0.09227067669172932, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3719, + "step": 3534 + }, + { + "epoch": 0.09233082706766917, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3253, + "step": 3535 + }, + { + "epoch": 0.09239097744360902, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3768, + "step": 3536 + }, + { + "epoch": 0.09245112781954887, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3986, + "step": 3537 + }, + { + "epoch": 0.09251127819548872, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3547, + "step": 3538 + }, + { + "epoch": 0.09257142857142857, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3134, + "step": 3539 + }, + { + "epoch": 0.09263157894736843, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.3665, + "step": 3540 + }, + { + "epoch": 0.09269172932330827, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3044, + "step": 3541 + }, + { + "epoch": 0.09275187969924813, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3432, + "step": 3542 + }, + { + "epoch": 0.09281203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3167, + "step": 3543 + }, + { + "epoch": 0.09287218045112781, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2765, + "step": 3544 + }, + { + "epoch": 0.09293233082706767, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3221, + "step": 3545 + }, + { + "epoch": 0.09299248120300752, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3758, + "step": 3546 + }, + { + "epoch": 0.09305263157894737, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.3084, + "step": 3547 + }, + { + "epoch": 0.09311278195488722, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3876, + "step": 3548 + }, + { + "epoch": 0.09317293233082707, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.378, + "step": 3549 + }, + { + "epoch": 0.09323308270676692, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.3253, + "step": 3550 + }, + { + "epoch": 0.09329323308270676, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3392, + "step": 3551 + }, + { + "epoch": 0.09335338345864662, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2889, + "step": 3552 + }, + { + "epoch": 0.09341353383458646, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.4093, + "step": 3553 + }, + { + "epoch": 0.09347368421052632, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3245, + "step": 3554 + }, + { + "epoch": 0.09353383458646616, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3339, + "step": 3555 + }, + { + "epoch": 0.09359398496240602, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3202, + "step": 3556 + }, + { + "epoch": 0.09365413533834586, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.5055, + "step": 3557 + }, + { + "epoch": 0.09371428571428571, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2961, + "step": 3558 + }, + { + "epoch": 0.09377443609022557, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2697, + "step": 3559 + }, + { + "epoch": 0.09383458646616541, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.345, + "step": 3560 + }, + { + "epoch": 0.09389473684210527, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3246, + "step": 3561 + }, + { + "epoch": 0.09395488721804511, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3411, + "step": 3562 + }, + { + "epoch": 0.09401503759398497, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2477, + "step": 3563 + }, + { + "epoch": 0.09407518796992481, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4232, + "step": 3564 + }, + { + "epoch": 0.09413533834586466, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.3657, + "step": 3565 + }, + { + "epoch": 0.09419548872180451, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3495, + "step": 3566 + }, + { + "epoch": 0.09425563909774436, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3275, + "step": 3567 + }, + { + "epoch": 0.09431578947368421, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3513, + "step": 3568 + }, + { + "epoch": 0.09437593984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3365, + "step": 3569 + }, + { + "epoch": 0.09443609022556392, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2482, + "step": 3570 + }, + { + "epoch": 0.09449624060150376, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2336, + "step": 3571 + }, + { + "epoch": 0.0945563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4048, + "step": 3572 + }, + { + "epoch": 0.09461654135338346, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.325, + "step": 3573 + }, + { + "epoch": 0.0946766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3756, + "step": 3574 + }, + { + "epoch": 0.09473684210526316, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3039, + "step": 3575 + }, + { + "epoch": 0.094796992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3794, + "step": 3576 + }, + { + "epoch": 0.09485714285714286, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.355, + "step": 3577 + }, + { + "epoch": 0.0949172932330827, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3262, + "step": 3578 + }, + { + "epoch": 0.09497744360902256, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.3965, + "step": 3579 + }, + { + "epoch": 0.0950375939849624, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3967, + "step": 3580 + }, + { + "epoch": 0.09509774436090225, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.402, + "step": 3581 + }, + { + "epoch": 0.09515789473684211, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2812, + "step": 3582 + }, + { + "epoch": 0.09521804511278195, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4338, + "step": 3583 + }, + { + "epoch": 0.09527819548872181, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2554, + "step": 3584 + }, + { + "epoch": 0.09533834586466165, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3234, + "step": 3585 + }, + { + "epoch": 0.09539849624060151, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.4299, + "step": 3586 + }, + { + "epoch": 0.09545864661654135, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.4308, + "step": 3587 + }, + { + "epoch": 0.0955187969924812, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.4292, + "step": 3588 + }, + { + "epoch": 0.09557894736842105, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3032, + "step": 3589 + }, + { + "epoch": 0.0956390977443609, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3263, + "step": 3590 + }, + { + "epoch": 0.09569924812030076, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3778, + "step": 3591 + }, + { + "epoch": 0.0957593984962406, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.396, + "step": 3592 + }, + { + "epoch": 0.09581954887218046, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3846, + "step": 3593 + }, + { + "epoch": 0.0958796992481203, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.3072, + "step": 3594 + }, + { + "epoch": 0.09593984962406014, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.3634, + "step": 3595 + }, + { + "epoch": 0.096, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3609, + "step": 3596 + }, + { + "epoch": 0.09606015037593985, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.332, + "step": 3597 + }, + { + "epoch": 0.0961203007518797, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3301, + "step": 3598 + }, + { + "epoch": 0.09618045112781955, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3301, + "step": 3599 + }, + { + "epoch": 0.0962406015037594, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.3751, + "step": 3600 + }, + { + "epoch": 0.09630075187969925, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3067, + "step": 3601 + }, + { + "epoch": 0.09636090225563909, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3275, + "step": 3602 + }, + { + "epoch": 0.09642105263157895, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3846, + "step": 3603 + }, + { + "epoch": 0.09648120300751879, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3627, + "step": 3604 + }, + { + "epoch": 0.09654135338345865, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3436, + "step": 3605 + }, + { + "epoch": 0.0966015037593985, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2679, + "step": 3606 + }, + { + "epoch": 0.09666165413533835, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.342, + "step": 3607 + }, + { + "epoch": 0.0967218045112782, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.273, + "step": 3608 + }, + { + "epoch": 0.09678195488721804, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3022, + "step": 3609 + }, + { + "epoch": 0.0968421052631579, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2299, + "step": 3610 + }, + { + "epoch": 0.09690225563909774, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3345, + "step": 3611 + }, + { + "epoch": 0.0969624060150376, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2772, + "step": 3612 + }, + { + "epoch": 0.09702255639097744, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3261, + "step": 3613 + }, + { + "epoch": 0.0970827067669173, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3615, + "step": 3614 + }, + { + "epoch": 0.09714285714285714, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.4201, + "step": 3615 + }, + { + "epoch": 0.097203007518797, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.4281, + "step": 3616 + }, + { + "epoch": 0.09726315789473684, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.5575, + "step": 3617 + }, + { + "epoch": 0.09732330827067669, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3144, + "step": 3618 + }, + { + "epoch": 0.09738345864661654, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3459, + "step": 3619 + }, + { + "epoch": 0.09744360902255639, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.317, + "step": 3620 + }, + { + "epoch": 0.09750375939849625, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2596, + "step": 3621 + }, + { + "epoch": 0.09756390977443609, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.361, + "step": 3622 + }, + { + "epoch": 0.09762406015037595, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3257, + "step": 3623 + }, + { + "epoch": 0.09768421052631579, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3584, + "step": 3624 + }, + { + "epoch": 0.09774436090225563, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.276, + "step": 3625 + }, + { + "epoch": 0.09780451127819549, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3849, + "step": 3626 + }, + { + "epoch": 0.09786466165413533, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3184, + "step": 3627 + }, + { + "epoch": 0.09792481203007519, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 3628 + }, + { + "epoch": 0.09798496240601504, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3872, + "step": 3629 + }, + { + "epoch": 0.0980451127819549, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.4398, + "step": 3630 + }, + { + "epoch": 0.09810526315789474, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2812, + "step": 3631 + }, + { + "epoch": 0.09816541353383458, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.335, + "step": 3632 + }, + { + "epoch": 0.09822556390977444, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4189, + "step": 3633 + }, + { + "epoch": 0.09828571428571428, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2821, + "step": 3634 + }, + { + "epoch": 0.09834586466165414, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3356, + "step": 3635 + }, + { + "epoch": 0.09840601503759398, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.4174, + "step": 3636 + }, + { + "epoch": 0.09846616541353384, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.4794, + "step": 3637 + }, + { + "epoch": 0.09852631578947368, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3232, + "step": 3638 + }, + { + "epoch": 0.09858646616541353, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.3245, + "step": 3639 + }, + { + "epoch": 0.09864661654135339, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3358, + "step": 3640 + }, + { + "epoch": 0.09870676691729323, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3655, + "step": 3641 + }, + { + "epoch": 0.09876691729323309, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4011, + "step": 3642 + }, + { + "epoch": 0.09882706766917293, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.364, + "step": 3643 + }, + { + "epoch": 0.09888721804511279, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3794, + "step": 3644 + }, + { + "epoch": 0.09894736842105263, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3468, + "step": 3645 + }, + { + "epoch": 0.09900751879699247, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4774, + "step": 3646 + }, + { + "epoch": 0.09906766917293233, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.585, + "step": 3647 + }, + { + "epoch": 0.09912781954887218, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3043, + "step": 3648 + }, + { + "epoch": 0.09918796992481203, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3732, + "step": 3649 + }, + { + "epoch": 0.09924812030075188, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 3650 + }, + { + "epoch": 0.09930827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3474, + "step": 3651 + }, + { + "epoch": 0.09936842105263158, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3427, + "step": 3652 + }, + { + "epoch": 0.09942857142857142, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.276, + "step": 3653 + }, + { + "epoch": 0.09948872180451128, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2958, + "step": 3654 + }, + { + "epoch": 0.09954887218045112, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2965, + "step": 3655 + }, + { + "epoch": 0.09960902255639098, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3163, + "step": 3656 + }, + { + "epoch": 0.09966917293233082, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.366, + "step": 3657 + }, + { + "epoch": 0.09972932330827068, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3018, + "step": 3658 + }, + { + "epoch": 0.09978947368421053, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3721, + "step": 3659 + }, + { + "epoch": 0.09984962406015038, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2466, + "step": 3660 + }, + { + "epoch": 0.09990977443609023, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3385, + "step": 3661 + }, + { + "epoch": 0.09996992481203007, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.3177, + "step": 3662 + }, + { + "epoch": 0.10003007518796993, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3878, + "step": 3663 + }, + { + "epoch": 0.10009022556390977, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3397, + "step": 3664 + }, + { + "epoch": 0.10015037593984963, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2933, + "step": 3665 + }, + { + "epoch": 0.10021052631578947, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.3621, + "step": 3666 + }, + { + "epoch": 0.10027067669172933, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3054, + "step": 3667 + }, + { + "epoch": 0.10033082706766917, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4166, + "step": 3668 + }, + { + "epoch": 0.10039097744360902, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2979, + "step": 3669 + }, + { + "epoch": 0.10045112781954887, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3309, + "step": 3670 + }, + { + "epoch": 0.10051127819548872, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3732, + "step": 3671 + }, + { + "epoch": 0.10057142857142858, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3109, + "step": 3672 + }, + { + "epoch": 0.10063157894736842, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2555, + "step": 3673 + }, + { + "epoch": 0.10069172932330828, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3202, + "step": 3674 + }, + { + "epoch": 0.10075187969924812, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.3102, + "step": 3675 + }, + { + "epoch": 0.10081203007518796, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2317, + "step": 3676 + }, + { + "epoch": 0.10087218045112782, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3494, + "step": 3677 + }, + { + "epoch": 0.10093233082706767, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.4055, + "step": 3678 + }, + { + "epoch": 0.10099248120300752, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2996, + "step": 3679 + }, + { + "epoch": 0.10105263157894737, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3246, + "step": 3680 + }, + { + "epoch": 0.10111278195488722, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3173, + "step": 3681 + }, + { + "epoch": 0.10117293233082707, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3673, + "step": 3682 + }, + { + "epoch": 0.10123308270676691, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3987, + "step": 3683 + }, + { + "epoch": 0.10129323308270677, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.328, + "step": 3684 + }, + { + "epoch": 0.10135338345864661, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.4355, + "step": 3685 + }, + { + "epoch": 0.10141353383458647, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3116, + "step": 3686 + }, + { + "epoch": 0.10147368421052631, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 3687 + }, + { + "epoch": 0.10153383458646617, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3038, + "step": 3688 + }, + { + "epoch": 0.10159398496240601, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3293, + "step": 3689 + }, + { + "epoch": 0.10165413533834586, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2907, + "step": 3690 + }, + { + "epoch": 0.10171428571428572, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2484, + "step": 3691 + }, + { + "epoch": 0.10177443609022556, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3287, + "step": 3692 + }, + { + "epoch": 0.10183458646616542, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.367, + "step": 3693 + }, + { + "epoch": 0.10189473684210526, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3911, + "step": 3694 + }, + { + "epoch": 0.10195488721804512, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2598, + "step": 3695 + }, + { + "epoch": 0.10201503759398496, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3238, + "step": 3696 + }, + { + "epoch": 0.10207518796992482, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4097, + "step": 3697 + }, + { + "epoch": 0.10213533834586466, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3315, + "step": 3698 + }, + { + "epoch": 0.1021954887218045, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3598, + "step": 3699 + }, + { + "epoch": 0.10225563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.274, + "step": 3700 + }, + { + "epoch": 0.10231578947368421, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3031, + "step": 3701 + }, + { + "epoch": 0.10237593984962406, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.326, + "step": 3702 + }, + { + "epoch": 0.10243609022556391, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3492, + "step": 3703 + }, + { + "epoch": 0.10249624060150377, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3213, + "step": 3704 + }, + { + "epoch": 0.10255639097744361, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.4807, + "step": 3705 + }, + { + "epoch": 0.10261654135338345, + "grad_norm": 0.431640625, + "learning_rate": 3e-05, + "loss": 1.3738, + "step": 3706 + }, + { + "epoch": 0.10267669172932331, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.3072, + "step": 3707 + }, + { + "epoch": 0.10273684210526315, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2902, + "step": 3708 + }, + { + "epoch": 0.10279699248120301, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4083, + "step": 3709 + }, + { + "epoch": 0.10285714285714286, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.3588, + "step": 3710 + }, + { + "epoch": 0.10291729323308271, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.411, + "step": 3711 + }, + { + "epoch": 0.10297744360902256, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3954, + "step": 3712 + }, + { + "epoch": 0.1030375939849624, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3059, + "step": 3713 + }, + { + "epoch": 0.10309774436090226, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.4941, + "step": 3714 + }, + { + "epoch": 0.1031578947368421, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.367, + "step": 3715 + }, + { + "epoch": 0.10321804511278196, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2913, + "step": 3716 + }, + { + "epoch": 0.1032781954887218, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.6474, + "step": 3717 + }, + { + "epoch": 0.10333834586466166, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.3352, + "step": 3718 + }, + { + "epoch": 0.1033984962406015, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3842, + "step": 3719 + }, + { + "epoch": 0.10345864661654135, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.3932, + "step": 3720 + }, + { + "epoch": 0.1035187969924812, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3083, + "step": 3721 + }, + { + "epoch": 0.10357894736842105, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3298, + "step": 3722 + }, + { + "epoch": 0.1036390977443609, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.237, + "step": 3723 + }, + { + "epoch": 0.10369924812030075, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.3019, + "step": 3724 + }, + { + "epoch": 0.1037593984962406, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3336, + "step": 3725 + }, + { + "epoch": 0.10381954887218045, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.3757, + "step": 3726 + }, + { + "epoch": 0.1038796992481203, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.3176, + "step": 3727 + }, + { + "epoch": 0.10393984962406015, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2759, + "step": 3728 + }, + { + "epoch": 0.104, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3415, + "step": 3729 + }, + { + "epoch": 0.10406015037593985, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3166, + "step": 3730 + }, + { + "epoch": 0.1041203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3663, + "step": 3731 + }, + { + "epoch": 0.10418045112781955, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.302, + "step": 3732 + }, + { + "epoch": 0.1042406015037594, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.3799, + "step": 3733 + }, + { + "epoch": 0.10430075187969925, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2751, + "step": 3734 + }, + { + "epoch": 0.1043609022556391, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3027, + "step": 3735 + }, + { + "epoch": 0.10442105263157894, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3918, + "step": 3736 + }, + { + "epoch": 0.1044812030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3327, + "step": 3737 + }, + { + "epoch": 0.10454135338345864, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3649, + "step": 3738 + }, + { + "epoch": 0.1046015037593985, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3288, + "step": 3739 + }, + { + "epoch": 0.10466165413533834, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.3427, + "step": 3740 + }, + { + "epoch": 0.1047218045112782, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 3741 + }, + { + "epoch": 0.10478195488721805, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4187, + "step": 3742 + }, + { + "epoch": 0.10484210526315789, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.317, + "step": 3743 + }, + { + "epoch": 0.10490225563909775, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2902, + "step": 3744 + }, + { + "epoch": 0.10496240601503759, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.4098, + "step": 3745 + }, + { + "epoch": 0.10502255639097745, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.3581, + "step": 3746 + }, + { + "epoch": 0.10508270676691729, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.4258, + "step": 3747 + }, + { + "epoch": 0.10514285714285715, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3124, + "step": 3748 + }, + { + "epoch": 0.10520300751879699, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2875, + "step": 3749 + }, + { + "epoch": 0.10526315789473684, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.25, + "step": 3750 + }, + { + "epoch": 0.1053233082706767, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.3193, + "step": 3751 + }, + { + "epoch": 0.10538345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2553, + "step": 3752 + }, + { + "epoch": 0.1054436090225564, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2726, + "step": 3753 + }, + { + "epoch": 0.10550375939849624, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3165, + "step": 3754 + }, + { + "epoch": 0.1055639097744361, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3233, + "step": 3755 + }, + { + "epoch": 0.10562406015037594, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2886, + "step": 3756 + }, + { + "epoch": 0.10568421052631578, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3795, + "step": 3757 + }, + { + "epoch": 0.10574436090225564, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2943, + "step": 3758 + }, + { + "epoch": 0.10580451127819548, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2986, + "step": 3759 + }, + { + "epoch": 0.10586466165413534, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2978, + "step": 3760 + }, + { + "epoch": 0.10592481203007519, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3106, + "step": 3761 + }, + { + "epoch": 0.10598496240601504, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.3371, + "step": 3762 + }, + { + "epoch": 0.10604511278195489, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3599, + "step": 3763 + }, + { + "epoch": 0.10610526315789473, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.3485, + "step": 3764 + }, + { + "epoch": 0.10616541353383459, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.4493, + "step": 3765 + }, + { + "epoch": 0.10622556390977443, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3716, + "step": 3766 + }, + { + "epoch": 0.10628571428571429, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3816, + "step": 3767 + }, + { + "epoch": 0.10634586466165413, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3831, + "step": 3768 + }, + { + "epoch": 0.10640601503759399, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2341, + "step": 3769 + }, + { + "epoch": 0.10646616541353383, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3444, + "step": 3770 + }, + { + "epoch": 0.10652631578947368, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4301, + "step": 3771 + }, + { + "epoch": 0.10658646616541353, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3465, + "step": 3772 + }, + { + "epoch": 0.10664661654135338, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4354, + "step": 3773 + }, + { + "epoch": 0.10670676691729324, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3153, + "step": 3774 + }, + { + "epoch": 0.10676691729323308, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2716, + "step": 3775 + }, + { + "epoch": 0.10682706766917294, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2824, + "step": 3776 + }, + { + "epoch": 0.10688721804511278, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2414, + "step": 3777 + }, + { + "epoch": 0.10694736842105264, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2904, + "step": 3778 + }, + { + "epoch": 0.10700751879699248, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2948, + "step": 3779 + }, + { + "epoch": 0.10706766917293233, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.3043, + "step": 3780 + }, + { + "epoch": 0.10712781954887218, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 3781 + }, + { + "epoch": 0.10718796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.5377, + "step": 3782 + }, + { + "epoch": 0.10724812030075188, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2981, + "step": 3783 + }, + { + "epoch": 0.10730827067669173, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.3541, + "step": 3784 + }, + { + "epoch": 0.10736842105263159, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.362, + "step": 3785 + }, + { + "epoch": 0.10742857142857143, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3423, + "step": 3786 + }, + { + "epoch": 0.10748872180451127, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2409, + "step": 3787 + }, + { + "epoch": 0.10754887218045113, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.355, + "step": 3788 + }, + { + "epoch": 0.10760902255639097, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3515, + "step": 3789 + }, + { + "epoch": 0.10766917293233083, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.4097, + "step": 3790 + }, + { + "epoch": 0.10772932330827067, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.3574, + "step": 3791 + }, + { + "epoch": 0.10778947368421053, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 3792 + }, + { + "epoch": 0.10784962406015038, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4156, + "step": 3793 + }, + { + "epoch": 0.10790977443609022, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3722, + "step": 3794 + }, + { + "epoch": 0.10796992481203008, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3377, + "step": 3795 + }, + { + "epoch": 0.10803007518796992, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3656, + "step": 3796 + }, + { + "epoch": 0.10809022556390978, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3371, + "step": 3797 + }, + { + "epoch": 0.10815037593984962, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3928, + "step": 3798 + }, + { + "epoch": 0.10821052631578948, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3077, + "step": 3799 + }, + { + "epoch": 0.10827067669172932, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2909, + "step": 3800 + }, + { + "epoch": 0.10833082706766917, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.3157, + "step": 3801 + }, + { + "epoch": 0.10839097744360902, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3164, + "step": 3802 + }, + { + "epoch": 0.10845112781954887, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.4527, + "step": 3803 + }, + { + "epoch": 0.10851127819548873, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2463, + "step": 3804 + }, + { + "epoch": 0.10857142857142857, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.4179, + "step": 3805 + }, + { + "epoch": 0.10863157894736843, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3709, + "step": 3806 + }, + { + "epoch": 0.10869172932330827, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3001, + "step": 3807 + }, + { + "epoch": 0.10875187969924811, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.4279, + "step": 3808 + }, + { + "epoch": 0.10881203007518797, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.3272, + "step": 3809 + }, + { + "epoch": 0.10887218045112781, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2643, + "step": 3810 + }, + { + "epoch": 0.10893233082706767, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3115, + "step": 3811 + }, + { + "epoch": 0.10899248120300752, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.3825, + "step": 3812 + }, + { + "epoch": 0.10905263157894737, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3267, + "step": 3813 + }, + { + "epoch": 0.10911278195488722, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3867, + "step": 3814 + }, + { + "epoch": 0.10917293233082707, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3471, + "step": 3815 + }, + { + "epoch": 0.10923308270676692, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3012, + "step": 3816 + }, + { + "epoch": 0.10929323308270676, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2847, + "step": 3817 + }, + { + "epoch": 0.10935338345864662, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4001, + "step": 3818 + }, + { + "epoch": 0.10941353383458646, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2732, + "step": 3819 + }, + { + "epoch": 0.10947368421052632, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3217, + "step": 3820 + }, + { + "epoch": 0.10953383458646616, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.2865, + "step": 3821 + }, + { + "epoch": 0.10959398496240602, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3436, + "step": 3822 + }, + { + "epoch": 0.10965413533834587, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.549, + "step": 3823 + }, + { + "epoch": 0.10971428571428571, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.301, + "step": 3824 + }, + { + "epoch": 0.10977443609022557, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2878, + "step": 3825 + }, + { + "epoch": 0.10983458646616541, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.4443, + "step": 3826 + }, + { + "epoch": 0.10989473684210527, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.3439, + "step": 3827 + }, + { + "epoch": 0.10995488721804511, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3799, + "step": 3828 + }, + { + "epoch": 0.11001503759398497, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2634, + "step": 3829 + }, + { + "epoch": 0.11007518796992481, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3776, + "step": 3830 + }, + { + "epoch": 0.11013533834586466, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2676, + "step": 3831 + }, + { + "epoch": 0.11019548872180451, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3182, + "step": 3832 + }, + { + "epoch": 0.11025563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3258, + "step": 3833 + }, + { + "epoch": 0.11031578947368421, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3219, + "step": 3834 + }, + { + "epoch": 0.11037593984962406, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4125, + "step": 3835 + }, + { + "epoch": 0.11043609022556392, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3174, + "step": 3836 + }, + { + "epoch": 0.11049624060150376, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3888, + "step": 3837 + }, + { + "epoch": 0.1105563909774436, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.2969, + "step": 3838 + }, + { + "epoch": 0.11061654135338346, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2751, + "step": 3839 + }, + { + "epoch": 0.1106766917293233, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2868, + "step": 3840 + }, + { + "epoch": 0.11073684210526316, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.4164, + "step": 3841 + }, + { + "epoch": 0.110796992481203, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.3298, + "step": 3842 + }, + { + "epoch": 0.11085714285714286, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.3777, + "step": 3843 + }, + { + "epoch": 0.1109172932330827, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2534, + "step": 3844 + }, + { + "epoch": 0.11097744360902255, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.4244, + "step": 3845 + }, + { + "epoch": 0.11103759398496241, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 3846 + }, + { + "epoch": 0.11109774436090225, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3459, + "step": 3847 + }, + { + "epoch": 0.11115789473684211, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3044, + "step": 3848 + }, + { + "epoch": 0.11121804511278195, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3436, + "step": 3849 + }, + { + "epoch": 0.11127819548872181, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3787, + "step": 3850 + }, + { + "epoch": 0.11133834586466165, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3223, + "step": 3851 + }, + { + "epoch": 0.11139849624060151, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3442, + "step": 3852 + }, + { + "epoch": 0.11145864661654135, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1861, + "step": 3853 + }, + { + "epoch": 0.1115187969924812, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2822, + "step": 3854 + }, + { + "epoch": 0.11157894736842106, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3975, + "step": 3855 + }, + { + "epoch": 0.1116390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.4649, + "step": 3856 + }, + { + "epoch": 0.11169924812030076, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.3099, + "step": 3857 + }, + { + "epoch": 0.1117593984962406, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.294, + "step": 3858 + }, + { + "epoch": 0.11181954887218046, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4539, + "step": 3859 + }, + { + "epoch": 0.1118796992481203, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.4329, + "step": 3860 + }, + { + "epoch": 0.11193984962406014, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.3178, + "step": 3861 + }, + { + "epoch": 0.112, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.3806, + "step": 3862 + }, + { + "epoch": 0.11206015037593985, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2822, + "step": 3863 + }, + { + "epoch": 0.1121203007518797, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3188, + "step": 3864 + }, + { + "epoch": 0.11218045112781955, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3749, + "step": 3865 + }, + { + "epoch": 0.1122406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3042, + "step": 3866 + }, + { + "epoch": 0.11230075187969925, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.3782, + "step": 3867 + }, + { + "epoch": 0.11236090225563909, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.4971, + "step": 3868 + }, + { + "epoch": 0.11242105263157895, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3696, + "step": 3869 + }, + { + "epoch": 0.11248120300751879, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3658, + "step": 3870 + }, + { + "epoch": 0.11254135338345865, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3645, + "step": 3871 + }, + { + "epoch": 0.1126015037593985, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3032, + "step": 3872 + }, + { + "epoch": 0.11266165413533835, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.2895, + "step": 3873 + }, + { + "epoch": 0.1127218045112782, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2844, + "step": 3874 + }, + { + "epoch": 0.11278195488721804, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.306, + "step": 3875 + }, + { + "epoch": 0.1128421052631579, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.4117, + "step": 3876 + }, + { + "epoch": 0.11290225563909774, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3411, + "step": 3877 + }, + { + "epoch": 0.1129624060150376, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3958, + "step": 3878 + }, + { + "epoch": 0.11302255639097744, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2839, + "step": 3879 + }, + { + "epoch": 0.1130827067669173, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2674, + "step": 3880 + }, + { + "epoch": 0.11314285714285714, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2732, + "step": 3881 + }, + { + "epoch": 0.11320300751879699, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.267, + "step": 3882 + }, + { + "epoch": 0.11326315789473684, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3566, + "step": 3883 + }, + { + "epoch": 0.11332330827067669, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4138, + "step": 3884 + }, + { + "epoch": 0.11338345864661654, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2737, + "step": 3885 + }, + { + "epoch": 0.11344360902255639, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3951, + "step": 3886 + }, + { + "epoch": 0.11350375939849625, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3266, + "step": 3887 + }, + { + "epoch": 0.11356390977443609, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 3888 + }, + { + "epoch": 0.11362406015037595, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1864, + "step": 3889 + }, + { + "epoch": 0.11368421052631579, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3706, + "step": 3890 + }, + { + "epoch": 0.11374436090225563, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.4081, + "step": 3891 + }, + { + "epoch": 0.11380451127819549, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3751, + "step": 3892 + }, + { + "epoch": 0.11386466165413534, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3039, + "step": 3893 + }, + { + "epoch": 0.11392481203007519, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3198, + "step": 3894 + }, + { + "epoch": 0.11398496240601504, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2661, + "step": 3895 + }, + { + "epoch": 0.1140451127819549, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3454, + "step": 3896 + }, + { + "epoch": 0.11410526315789474, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3369, + "step": 3897 + }, + { + "epoch": 0.11416541353383458, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.3038, + "step": 3898 + }, + { + "epoch": 0.11422556390977444, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.278, + "step": 3899 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3223, + "step": 3900 + }, + { + "epoch": 0.11434586466165414, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3849, + "step": 3901 + }, + { + "epoch": 0.11440601503759398, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.3095, + "step": 3902 + }, + { + "epoch": 0.11446616541353384, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3897, + "step": 3903 + }, + { + "epoch": 0.11452631578947368, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.5481, + "step": 3904 + }, + { + "epoch": 0.11458646616541353, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3283, + "step": 3905 + }, + { + "epoch": 0.11464661654135339, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2917, + "step": 3906 + }, + { + "epoch": 0.11470676691729323, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3089, + "step": 3907 + }, + { + "epoch": 0.11476691729323309, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2738, + "step": 3908 + }, + { + "epoch": 0.11482706766917293, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3431, + "step": 3909 + }, + { + "epoch": 0.11488721804511279, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3261, + "step": 3910 + }, + { + "epoch": 0.11494736842105263, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3544, + "step": 3911 + }, + { + "epoch": 0.11500751879699248, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3227, + "step": 3912 + }, + { + "epoch": 0.11506766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3008, + "step": 3913 + }, + { + "epoch": 0.11512781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3254, + "step": 3914 + }, + { + "epoch": 0.11518796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2872, + "step": 3915 + }, + { + "epoch": 0.11524812030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2514, + "step": 3916 + }, + { + "epoch": 0.11530827067669173, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3422, + "step": 3917 + }, + { + "epoch": 0.11536842105263158, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.4282, + "step": 3918 + }, + { + "epoch": 0.11542857142857142, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.4443, + "step": 3919 + }, + { + "epoch": 0.11548872180451128, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3636, + "step": 3920 + }, + { + "epoch": 0.11554887218045112, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.4007, + "step": 3921 + }, + { + "epoch": 0.11560902255639098, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.312, + "step": 3922 + }, + { + "epoch": 0.11566917293233082, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.3726, + "step": 3923 + }, + { + "epoch": 0.11572932330827068, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2567, + "step": 3924 + }, + { + "epoch": 0.11578947368421053, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3212, + "step": 3925 + }, + { + "epoch": 0.11584962406015037, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2763, + "step": 3926 + }, + { + "epoch": 0.11590977443609023, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2841, + "step": 3927 + }, + { + "epoch": 0.11596992481203007, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.3986, + "step": 3928 + }, + { + "epoch": 0.11603007518796993, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3722, + "step": 3929 + }, + { + "epoch": 0.11609022556390977, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3964, + "step": 3930 + }, + { + "epoch": 0.11615037593984963, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.4165, + "step": 3931 + }, + { + "epoch": 0.11621052631578947, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4144, + "step": 3932 + }, + { + "epoch": 0.11627067669172933, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3592, + "step": 3933 + }, + { + "epoch": 0.11633082706766917, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3571, + "step": 3934 + }, + { + "epoch": 0.11639097744360902, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2998, + "step": 3935 + }, + { + "epoch": 0.11645112781954887, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.352, + "step": 3936 + }, + { + "epoch": 0.11651127819548872, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3864, + "step": 3937 + }, + { + "epoch": 0.11657142857142858, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.4907, + "step": 3938 + }, + { + "epoch": 0.11663157894736842, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2652, + "step": 3939 + }, + { + "epoch": 0.11669172932330828, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.3395, + "step": 3940 + }, + { + "epoch": 0.11675187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3766, + "step": 3941 + }, + { + "epoch": 0.11681203007518796, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.3191, + "step": 3942 + }, + { + "epoch": 0.11687218045112782, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2756, + "step": 3943 + }, + { + "epoch": 0.11693233082706767, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3358, + "step": 3944 + }, + { + "epoch": 0.11699248120300752, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3656, + "step": 3945 + }, + { + "epoch": 0.11705263157894737, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3847, + "step": 3946 + }, + { + "epoch": 0.11711278195488722, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.295, + "step": 3947 + }, + { + "epoch": 0.11717293233082707, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.29, + "step": 3948 + }, + { + "epoch": 0.11723308270676691, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.284, + "step": 3949 + }, + { + "epoch": 0.11729323308270677, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3241, + "step": 3950 + }, + { + "epoch": 0.11735338345864661, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3015, + "step": 3951 + }, + { + "epoch": 0.11741353383458647, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3222, + "step": 3952 + }, + { + "epoch": 0.11747368421052631, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 3953 + }, + { + "epoch": 0.11753383458646617, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3127, + "step": 3954 + }, + { + "epoch": 0.11759398496240601, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2889, + "step": 3955 + }, + { + "epoch": 0.11765413533834586, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3292, + "step": 3956 + }, + { + "epoch": 0.11771428571428572, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 3957 + }, + { + "epoch": 0.11777443609022556, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.333, + "step": 3958 + }, + { + "epoch": 0.11783458646616542, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3684, + "step": 3959 + }, + { + "epoch": 0.11789473684210526, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.4087, + "step": 3960 + }, + { + "epoch": 0.11795488721804512, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3715, + "step": 3961 + }, + { + "epoch": 0.11801503759398496, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3001, + "step": 3962 + }, + { + "epoch": 0.1180751879699248, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3541, + "step": 3963 + }, + { + "epoch": 0.11813533834586466, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4036, + "step": 3964 + }, + { + "epoch": 0.1181954887218045, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3981, + "step": 3965 + }, + { + "epoch": 0.11825563909774436, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 3966 + }, + { + "epoch": 0.11831578947368421, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2943, + "step": 3967 + }, + { + "epoch": 0.11837593984962407, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2914, + "step": 3968 + }, + { + "epoch": 0.11843609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3252, + "step": 3969 + }, + { + "epoch": 0.11849624060150377, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2971, + "step": 3970 + }, + { + "epoch": 0.11855639097744361, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3331, + "step": 3971 + }, + { + "epoch": 0.11861654135338345, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.29, + "step": 3972 + }, + { + "epoch": 0.11867669172932331, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2707, + "step": 3973 + }, + { + "epoch": 0.11873684210526315, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3322, + "step": 3974 + }, + { + "epoch": 0.11879699248120301, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3361, + "step": 3975 + }, + { + "epoch": 0.11885714285714286, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2885, + "step": 3976 + }, + { + "epoch": 0.11891729323308271, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3781, + "step": 3977 + }, + { + "epoch": 0.11897744360902256, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2815, + "step": 3978 + }, + { + "epoch": 0.1190375939849624, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3405, + "step": 3979 + }, + { + "epoch": 0.11909774436090226, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3421, + "step": 3980 + }, + { + "epoch": 0.1191578947368421, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2838, + "step": 3981 + }, + { + "epoch": 0.11921804511278196, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.3124, + "step": 3982 + }, + { + "epoch": 0.1192781954887218, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.32, + "step": 3983 + }, + { + "epoch": 0.11933834586466166, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.4082, + "step": 3984 + }, + { + "epoch": 0.1193984962406015, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3605, + "step": 3985 + }, + { + "epoch": 0.11945864661654135, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3301, + "step": 3986 + }, + { + "epoch": 0.1195187969924812, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 3987 + }, + { + "epoch": 0.11957894736842105, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2952, + "step": 3988 + }, + { + "epoch": 0.1196390977443609, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.287, + "step": 3989 + }, + { + "epoch": 0.11969924812030075, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.3908, + "step": 3990 + }, + { + "epoch": 0.11975939849624061, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2704, + "step": 3991 + }, + { + "epoch": 0.11981954887218045, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.3559, + "step": 3992 + }, + { + "epoch": 0.1198796992481203, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.4195, + "step": 3993 + }, + { + "epoch": 0.11993984962406015, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3291, + "step": 3994 + }, + { + "epoch": 0.12, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3096, + "step": 3995 + }, + { + "epoch": 0.12006015037593985, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.313, + "step": 3996 + }, + { + "epoch": 0.1201203007518797, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.3641, + "step": 3997 + }, + { + "epoch": 0.12018045112781955, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3601, + "step": 3998 + }, + { + "epoch": 0.1202406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.341, + "step": 3999 + }, + { + "epoch": 0.12030075187969924, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.4395, + "step": 4000 + }, + { + "epoch": 0.1203609022556391, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2964, + "step": 4001 + }, + { + "epoch": 0.12042105263157894, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2984, + "step": 4002 + }, + { + "epoch": 0.1204812030075188, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 4003 + }, + { + "epoch": 0.12054135338345864, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2277, + "step": 4004 + }, + { + "epoch": 0.1206015037593985, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3059, + "step": 4005 + }, + { + "epoch": 0.12066165413533834, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3289, + "step": 4006 + }, + { + "epoch": 0.1207218045112782, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.3089, + "step": 4007 + }, + { + "epoch": 0.12078195488721805, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3461, + "step": 4008 + }, + { + "epoch": 0.12084210526315789, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3431, + "step": 4009 + }, + { + "epoch": 0.12090225563909775, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.3997, + "step": 4010 + }, + { + "epoch": 0.12096240601503759, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2902, + "step": 4011 + }, + { + "epoch": 0.12102255639097745, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2488, + "step": 4012 + }, + { + "epoch": 0.12108270676691729, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2992, + "step": 4013 + }, + { + "epoch": 0.12114285714285715, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3329, + "step": 4014 + }, + { + "epoch": 0.12120300751879699, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 4015 + }, + { + "epoch": 0.12126315789473684, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4373, + "step": 4016 + }, + { + "epoch": 0.1213233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4043, + "step": 4017 + }, + { + "epoch": 0.12138345864661654, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3658, + "step": 4018 + }, + { + "epoch": 0.1214436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.4059, + "step": 4019 + }, + { + "epoch": 0.12150375939849624, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.243, + "step": 4020 + }, + { + "epoch": 0.1215639097744361, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2965, + "step": 4021 + }, + { + "epoch": 0.12162406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3103, + "step": 4022 + }, + { + "epoch": 0.12168421052631578, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.33, + "step": 4023 + }, + { + "epoch": 0.12174436090225564, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3005, + "step": 4024 + }, + { + "epoch": 0.12180451127819548, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3914, + "step": 4025 + }, + { + "epoch": 0.12186466165413534, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2127, + "step": 4026 + }, + { + "epoch": 0.12192481203007519, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3449, + "step": 4027 + }, + { + "epoch": 0.12198496240601504, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2992, + "step": 4028 + }, + { + "epoch": 0.12204511278195489, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3489, + "step": 4029 + }, + { + "epoch": 0.12210526315789473, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2331, + "step": 4030 + }, + { + "epoch": 0.12216541353383459, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2908, + "step": 4031 + }, + { + "epoch": 0.12222556390977443, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2842, + "step": 4032 + }, + { + "epoch": 0.12228571428571429, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 4033 + }, + { + "epoch": 0.12234586466165413, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.473, + "step": 4034 + }, + { + "epoch": 0.12240601503759399, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2778, + "step": 4035 + }, + { + "epoch": 0.12246616541353383, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3403, + "step": 4036 + }, + { + "epoch": 0.12252631578947368, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2721, + "step": 4037 + }, + { + "epoch": 0.12258646616541354, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2746, + "step": 4038 + }, + { + "epoch": 0.12264661654135338, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.3071, + "step": 4039 + }, + { + "epoch": 0.12270676691729324, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.3703, + "step": 4040 + }, + { + "epoch": 0.12276691729323308, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2261, + "step": 4041 + }, + { + "epoch": 0.12282706766917294, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.4549, + "step": 4042 + }, + { + "epoch": 0.12288721804511278, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3212, + "step": 4043 + }, + { + "epoch": 0.12294736842105262, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3894, + "step": 4044 + }, + { + "epoch": 0.12300751879699248, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2513, + "step": 4045 + }, + { + "epoch": 0.12306766917293233, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2636, + "step": 4046 + }, + { + "epoch": 0.12312781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3815, + "step": 4047 + }, + { + "epoch": 0.12318796992481203, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3048, + "step": 4048 + }, + { + "epoch": 0.12324812030075188, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2994, + "step": 4049 + }, + { + "epoch": 0.12330827067669173, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3675, + "step": 4050 + }, + { + "epoch": 0.12336842105263159, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2583, + "step": 4051 + }, + { + "epoch": 0.12342857142857143, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3492, + "step": 4052 + }, + { + "epoch": 0.12348872180451127, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3469, + "step": 4053 + }, + { + "epoch": 0.12354887218045113, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.281, + "step": 4054 + }, + { + "epoch": 0.12360902255639097, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3444, + "step": 4055 + }, + { + "epoch": 0.12366917293233083, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3204, + "step": 4056 + }, + { + "epoch": 0.12372932330827068, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3092, + "step": 4057 + }, + { + "epoch": 0.12378947368421053, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.336, + "step": 4058 + }, + { + "epoch": 0.12384962406015038, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2623, + "step": 4059 + }, + { + "epoch": 0.12390977443609022, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.323, + "step": 4060 + }, + { + "epoch": 0.12396992481203008, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2878, + "step": 4061 + }, + { + "epoch": 0.12403007518796992, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2891, + "step": 4062 + }, + { + "epoch": 0.12409022556390978, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3908, + "step": 4063 + }, + { + "epoch": 0.12415037593984962, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2887, + "step": 4064 + }, + { + "epoch": 0.12421052631578948, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3382, + "step": 4065 + }, + { + "epoch": 0.12427067669172932, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2973, + "step": 4066 + }, + { + "epoch": 0.12433082706766917, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2766, + "step": 4067 + }, + { + "epoch": 0.12439097744360902, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2924, + "step": 4068 + }, + { + "epoch": 0.12445112781954887, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2951, + "step": 4069 + }, + { + "epoch": 0.12451127819548873, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2872, + "step": 4070 + }, + { + "epoch": 0.12457142857142857, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4478, + "step": 4071 + }, + { + "epoch": 0.12463157894736843, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3658, + "step": 4072 + }, + { + "epoch": 0.12469172932330827, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3725, + "step": 4073 + }, + { + "epoch": 0.12475187969924811, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3424, + "step": 4074 + }, + { + "epoch": 0.12481203007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2833, + "step": 4075 + }, + { + "epoch": 0.12487218045112781, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3025, + "step": 4076 + }, + { + "epoch": 0.12493233082706767, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3495, + "step": 4077 + }, + { + "epoch": 0.12499248120300752, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3116, + "step": 4078 + }, + { + "epoch": 0.12505263157894736, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.4234, + "step": 4079 + }, + { + "epoch": 0.12511278195488723, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.299, + "step": 4080 + }, + { + "epoch": 0.12517293233082707, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3277, + "step": 4081 + }, + { + "epoch": 0.12523308270676692, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4297, + "step": 4082 + }, + { + "epoch": 0.12529323308270676, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3519, + "step": 4083 + }, + { + "epoch": 0.1253533834586466, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2998, + "step": 4084 + }, + { + "epoch": 0.12541353383458648, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3426, + "step": 4085 + }, + { + "epoch": 0.12547368421052632, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.319, + "step": 4086 + }, + { + "epoch": 0.12553383458646616, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.4556, + "step": 4087 + }, + { + "epoch": 0.125593984962406, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3909, + "step": 4088 + }, + { + "epoch": 0.12565413533834585, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2572, + "step": 4089 + }, + { + "epoch": 0.12571428571428572, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2845, + "step": 4090 + }, + { + "epoch": 0.12577443609022557, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3375, + "step": 4091 + }, + { + "epoch": 0.1258345864661654, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.4156, + "step": 4092 + }, + { + "epoch": 0.12589473684210525, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4757, + "step": 4093 + }, + { + "epoch": 0.12595488721804513, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.287, + "step": 4094 + }, + { + "epoch": 0.12601503759398497, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3233, + "step": 4095 + }, + { + "epoch": 0.1260751879699248, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3742, + "step": 4096 + }, + { + "epoch": 0.12613533834586466, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3321, + "step": 4097 + }, + { + "epoch": 0.1261954887218045, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3337, + "step": 4098 + }, + { + "epoch": 0.12625563909774437, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3712, + "step": 4099 + }, + { + "epoch": 0.12631578947368421, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2932, + "step": 4100 + }, + { + "epoch": 0.12637593984962406, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2775, + "step": 4101 + }, + { + "epoch": 0.1264360902255639, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3428, + "step": 4102 + }, + { + "epoch": 0.12649624060150377, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3182, + "step": 4103 + }, + { + "epoch": 0.12655639097744362, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 4104 + }, + { + "epoch": 0.12661654135338346, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3436, + "step": 4105 + }, + { + "epoch": 0.1266766917293233, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.4061, + "step": 4106 + }, + { + "epoch": 0.12673684210526315, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3284, + "step": 4107 + }, + { + "epoch": 0.12679699248120302, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3452, + "step": 4108 + }, + { + "epoch": 0.12685714285714286, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.329, + "step": 4109 + }, + { + "epoch": 0.1269172932330827, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.2669, + "step": 4110 + }, + { + "epoch": 0.12697744360902255, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.3112, + "step": 4111 + }, + { + "epoch": 0.1270375939849624, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3274, + "step": 4112 + }, + { + "epoch": 0.12709774436090227, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2719, + "step": 4113 + }, + { + "epoch": 0.1271578947368421, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3402, + "step": 4114 + }, + { + "epoch": 0.12721804511278195, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2711, + "step": 4115 + }, + { + "epoch": 0.1272781954887218, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3522, + "step": 4116 + }, + { + "epoch": 0.12733834586466167, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.445, + "step": 4117 + }, + { + "epoch": 0.1273984962406015, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3584, + "step": 4118 + }, + { + "epoch": 0.12745864661654135, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3797, + "step": 4119 + }, + { + "epoch": 0.1275187969924812, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3453, + "step": 4120 + }, + { + "epoch": 0.12757894736842104, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2688, + "step": 4121 + }, + { + "epoch": 0.1276390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.341, + "step": 4122 + }, + { + "epoch": 0.12769924812030076, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3748, + "step": 4123 + }, + { + "epoch": 0.1277593984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3456, + "step": 4124 + }, + { + "epoch": 0.12781954887218044, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2696, + "step": 4125 + }, + { + "epoch": 0.1278796992481203, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3712, + "step": 4126 + }, + { + "epoch": 0.12793984962406016, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3939, + "step": 4127 + }, + { + "epoch": 0.128, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2649, + "step": 4128 + }, + { + "epoch": 0.12806015037593985, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2891, + "step": 4129 + }, + { + "epoch": 0.1281203007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2868, + "step": 4130 + }, + { + "epoch": 0.12818045112781956, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3836, + "step": 4131 + }, + { + "epoch": 0.1282406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2317, + "step": 4132 + }, + { + "epoch": 0.12830075187969925, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.3363, + "step": 4133 + }, + { + "epoch": 0.1283609022556391, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.3369, + "step": 4134 + }, + { + "epoch": 0.12842105263157894, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3752, + "step": 4135 + }, + { + "epoch": 0.1284812030075188, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2776, + "step": 4136 + }, + { + "epoch": 0.12854135338345865, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2685, + "step": 4137 + }, + { + "epoch": 0.1286015037593985, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2956, + "step": 4138 + }, + { + "epoch": 0.12866165413533834, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3132, + "step": 4139 + }, + { + "epoch": 0.12872180451127818, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2599, + "step": 4140 + }, + { + "epoch": 0.12878195488721805, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2813, + "step": 4141 + }, + { + "epoch": 0.1288421052631579, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.4185, + "step": 4142 + }, + { + "epoch": 0.12890225563909774, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3542, + "step": 4143 + }, + { + "epoch": 0.12896240601503758, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3614, + "step": 4144 + }, + { + "epoch": 0.12902255639097746, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.3277, + "step": 4145 + }, + { + "epoch": 0.1290827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3722, + "step": 4146 + }, + { + "epoch": 0.12914285714285714, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3327, + "step": 4147 + }, + { + "epoch": 0.129203007518797, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.3805, + "step": 4148 + }, + { + "epoch": 0.12926315789473683, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3895, + "step": 4149 + }, + { + "epoch": 0.1293233082706767, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.295, + "step": 4150 + }, + { + "epoch": 0.12938345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3644, + "step": 4151 + }, + { + "epoch": 0.1294436090225564, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2875, + "step": 4152 + }, + { + "epoch": 0.12950375939849623, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3084, + "step": 4153 + }, + { + "epoch": 0.1295639097744361, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3496, + "step": 4154 + }, + { + "epoch": 0.12962406015037595, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2711, + "step": 4155 + }, + { + "epoch": 0.1296842105263158, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3051, + "step": 4156 + }, + { + "epoch": 0.12974436090225563, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2902, + "step": 4157 + }, + { + "epoch": 0.12980451127819548, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.4, + "step": 4158 + }, + { + "epoch": 0.12986466165413535, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2571, + "step": 4159 + }, + { + "epoch": 0.1299248120300752, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2402, + "step": 4160 + }, + { + "epoch": 0.12998496240601504, + "grad_norm": 1.3984375, + "learning_rate": 3e-05, + "loss": 1.3289, + "step": 4161 + }, + { + "epoch": 0.13004511278195488, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3657, + "step": 4162 + }, + { + "epoch": 0.13010526315789472, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3073, + "step": 4163 + }, + { + "epoch": 0.1301654135338346, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.334, + "step": 4164 + }, + { + "epoch": 0.13022556390977444, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3674, + "step": 4165 + }, + { + "epoch": 0.13028571428571428, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2766, + "step": 4166 + }, + { + "epoch": 0.13034586466165413, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.3632, + "step": 4167 + }, + { + "epoch": 0.130406015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2545, + "step": 4168 + }, + { + "epoch": 0.13046616541353384, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 4169 + }, + { + "epoch": 0.13052631578947368, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3971, + "step": 4170 + }, + { + "epoch": 0.13058646616541353, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3439, + "step": 4171 + }, + { + "epoch": 0.13064661654135337, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.3518, + "step": 4172 + }, + { + "epoch": 0.13070676691729324, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3203, + "step": 4173 + }, + { + "epoch": 0.1307669172932331, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3868, + "step": 4174 + }, + { + "epoch": 0.13082706766917293, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3133, + "step": 4175 + }, + { + "epoch": 0.13088721804511277, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2632, + "step": 4176 + }, + { + "epoch": 0.13094736842105262, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2514, + "step": 4177 + }, + { + "epoch": 0.1310075187969925, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3611, + "step": 4178 + }, + { + "epoch": 0.13106766917293233, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3905, + "step": 4179 + }, + { + "epoch": 0.13112781954887218, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.4032, + "step": 4180 + }, + { + "epoch": 0.13118796992481202, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3059, + "step": 4181 + }, + { + "epoch": 0.1312481203007519, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3256, + "step": 4182 + }, + { + "epoch": 0.13130827067669174, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3287, + "step": 4183 + }, + { + "epoch": 0.13136842105263158, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2927, + "step": 4184 + }, + { + "epoch": 0.13142857142857142, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3243, + "step": 4185 + }, + { + "epoch": 0.13148872180451127, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.323, + "step": 4186 + }, + { + "epoch": 0.13154887218045114, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3328, + "step": 4187 + }, + { + "epoch": 0.13160902255639098, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.3337, + "step": 4188 + }, + { + "epoch": 0.13166917293233082, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3397, + "step": 4189 + }, + { + "epoch": 0.13172932330827067, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.4148, + "step": 4190 + }, + { + "epoch": 0.13178947368421054, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3271, + "step": 4191 + }, + { + "epoch": 0.13184962406015038, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2943, + "step": 4192 + }, + { + "epoch": 0.13190977443609023, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.3092, + "step": 4193 + }, + { + "epoch": 0.13196992481203007, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.4011, + "step": 4194 + }, + { + "epoch": 0.13203007518796991, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3497, + "step": 4195 + }, + { + "epoch": 0.13209022556390979, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.301, + "step": 4196 + }, + { + "epoch": 0.13215037593984963, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3662, + "step": 4197 + }, + { + "epoch": 0.13221052631578947, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3243, + "step": 4198 + }, + { + "epoch": 0.13227067669172932, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3447, + "step": 4199 + }, + { + "epoch": 0.13233082706766916, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.4253, + "step": 4200 + }, + { + "epoch": 0.13239097744360903, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.4458, + "step": 4201 + }, + { + "epoch": 0.13245112781954888, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2401, + "step": 4202 + }, + { + "epoch": 0.13251127819548872, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3765, + "step": 4203 + }, + { + "epoch": 0.13257142857142856, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.331, + "step": 4204 + }, + { + "epoch": 0.13263157894736843, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4057, + "step": 4205 + }, + { + "epoch": 0.13269172932330828, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3232, + "step": 4206 + }, + { + "epoch": 0.13275187969924812, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2926, + "step": 4207 + }, + { + "epoch": 0.13281203007518796, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3613, + "step": 4208 + }, + { + "epoch": 0.1328721804511278, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2717, + "step": 4209 + }, + { + "epoch": 0.13293233082706768, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3405, + "step": 4210 + }, + { + "epoch": 0.13299248120300752, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.4011, + "step": 4211 + }, + { + "epoch": 0.13305263157894737, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.308, + "step": 4212 + }, + { + "epoch": 0.1331127819548872, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2756, + "step": 4213 + }, + { + "epoch": 0.13317293233082705, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3789, + "step": 4214 + }, + { + "epoch": 0.13323308270676693, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3536, + "step": 4215 + }, + { + "epoch": 0.13329323308270677, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2162, + "step": 4216 + }, + { + "epoch": 0.1333533834586466, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2738, + "step": 4217 + }, + { + "epoch": 0.13341353383458646, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3207, + "step": 4218 + }, + { + "epoch": 0.13347368421052633, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3906, + "step": 4219 + }, + { + "epoch": 0.13353383458646617, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2927, + "step": 4220 + }, + { + "epoch": 0.13359398496240601, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2763, + "step": 4221 + }, + { + "epoch": 0.13365413533834586, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.288, + "step": 4222 + }, + { + "epoch": 0.1337142857142857, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3681, + "step": 4223 + }, + { + "epoch": 0.13377443609022557, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3128, + "step": 4224 + }, + { + "epoch": 0.13383458646616542, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.32, + "step": 4225 + }, + { + "epoch": 0.13389473684210526, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3068, + "step": 4226 + }, + { + "epoch": 0.1339548872180451, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2555, + "step": 4227 + }, + { + "epoch": 0.13401503759398498, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2752, + "step": 4228 + }, + { + "epoch": 0.13407518796992482, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.34, + "step": 4229 + }, + { + "epoch": 0.13413533834586466, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.3128, + "step": 4230 + }, + { + "epoch": 0.1341954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 4231 + }, + { + "epoch": 0.13425563909774435, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.5393, + "step": 4232 + }, + { + "epoch": 0.13431578947368422, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3059, + "step": 4233 + }, + { + "epoch": 0.13437593984962407, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3713, + "step": 4234 + }, + { + "epoch": 0.1344360902255639, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.3759, + "step": 4235 + }, + { + "epoch": 0.13449624060150375, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.4049, + "step": 4236 + }, + { + "epoch": 0.1345563909774436, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3416, + "step": 4237 + }, + { + "epoch": 0.13461654135338347, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3239, + "step": 4238 + }, + { + "epoch": 0.1346766917293233, + "grad_norm": 1.8125, + "learning_rate": 3e-05, + "loss": 1.3789, + "step": 4239 + }, + { + "epoch": 0.13473684210526315, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2883, + "step": 4240 + }, + { + "epoch": 0.134796992481203, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3925, + "step": 4241 + }, + { + "epoch": 0.13485714285714287, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2648, + "step": 4242 + }, + { + "epoch": 0.1349172932330827, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3304, + "step": 4243 + }, + { + "epoch": 0.13497744360902256, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2406, + "step": 4244 + }, + { + "epoch": 0.1350375939849624, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3133, + "step": 4245 + }, + { + "epoch": 0.13509774436090224, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3764, + "step": 4246 + }, + { + "epoch": 0.13515789473684212, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.336, + "step": 4247 + }, + { + "epoch": 0.13521804511278196, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4281, + "step": 4248 + }, + { + "epoch": 0.1352781954887218, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.3339, + "step": 4249 + }, + { + "epoch": 0.13533834586466165, + "grad_norm": 0.427734375, + "learning_rate": 3e-05, + "loss": 1.3126, + "step": 4250 + }, + { + "epoch": 0.1353984962406015, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3587, + "step": 4251 + }, + { + "epoch": 0.13545864661654136, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.3355, + "step": 4252 + }, + { + "epoch": 0.1355187969924812, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3526, + "step": 4253 + }, + { + "epoch": 0.13557894736842105, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3288, + "step": 4254 + }, + { + "epoch": 0.1356390977443609, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3114, + "step": 4255 + }, + { + "epoch": 0.13569924812030076, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2572, + "step": 4256 + }, + { + "epoch": 0.1357593984962406, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.3843, + "step": 4257 + }, + { + "epoch": 0.13581954887218045, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3526, + "step": 4258 + }, + { + "epoch": 0.1358796992481203, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 4259 + }, + { + "epoch": 0.13593984962406014, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2561, + "step": 4260 + }, + { + "epoch": 0.136, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3362, + "step": 4261 + }, + { + "epoch": 0.13606015037593985, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4043, + "step": 4262 + }, + { + "epoch": 0.1361203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2984, + "step": 4263 + }, + { + "epoch": 0.13618045112781954, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2779, + "step": 4264 + }, + { + "epoch": 0.1362406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2634, + "step": 4265 + }, + { + "epoch": 0.13630075187969926, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.305, + "step": 4266 + }, + { + "epoch": 0.1363609022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3412, + "step": 4267 + }, + { + "epoch": 0.13642105263157894, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3153, + "step": 4268 + }, + { + "epoch": 0.1364812030075188, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2911, + "step": 4269 + }, + { + "epoch": 0.13654135338345866, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3568, + "step": 4270 + }, + { + "epoch": 0.1366015037593985, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3031, + "step": 4271 + }, + { + "epoch": 0.13666165413533835, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3104, + "step": 4272 + }, + { + "epoch": 0.1367218045112782, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 4273 + }, + { + "epoch": 0.13678195488721803, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3385, + "step": 4274 + }, + { + "epoch": 0.1368421052631579, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3439, + "step": 4275 + }, + { + "epoch": 0.13690225563909775, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2526, + "step": 4276 + }, + { + "epoch": 0.1369624060150376, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2451, + "step": 4277 + }, + { + "epoch": 0.13702255639097743, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.3223, + "step": 4278 + }, + { + "epoch": 0.1370827067669173, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.3451, + "step": 4279 + }, + { + "epoch": 0.13714285714285715, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.369, + "step": 4280 + }, + { + "epoch": 0.137203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2811, + "step": 4281 + }, + { + "epoch": 0.13726315789473684, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2907, + "step": 4282 + }, + { + "epoch": 0.13732330827067668, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1861, + "step": 4283 + }, + { + "epoch": 0.13738345864661655, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3022, + "step": 4284 + }, + { + "epoch": 0.1374436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2568, + "step": 4285 + }, + { + "epoch": 0.13750375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3805, + "step": 4286 + }, + { + "epoch": 0.13756390977443608, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.3969, + "step": 4287 + }, + { + "epoch": 0.13762406015037593, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 4288 + }, + { + "epoch": 0.1376842105263158, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.4394, + "step": 4289 + }, + { + "epoch": 0.13774436090225564, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3643, + "step": 4290 + }, + { + "epoch": 0.13780451127819549, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2997, + "step": 4291 + }, + { + "epoch": 0.13786466165413533, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3105, + "step": 4292 + }, + { + "epoch": 0.1379248120300752, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3361, + "step": 4293 + }, + { + "epoch": 0.13798496240601504, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.4008, + "step": 4294 + }, + { + "epoch": 0.1380451127819549, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3607, + "step": 4295 + }, + { + "epoch": 0.13810526315789473, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3117, + "step": 4296 + }, + { + "epoch": 0.13816541353383457, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3337, + "step": 4297 + }, + { + "epoch": 0.13822556390977445, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3449, + "step": 4298 + }, + { + "epoch": 0.1382857142857143, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3012, + "step": 4299 + }, + { + "epoch": 0.13834586466165413, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2874, + "step": 4300 + }, + { + "epoch": 0.13840601503759398, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2504, + "step": 4301 + }, + { + "epoch": 0.13846616541353385, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2608, + "step": 4302 + }, + { + "epoch": 0.1385263157894737, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3192, + "step": 4303 + }, + { + "epoch": 0.13858646616541354, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.324, + "step": 4304 + }, + { + "epoch": 0.13864661654135338, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2955, + "step": 4305 + }, + { + "epoch": 0.13870676691729322, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3317, + "step": 4306 + }, + { + "epoch": 0.1387669172932331, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3475, + "step": 4307 + }, + { + "epoch": 0.13882706766917294, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.355, + "step": 4308 + }, + { + "epoch": 0.13888721804511278, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3205, + "step": 4309 + }, + { + "epoch": 0.13894736842105262, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3745, + "step": 4310 + }, + { + "epoch": 0.13900751879699247, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4283, + "step": 4311 + }, + { + "epoch": 0.13906766917293234, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3442, + "step": 4312 + }, + { + "epoch": 0.13912781954887218, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.28, + "step": 4313 + }, + { + "epoch": 0.13918796992481203, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.5474, + "step": 4314 + }, + { + "epoch": 0.13924812030075187, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3536, + "step": 4315 + }, + { + "epoch": 0.13930827067669174, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3198, + "step": 4316 + }, + { + "epoch": 0.13936842105263159, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3041, + "step": 4317 + }, + { + "epoch": 0.13942857142857143, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 4318 + }, + { + "epoch": 0.13948872180451127, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2518, + "step": 4319 + }, + { + "epoch": 0.13954887218045112, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3757, + "step": 4320 + }, + { + "epoch": 0.139609022556391, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2056, + "step": 4321 + }, + { + "epoch": 0.13966917293233083, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3752, + "step": 4322 + }, + { + "epoch": 0.13972932330827068, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2457, + "step": 4323 + }, + { + "epoch": 0.13978947368421052, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3323, + "step": 4324 + }, + { + "epoch": 0.13984962406015036, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3247, + "step": 4325 + }, + { + "epoch": 0.13990977443609023, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.4109, + "step": 4326 + }, + { + "epoch": 0.13996992481203008, + "grad_norm": 0.4296875, + "learning_rate": 3e-05, + "loss": 1.4268, + "step": 4327 + }, + { + "epoch": 0.14003007518796992, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2807, + "step": 4328 + }, + { + "epoch": 0.14009022556390976, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3935, + "step": 4329 + }, + { + "epoch": 0.14015037593984964, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.302, + "step": 4330 + }, + { + "epoch": 0.14021052631578948, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.3765, + "step": 4331 + }, + { + "epoch": 0.14027067669172932, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2942, + "step": 4332 + }, + { + "epoch": 0.14033082706766917, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.329, + "step": 4333 + }, + { + "epoch": 0.140390977443609, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.4004, + "step": 4334 + }, + { + "epoch": 0.14045112781954888, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.322, + "step": 4335 + }, + { + "epoch": 0.14051127819548873, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3989, + "step": 4336 + }, + { + "epoch": 0.14057142857142857, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 4337 + }, + { + "epoch": 0.1406315789473684, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3013, + "step": 4338 + }, + { + "epoch": 0.14069172932330828, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3079, + "step": 4339 + }, + { + "epoch": 0.14075187969924813, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.2829, + "step": 4340 + }, + { + "epoch": 0.14081203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3519, + "step": 4341 + }, + { + "epoch": 0.14087218045112782, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3494, + "step": 4342 + }, + { + "epoch": 0.14093233082706766, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.3063, + "step": 4343 + }, + { + "epoch": 0.14099248120300753, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3564, + "step": 4344 + }, + { + "epoch": 0.14105263157894737, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3216, + "step": 4345 + }, + { + "epoch": 0.14111278195488722, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3747, + "step": 4346 + }, + { + "epoch": 0.14117293233082706, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3703, + "step": 4347 + }, + { + "epoch": 0.1412330827067669, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3204, + "step": 4348 + }, + { + "epoch": 0.14129323308270678, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.233, + "step": 4349 + }, + { + "epoch": 0.14135338345864662, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2367, + "step": 4350 + }, + { + "epoch": 0.14141353383458646, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3245, + "step": 4351 + }, + { + "epoch": 0.1414736842105263, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3631, + "step": 4352 + }, + { + "epoch": 0.14153383458646618, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 4353 + }, + { + "epoch": 0.14159398496240602, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 4354 + }, + { + "epoch": 0.14165413533834587, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3299, + "step": 4355 + }, + { + "epoch": 0.1417142857142857, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3042, + "step": 4356 + }, + { + "epoch": 0.14177443609022555, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2513, + "step": 4357 + }, + { + "epoch": 0.14183458646616542, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1909, + "step": 4358 + }, + { + "epoch": 0.14189473684210527, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3211, + "step": 4359 + }, + { + "epoch": 0.1419548872180451, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.4091, + "step": 4360 + }, + { + "epoch": 0.14201503759398496, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3022, + "step": 4361 + }, + { + "epoch": 0.1420751879699248, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3875, + "step": 4362 + }, + { + "epoch": 0.14213533834586467, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2961, + "step": 4363 + }, + { + "epoch": 0.1421954887218045, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3653, + "step": 4364 + }, + { + "epoch": 0.14225563909774436, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3499, + "step": 4365 + }, + { + "epoch": 0.1423157894736842, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2331, + "step": 4366 + }, + { + "epoch": 0.14237593984962407, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3768, + "step": 4367 + }, + { + "epoch": 0.14243609022556392, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4238, + "step": 4368 + }, + { + "epoch": 0.14249624060150376, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3455, + "step": 4369 + }, + { + "epoch": 0.1425563909774436, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3442, + "step": 4370 + }, + { + "epoch": 0.14261654135338345, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.2652, + "step": 4371 + }, + { + "epoch": 0.14267669172932332, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2637, + "step": 4372 + }, + { + "epoch": 0.14273684210526316, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3592, + "step": 4373 + }, + { + "epoch": 0.142796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3887, + "step": 4374 + }, + { + "epoch": 0.14285714285714285, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3455, + "step": 4375 + }, + { + "epoch": 0.14291729323308272, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2397, + "step": 4376 + }, + { + "epoch": 0.14297744360902256, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 4377 + }, + { + "epoch": 0.1430375939849624, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2091, + "step": 4378 + }, + { + "epoch": 0.14309774436090225, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3927, + "step": 4379 + }, + { + "epoch": 0.1431578947368421, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3074, + "step": 4380 + }, + { + "epoch": 0.14321804511278197, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3251, + "step": 4381 + }, + { + "epoch": 0.1432781954887218, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 4382 + }, + { + "epoch": 0.14333834586466165, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3108, + "step": 4383 + }, + { + "epoch": 0.1433984962406015, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3301, + "step": 4384 + }, + { + "epoch": 0.14345864661654134, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3527, + "step": 4385 + }, + { + "epoch": 0.1435187969924812, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2857, + "step": 4386 + }, + { + "epoch": 0.14357894736842106, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2707, + "step": 4387 + }, + { + "epoch": 0.1436390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3533, + "step": 4388 + }, + { + "epoch": 0.14369924812030074, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2666, + "step": 4389 + }, + { + "epoch": 0.14375939849624061, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3057, + "step": 4390 + }, + { + "epoch": 0.14381954887218046, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.5619, + "step": 4391 + }, + { + "epoch": 0.1438796992481203, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3324, + "step": 4392 + }, + { + "epoch": 0.14393984962406015, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2754, + "step": 4393 + }, + { + "epoch": 0.144, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.286, + "step": 4394 + }, + { + "epoch": 0.14406015037593986, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3016, + "step": 4395 + }, + { + "epoch": 0.1441203007518797, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.416, + "step": 4396 + }, + { + "epoch": 0.14418045112781955, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2956, + "step": 4397 + }, + { + "epoch": 0.1442406015037594, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2556, + "step": 4398 + }, + { + "epoch": 0.14430075187969924, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3454, + "step": 4399 + }, + { + "epoch": 0.1443609022556391, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.5032, + "step": 4400 + }, + { + "epoch": 0.14442105263157895, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2019, + "step": 4401 + }, + { + "epoch": 0.1444812030075188, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.3232, + "step": 4402 + }, + { + "epoch": 0.14454135338345864, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2692, + "step": 4403 + }, + { + "epoch": 0.1446015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2845, + "step": 4404 + }, + { + "epoch": 0.14466165413533835, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2897, + "step": 4405 + }, + { + "epoch": 0.1447218045112782, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3219, + "step": 4406 + }, + { + "epoch": 0.14478195488721804, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2297, + "step": 4407 + }, + { + "epoch": 0.14484210526315788, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.28, + "step": 4408 + }, + { + "epoch": 0.14490225563909775, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3044, + "step": 4409 + }, + { + "epoch": 0.1449624060150376, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2742, + "step": 4410 + }, + { + "epoch": 0.14502255639097744, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3044, + "step": 4411 + }, + { + "epoch": 0.14508270676691729, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.381, + "step": 4412 + }, + { + "epoch": 0.14514285714285713, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3618, + "step": 4413 + }, + { + "epoch": 0.145203007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.335, + "step": 4414 + }, + { + "epoch": 0.14526315789473684, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.3764, + "step": 4415 + }, + { + "epoch": 0.1453233082706767, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3754, + "step": 4416 + }, + { + "epoch": 0.14538345864661653, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2714, + "step": 4417 + }, + { + "epoch": 0.1454436090225564, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2729, + "step": 4418 + }, + { + "epoch": 0.14550375939849625, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2801, + "step": 4419 + }, + { + "epoch": 0.1455639097744361, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2661, + "step": 4420 + }, + { + "epoch": 0.14562406015037593, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.39, + "step": 4421 + }, + { + "epoch": 0.14568421052631578, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3485, + "step": 4422 + }, + { + "epoch": 0.14574436090225565, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2867, + "step": 4423 + }, + { + "epoch": 0.1458045112781955, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 4424 + }, + { + "epoch": 0.14586466165413534, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.404, + "step": 4425 + }, + { + "epoch": 0.14592481203007518, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3731, + "step": 4426 + }, + { + "epoch": 0.14598496240601505, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2592, + "step": 4427 + }, + { + "epoch": 0.1460451127819549, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3112, + "step": 4428 + }, + { + "epoch": 0.14610526315789474, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3515, + "step": 4429 + }, + { + "epoch": 0.14616541353383458, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.4164, + "step": 4430 + }, + { + "epoch": 0.14622556390977443, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3216, + "step": 4431 + }, + { + "epoch": 0.1462857142857143, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2837, + "step": 4432 + }, + { + "epoch": 0.14634586466165414, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.3742, + "step": 4433 + }, + { + "epoch": 0.14640601503759398, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3292, + "step": 4434 + }, + { + "epoch": 0.14646616541353383, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3168, + "step": 4435 + }, + { + "epoch": 0.14652631578947367, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2918, + "step": 4436 + }, + { + "epoch": 0.14658646616541354, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2409, + "step": 4437 + }, + { + "epoch": 0.1466466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3571, + "step": 4438 + }, + { + "epoch": 0.14670676691729323, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3109, + "step": 4439 + }, + { + "epoch": 0.14676691729323307, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2851, + "step": 4440 + }, + { + "epoch": 0.14682706766917294, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.3495, + "step": 4441 + }, + { + "epoch": 0.1468872180451128, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2712, + "step": 4442 + }, + { + "epoch": 0.14694736842105263, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.339, + "step": 4443 + }, + { + "epoch": 0.14700751879699248, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3003, + "step": 4444 + }, + { + "epoch": 0.14706766917293232, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.3642, + "step": 4445 + }, + { + "epoch": 0.1471278195488722, + "grad_norm": 0.42578125, + "learning_rate": 3e-05, + "loss": 1.3107, + "step": 4446 + }, + { + "epoch": 0.14718796992481203, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.4299, + "step": 4447 + }, + { + "epoch": 0.14724812030075188, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.2966, + "step": 4448 + }, + { + "epoch": 0.14730827067669172, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2564, + "step": 4449 + }, + { + "epoch": 0.14736842105263157, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.3587, + "step": 4450 + }, + { + "epoch": 0.14742857142857144, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 4451 + }, + { + "epoch": 0.14748872180451128, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2822, + "step": 4452 + }, + { + "epoch": 0.14754887218045112, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3379, + "step": 4453 + }, + { + "epoch": 0.14760902255639097, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3415, + "step": 4454 + }, + { + "epoch": 0.14766917293233084, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2737, + "step": 4455 + }, + { + "epoch": 0.14772932330827068, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.331, + "step": 4456 + }, + { + "epoch": 0.14778947368421053, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.29, + "step": 4457 + }, + { + "epoch": 0.14784962406015037, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2422, + "step": 4458 + }, + { + "epoch": 0.1479097744360902, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3152, + "step": 4459 + }, + { + "epoch": 0.14796992481203008, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.221, + "step": 4460 + }, + { + "epoch": 0.14803007518796993, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3691, + "step": 4461 + }, + { + "epoch": 0.14809022556390977, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2791, + "step": 4462 + }, + { + "epoch": 0.14815037593984962, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2687, + "step": 4463 + }, + { + "epoch": 0.1482105263157895, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2703, + "step": 4464 + }, + { + "epoch": 0.14827067669172933, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2967, + "step": 4465 + }, + { + "epoch": 0.14833082706766917, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.371, + "step": 4466 + }, + { + "epoch": 0.14839097744360902, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3162, + "step": 4467 + }, + { + "epoch": 0.14845112781954886, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3106, + "step": 4468 + }, + { + "epoch": 0.14851127819548873, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.3285, + "step": 4469 + }, + { + "epoch": 0.14857142857142858, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3106, + "step": 4470 + }, + { + "epoch": 0.14863157894736842, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3085, + "step": 4471 + }, + { + "epoch": 0.14869172932330826, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2577, + "step": 4472 + }, + { + "epoch": 0.1487518796992481, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3116, + "step": 4473 + }, + { + "epoch": 0.14881203007518798, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3871, + "step": 4474 + }, + { + "epoch": 0.14887218045112782, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.3083, + "step": 4475 + }, + { + "epoch": 0.14893233082706767, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3592, + "step": 4476 + }, + { + "epoch": 0.1489924812030075, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.4007, + "step": 4477 + }, + { + "epoch": 0.14905263157894738, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2223, + "step": 4478 + }, + { + "epoch": 0.14911278195488722, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.2764, + "step": 4479 + }, + { + "epoch": 0.14917293233082707, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2474, + "step": 4480 + }, + { + "epoch": 0.1492330827067669, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2523, + "step": 4481 + }, + { + "epoch": 0.14929323308270676, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3008, + "step": 4482 + }, + { + "epoch": 0.14935338345864663, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3275, + "step": 4483 + }, + { + "epoch": 0.14941353383458647, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3056, + "step": 4484 + }, + { + "epoch": 0.14947368421052631, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 4485 + }, + { + "epoch": 0.14953383458646616, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3597, + "step": 4486 + }, + { + "epoch": 0.149593984962406, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3223, + "step": 4487 + }, + { + "epoch": 0.14965413533834587, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3484, + "step": 4488 + }, + { + "epoch": 0.14971428571428572, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3737, + "step": 4489 + }, + { + "epoch": 0.14977443609022556, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3541, + "step": 4490 + }, + { + "epoch": 0.1498345864661654, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3101, + "step": 4491 + }, + { + "epoch": 0.14989473684210528, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2774, + "step": 4492 + }, + { + "epoch": 0.14995488721804512, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.3213, + "step": 4493 + }, + { + "epoch": 0.15001503759398496, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2526, + "step": 4494 + }, + { + "epoch": 0.1500751879699248, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2753, + "step": 4495 + }, + { + "epoch": 0.15013533834586465, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3112, + "step": 4496 + }, + { + "epoch": 0.15019548872180452, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2499, + "step": 4497 + }, + { + "epoch": 0.15025563909774436, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2409, + "step": 4498 + }, + { + "epoch": 0.1503157894736842, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.259, + "step": 4499 + }, + { + "epoch": 0.15037593984962405, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3101, + "step": 4500 + }, + { + "epoch": 0.15043609022556392, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2871, + "step": 4501 + }, + { + "epoch": 0.15049624060150377, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3171, + "step": 4502 + }, + { + "epoch": 0.1505563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3688, + "step": 4503 + }, + { + "epoch": 0.15061654135338345, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2577, + "step": 4504 + }, + { + "epoch": 0.1506766917293233, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 4505 + }, + { + "epoch": 0.15073684210526317, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3798, + "step": 4506 + }, + { + "epoch": 0.150796992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 4507 + }, + { + "epoch": 0.15085714285714286, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2569, + "step": 4508 + }, + { + "epoch": 0.1509172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.37, + "step": 4509 + }, + { + "epoch": 0.15097744360902254, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2795, + "step": 4510 + }, + { + "epoch": 0.15103759398496241, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2851, + "step": 4511 + }, + { + "epoch": 0.15109774436090226, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3553, + "step": 4512 + }, + { + "epoch": 0.1511578947368421, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2716, + "step": 4513 + }, + { + "epoch": 0.15121804511278195, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 4514 + }, + { + "epoch": 0.15127819548872182, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3924, + "step": 4515 + }, + { + "epoch": 0.15133834586466166, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 4516 + }, + { + "epoch": 0.1513984962406015, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2633, + "step": 4517 + }, + { + "epoch": 0.15145864661654135, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2741, + "step": 4518 + }, + { + "epoch": 0.1515187969924812, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2916, + "step": 4519 + }, + { + "epoch": 0.15157894736842106, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2238, + "step": 4520 + }, + { + "epoch": 0.1516390977443609, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3056, + "step": 4521 + }, + { + "epoch": 0.15169924812030075, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.3046, + "step": 4522 + }, + { + "epoch": 0.1517593984962406, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3104, + "step": 4523 + }, + { + "epoch": 0.15181954887218044, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.349, + "step": 4524 + }, + { + "epoch": 0.1518796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3043, + "step": 4525 + }, + { + "epoch": 0.15193984962406015, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3268, + "step": 4526 + }, + { + "epoch": 0.152, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2155, + "step": 4527 + }, + { + "epoch": 0.15206015037593984, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.357, + "step": 4528 + }, + { + "epoch": 0.1521203007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3603, + "step": 4529 + }, + { + "epoch": 0.15218045112781955, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2484, + "step": 4530 + }, + { + "epoch": 0.1522406015037594, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3852, + "step": 4531 + }, + { + "epoch": 0.15230075187969924, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2832, + "step": 4532 + }, + { + "epoch": 0.15236090225563909, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3301, + "step": 4533 + }, + { + "epoch": 0.15242105263157896, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3456, + "step": 4534 + }, + { + "epoch": 0.1524812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3156, + "step": 4535 + }, + { + "epoch": 0.15254135338345864, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3763, + "step": 4536 + }, + { + "epoch": 0.1526015037593985, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.3665, + "step": 4537 + }, + { + "epoch": 0.15266165413533836, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2364, + "step": 4538 + }, + { + "epoch": 0.1527218045112782, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3852, + "step": 4539 + }, + { + "epoch": 0.15278195488721805, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2592, + "step": 4540 + }, + { + "epoch": 0.1528421052631579, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2788, + "step": 4541 + }, + { + "epoch": 0.15290225563909773, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.2926, + "step": 4542 + }, + { + "epoch": 0.1529624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.347, + "step": 4543 + }, + { + "epoch": 0.15302255639097745, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3747, + "step": 4544 + }, + { + "epoch": 0.1530827067669173, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3177, + "step": 4545 + }, + { + "epoch": 0.15314285714285714, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.3126, + "step": 4546 + }, + { + "epoch": 0.15320300751879698, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2691, + "step": 4547 + }, + { + "epoch": 0.15326315789473685, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3625, + "step": 4548 + }, + { + "epoch": 0.1533233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3358, + "step": 4549 + }, + { + "epoch": 0.15338345864661654, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2746, + "step": 4550 + }, + { + "epoch": 0.15344360902255638, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3718, + "step": 4551 + }, + { + "epoch": 0.15350375939849625, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2683, + "step": 4552 + }, + { + "epoch": 0.1535639097744361, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2697, + "step": 4553 + }, + { + "epoch": 0.15362406015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.325, + "step": 4554 + }, + { + "epoch": 0.15368421052631578, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3459, + "step": 4555 + }, + { + "epoch": 0.15374436090225563, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.438, + "step": 4556 + }, + { + "epoch": 0.1538045112781955, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2738, + "step": 4557 + }, + { + "epoch": 0.15386466165413534, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.4603, + "step": 4558 + }, + { + "epoch": 0.1539248120300752, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2391, + "step": 4559 + }, + { + "epoch": 0.15398496240601503, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2949, + "step": 4560 + }, + { + "epoch": 0.15404511278195487, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3503, + "step": 4561 + }, + { + "epoch": 0.15410526315789475, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3154, + "step": 4562 + }, + { + "epoch": 0.1541654135338346, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2181, + "step": 4563 + }, + { + "epoch": 0.15422556390977443, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 4564 + }, + { + "epoch": 0.15428571428571428, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2935, + "step": 4565 + }, + { + "epoch": 0.15434586466165415, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2885, + "step": 4566 + }, + { + "epoch": 0.154406015037594, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3685, + "step": 4567 + }, + { + "epoch": 0.15446616541353383, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.331, + "step": 4568 + }, + { + "epoch": 0.15452631578947368, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.346, + "step": 4569 + }, + { + "epoch": 0.15458646616541352, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3557, + "step": 4570 + }, + { + "epoch": 0.1546466165413534, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3115, + "step": 4571 + }, + { + "epoch": 0.15470676691729324, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.5521, + "step": 4572 + }, + { + "epoch": 0.15476691729323308, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.36, + "step": 4573 + }, + { + "epoch": 0.15482706766917292, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3215, + "step": 4574 + }, + { + "epoch": 0.1548872180451128, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.355, + "step": 4575 + }, + { + "epoch": 0.15494736842105264, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.3446, + "step": 4576 + }, + { + "epoch": 0.15500751879699248, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3758, + "step": 4577 + }, + { + "epoch": 0.15506766917293233, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.301, + "step": 4578 + }, + { + "epoch": 0.15512781954887217, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3948, + "step": 4579 + }, + { + "epoch": 0.15518796992481204, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2824, + "step": 4580 + }, + { + "epoch": 0.15524812030075189, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2914, + "step": 4581 + }, + { + "epoch": 0.15530827067669173, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3726, + "step": 4582 + }, + { + "epoch": 0.15536842105263157, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2645, + "step": 4583 + }, + { + "epoch": 0.15542857142857142, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2825, + "step": 4584 + }, + { + "epoch": 0.1554887218045113, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3071, + "step": 4585 + }, + { + "epoch": 0.15554887218045113, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3218, + "step": 4586 + }, + { + "epoch": 0.15560902255639097, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.3605, + "step": 4587 + }, + { + "epoch": 0.15566917293233082, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3124, + "step": 4588 + }, + { + "epoch": 0.1557293233082707, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3131, + "step": 4589 + }, + { + "epoch": 0.15578947368421053, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2674, + "step": 4590 + }, + { + "epoch": 0.15584962406015038, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2679, + "step": 4591 + }, + { + "epoch": 0.15590977443609022, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3352, + "step": 4592 + }, + { + "epoch": 0.15596992481203006, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2245, + "step": 4593 + }, + { + "epoch": 0.15603007518796994, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.3243, + "step": 4594 + }, + { + "epoch": 0.15609022556390978, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3367, + "step": 4595 + }, + { + "epoch": 0.15615037593984962, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4018, + "step": 4596 + }, + { + "epoch": 0.15621052631578947, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3389, + "step": 4597 + }, + { + "epoch": 0.1562706766917293, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2853, + "step": 4598 + }, + { + "epoch": 0.15633082706766918, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4232, + "step": 4599 + }, + { + "epoch": 0.15639097744360902, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3456, + "step": 4600 + }, + { + "epoch": 0.15645112781954887, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.387, + "step": 4601 + }, + { + "epoch": 0.1565112781954887, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 4602 + }, + { + "epoch": 0.15657142857142858, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3502, + "step": 4603 + }, + { + "epoch": 0.15663157894736843, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2965, + "step": 4604 + }, + { + "epoch": 0.15669172932330827, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.3219, + "step": 4605 + }, + { + "epoch": 0.15675187969924811, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.279, + "step": 4606 + }, + { + "epoch": 0.15681203007518796, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3388, + "step": 4607 + }, + { + "epoch": 0.15687218045112783, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.3909, + "step": 4608 + }, + { + "epoch": 0.15693233082706767, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3912, + "step": 4609 + }, + { + "epoch": 0.15699248120300752, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3395, + "step": 4610 + }, + { + "epoch": 0.15705263157894736, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3188, + "step": 4611 + }, + { + "epoch": 0.15711278195488723, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2449, + "step": 4612 + }, + { + "epoch": 0.15717293233082708, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3513, + "step": 4613 + }, + { + "epoch": 0.15723308270676692, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3901, + "step": 4614 + }, + { + "epoch": 0.15729323308270676, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 4615 + }, + { + "epoch": 0.1573533834586466, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3557, + "step": 4616 + }, + { + "epoch": 0.15741353383458648, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.3824, + "step": 4617 + }, + { + "epoch": 0.15747368421052632, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2655, + "step": 4618 + }, + { + "epoch": 0.15753383458646616, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2083, + "step": 4619 + }, + { + "epoch": 0.157593984962406, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3293, + "step": 4620 + }, + { + "epoch": 0.15765413533834585, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3873, + "step": 4621 + }, + { + "epoch": 0.15771428571428572, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.4121, + "step": 4622 + }, + { + "epoch": 0.15777443609022557, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3513, + "step": 4623 + }, + { + "epoch": 0.1578345864661654, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3525, + "step": 4624 + }, + { + "epoch": 0.15789473684210525, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3037, + "step": 4625 + }, + { + "epoch": 0.15795488721804513, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2756, + "step": 4626 + }, + { + "epoch": 0.15801503759398497, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3573, + "step": 4627 + }, + { + "epoch": 0.1580751879699248, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.306, + "step": 4628 + }, + { + "epoch": 0.15813533834586466, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3353, + "step": 4629 + }, + { + "epoch": 0.1581954887218045, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.5209, + "step": 4630 + }, + { + "epoch": 0.15825563909774437, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2579, + "step": 4631 + }, + { + "epoch": 0.15831578947368422, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3579, + "step": 4632 + }, + { + "epoch": 0.15837593984962406, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.365, + "step": 4633 + }, + { + "epoch": 0.1584360902255639, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2662, + "step": 4634 + }, + { + "epoch": 0.15849624060150375, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2982, + "step": 4635 + }, + { + "epoch": 0.15855639097744362, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 4636 + }, + { + "epoch": 0.15861654135338346, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2397, + "step": 4637 + }, + { + "epoch": 0.1586766917293233, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2523, + "step": 4638 + }, + { + "epoch": 0.15873684210526315, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3417, + "step": 4639 + }, + { + "epoch": 0.15879699248120302, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.3191, + "step": 4640 + }, + { + "epoch": 0.15885714285714286, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3659, + "step": 4641 + }, + { + "epoch": 0.1589172932330827, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3216, + "step": 4642 + }, + { + "epoch": 0.15897744360902255, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3556, + "step": 4643 + }, + { + "epoch": 0.1590375939849624, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2752, + "step": 4644 + }, + { + "epoch": 0.15909774436090227, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2303, + "step": 4645 + }, + { + "epoch": 0.1591578947368421, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3165, + "step": 4646 + }, + { + "epoch": 0.15921804511278195, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.317, + "step": 4647 + }, + { + "epoch": 0.1592781954887218, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2909, + "step": 4648 + }, + { + "epoch": 0.15933834586466167, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3853, + "step": 4649 + }, + { + "epoch": 0.1593984962406015, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3226, + "step": 4650 + }, + { + "epoch": 0.15945864661654136, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3194, + "step": 4651 + }, + { + "epoch": 0.1595187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3025, + "step": 4652 + }, + { + "epoch": 0.15957894736842104, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2688, + "step": 4653 + }, + { + "epoch": 0.1596390977443609, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.2629, + "step": 4654 + }, + { + "epoch": 0.15969924812030076, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3188, + "step": 4655 + }, + { + "epoch": 0.1597593984962406, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2583, + "step": 4656 + }, + { + "epoch": 0.15981954887218044, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3396, + "step": 4657 + }, + { + "epoch": 0.1598796992481203, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2772, + "step": 4658 + }, + { + "epoch": 0.15993984962406016, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.284, + "step": 4659 + }, + { + "epoch": 0.16, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3143, + "step": 4660 + }, + { + "epoch": 0.16006015037593985, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3272, + "step": 4661 + }, + { + "epoch": 0.1601203007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 4662 + }, + { + "epoch": 0.16018045112781956, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3204, + "step": 4663 + }, + { + "epoch": 0.1602406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3328, + "step": 4664 + }, + { + "epoch": 0.16030075187969925, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.3343, + "step": 4665 + }, + { + "epoch": 0.1603609022556391, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3013, + "step": 4666 + }, + { + "epoch": 0.16042105263157894, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 4667 + }, + { + "epoch": 0.1604812030075188, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2827, + "step": 4668 + }, + { + "epoch": 0.16054135338345865, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2346, + "step": 4669 + }, + { + "epoch": 0.1606015037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2223, + "step": 4670 + }, + { + "epoch": 0.16066165413533834, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.3589, + "step": 4671 + }, + { + "epoch": 0.16072180451127818, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3096, + "step": 4672 + }, + { + "epoch": 0.16078195488721805, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2532, + "step": 4673 + }, + { + "epoch": 0.1608421052631579, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2681, + "step": 4674 + }, + { + "epoch": 0.16090225563909774, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2183, + "step": 4675 + }, + { + "epoch": 0.16096240601503758, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2864, + "step": 4676 + }, + { + "epoch": 0.16102255639097746, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2694, + "step": 4677 + }, + { + "epoch": 0.1610827067669173, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2742, + "step": 4678 + }, + { + "epoch": 0.16114285714285714, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3005, + "step": 4679 + }, + { + "epoch": 0.161203007518797, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.307, + "step": 4680 + }, + { + "epoch": 0.16126315789473683, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3212, + "step": 4681 + }, + { + "epoch": 0.1613233082706767, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3383, + "step": 4682 + }, + { + "epoch": 0.16138345864661655, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2283, + "step": 4683 + }, + { + "epoch": 0.1614436090225564, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 4684 + }, + { + "epoch": 0.16150375939849623, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2482, + "step": 4685 + }, + { + "epoch": 0.16156390977443608, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3329, + "step": 4686 + }, + { + "epoch": 0.16162406015037595, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2974, + "step": 4687 + }, + { + "epoch": 0.1616842105263158, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2567, + "step": 4688 + }, + { + "epoch": 0.16174436090225564, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2624, + "step": 4689 + }, + { + "epoch": 0.16180451127819548, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2349, + "step": 4690 + }, + { + "epoch": 0.16186466165413535, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3498, + "step": 4691 + }, + { + "epoch": 0.1619248120300752, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3319, + "step": 4692 + }, + { + "epoch": 0.16198496240601504, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2943, + "step": 4693 + }, + { + "epoch": 0.16204511278195488, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3598, + "step": 4694 + }, + { + "epoch": 0.16210526315789472, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2562, + "step": 4695 + }, + { + "epoch": 0.1621654135338346, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.244, + "step": 4696 + }, + { + "epoch": 0.16222556390977444, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2525, + "step": 4697 + }, + { + "epoch": 0.16228571428571428, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3465, + "step": 4698 + }, + { + "epoch": 0.16234586466165413, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3998, + "step": 4699 + }, + { + "epoch": 0.162406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3533, + "step": 4700 + }, + { + "epoch": 0.16246616541353384, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.279, + "step": 4701 + }, + { + "epoch": 0.16252631578947369, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.3321, + "step": 4702 + }, + { + "epoch": 0.16258646616541353, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3301, + "step": 4703 + }, + { + "epoch": 0.16264661654135337, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3156, + "step": 4704 + }, + { + "epoch": 0.16270676691729324, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3763, + "step": 4705 + }, + { + "epoch": 0.1627669172932331, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3546, + "step": 4706 + }, + { + "epoch": 0.16282706766917293, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2844, + "step": 4707 + }, + { + "epoch": 0.16288721804511277, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3415, + "step": 4708 + }, + { + "epoch": 0.16294736842105262, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.327, + "step": 4709 + }, + { + "epoch": 0.1630075187969925, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.321, + "step": 4710 + }, + { + "epoch": 0.16306766917293233, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3246, + "step": 4711 + }, + { + "epoch": 0.16312781954887218, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3011, + "step": 4712 + }, + { + "epoch": 0.16318796992481202, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3476, + "step": 4713 + }, + { + "epoch": 0.1632481203007519, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2289, + "step": 4714 + }, + { + "epoch": 0.16330827067669174, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3734, + "step": 4715 + }, + { + "epoch": 0.16336842105263158, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.328, + "step": 4716 + }, + { + "epoch": 0.16342857142857142, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2735, + "step": 4717 + }, + { + "epoch": 0.16348872180451127, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.298, + "step": 4718 + }, + { + "epoch": 0.16354887218045114, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3315, + "step": 4719 + }, + { + "epoch": 0.16360902255639098, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.3258, + "step": 4720 + }, + { + "epoch": 0.16366917293233083, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2445, + "step": 4721 + }, + { + "epoch": 0.16372932330827067, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3273, + "step": 4722 + }, + { + "epoch": 0.1637894736842105, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2509, + "step": 4723 + }, + { + "epoch": 0.16384962406015038, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2795, + "step": 4724 + }, + { + "epoch": 0.16390977443609023, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.5046, + "step": 4725 + }, + { + "epoch": 0.16396992481203007, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3715, + "step": 4726 + }, + { + "epoch": 0.16403007518796991, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2392, + "step": 4727 + }, + { + "epoch": 0.1640902255639098, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.38, + "step": 4728 + }, + { + "epoch": 0.16415037593984963, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2201, + "step": 4729 + }, + { + "epoch": 0.16421052631578947, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3424, + "step": 4730 + }, + { + "epoch": 0.16427067669172932, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3105, + "step": 4731 + }, + { + "epoch": 0.16433082706766916, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3753, + "step": 4732 + }, + { + "epoch": 0.16439097744360903, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3191, + "step": 4733 + }, + { + "epoch": 0.16445112781954888, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3625, + "step": 4734 + }, + { + "epoch": 0.16451127819548872, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.4569, + "step": 4735 + }, + { + "epoch": 0.16457142857142856, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2832, + "step": 4736 + }, + { + "epoch": 0.16463157894736843, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3127, + "step": 4737 + }, + { + "epoch": 0.16469172932330828, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2543, + "step": 4738 + }, + { + "epoch": 0.16475187969924812, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3027, + "step": 4739 + }, + { + "epoch": 0.16481203007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2745, + "step": 4740 + }, + { + "epoch": 0.1648721804511278, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.3476, + "step": 4741 + }, + { + "epoch": 0.16493233082706768, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3368, + "step": 4742 + }, + { + "epoch": 0.16499248120300752, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3665, + "step": 4743 + }, + { + "epoch": 0.16505263157894737, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2289, + "step": 4744 + }, + { + "epoch": 0.1651127819548872, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3858, + "step": 4745 + }, + { + "epoch": 0.16517293233082705, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3959, + "step": 4746 + }, + { + "epoch": 0.16523308270676693, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.344, + "step": 4747 + }, + { + "epoch": 0.16529323308270677, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2557, + "step": 4748 + }, + { + "epoch": 0.1653533834586466, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3293, + "step": 4749 + }, + { + "epoch": 0.16541353383458646, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3262, + "step": 4750 + }, + { + "epoch": 0.16547368421052633, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2864, + "step": 4751 + }, + { + "epoch": 0.16553383458646617, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2951, + "step": 4752 + }, + { + "epoch": 0.16559398496240602, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2864, + "step": 4753 + }, + { + "epoch": 0.16565413533834586, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2599, + "step": 4754 + }, + { + "epoch": 0.1657142857142857, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3824, + "step": 4755 + }, + { + "epoch": 0.16577443609022557, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2944, + "step": 4756 + }, + { + "epoch": 0.16583458646616542, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2511, + "step": 4757 + }, + { + "epoch": 0.16589473684210526, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.4233, + "step": 4758 + }, + { + "epoch": 0.1659548872180451, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2616, + "step": 4759 + }, + { + "epoch": 0.16601503759398495, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.288, + "step": 4760 + }, + { + "epoch": 0.16607518796992482, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.334, + "step": 4761 + }, + { + "epoch": 0.16613533834586466, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3154, + "step": 4762 + }, + { + "epoch": 0.1661954887218045, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2954, + "step": 4763 + }, + { + "epoch": 0.16625563909774435, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2816, + "step": 4764 + }, + { + "epoch": 0.16631578947368422, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2307, + "step": 4765 + }, + { + "epoch": 0.16637593984962407, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2945, + "step": 4766 + }, + { + "epoch": 0.1664360902255639, + "grad_norm": 0.421875, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 4767 + }, + { + "epoch": 0.16649624060150375, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3623, + "step": 4768 + }, + { + "epoch": 0.1665563909774436, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2768, + "step": 4769 + }, + { + "epoch": 0.16661654135338347, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.4101, + "step": 4770 + }, + { + "epoch": 0.1666766917293233, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.3724, + "step": 4771 + }, + { + "epoch": 0.16673684210526316, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2682, + "step": 4772 + }, + { + "epoch": 0.166796992481203, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2801, + "step": 4773 + }, + { + "epoch": 0.16685714285714287, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2587, + "step": 4774 + }, + { + "epoch": 0.16691729323308271, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3137, + "step": 4775 + }, + { + "epoch": 0.16697744360902256, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3102, + "step": 4776 + }, + { + "epoch": 0.1670375939849624, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2074, + "step": 4777 + }, + { + "epoch": 0.16709774436090225, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2191, + "step": 4778 + }, + { + "epoch": 0.16715789473684212, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 4779 + }, + { + "epoch": 0.16721804511278196, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2224, + "step": 4780 + }, + { + "epoch": 0.1672781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2975, + "step": 4781 + }, + { + "epoch": 0.16733834586466165, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3473, + "step": 4782 + }, + { + "epoch": 0.1673984962406015, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3266, + "step": 4783 + }, + { + "epoch": 0.16745864661654136, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2677, + "step": 4784 + }, + { + "epoch": 0.1675187969924812, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2858, + "step": 4785 + }, + { + "epoch": 0.16757894736842105, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.2479, + "step": 4786 + }, + { + "epoch": 0.1676390977443609, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.3456, + "step": 4787 + }, + { + "epoch": 0.16769924812030076, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3249, + "step": 4788 + }, + { + "epoch": 0.1677593984962406, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2254, + "step": 4789 + }, + { + "epoch": 0.16781954887218045, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.316, + "step": 4790 + }, + { + "epoch": 0.1678796992481203, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 4791 + }, + { + "epoch": 0.16793984962406014, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2119, + "step": 4792 + }, + { + "epoch": 0.168, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2611, + "step": 4793 + }, + { + "epoch": 0.16806015037593985, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2566, + "step": 4794 + }, + { + "epoch": 0.1681203007518797, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2843, + "step": 4795 + }, + { + "epoch": 0.16818045112781954, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3332, + "step": 4796 + }, + { + "epoch": 0.16824060150375938, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.367, + "step": 4797 + }, + { + "epoch": 0.16830075187969926, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2539, + "step": 4798 + }, + { + "epoch": 0.1683609022556391, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2472, + "step": 4799 + }, + { + "epoch": 0.16842105263157894, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2397, + "step": 4800 + }, + { + "epoch": 0.1684812030075188, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2986, + "step": 4801 + }, + { + "epoch": 0.16854135338345866, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.336, + "step": 4802 + }, + { + "epoch": 0.1686015037593985, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3041, + "step": 4803 + }, + { + "epoch": 0.16866165413533835, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2852, + "step": 4804 + }, + { + "epoch": 0.1687218045112782, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3023, + "step": 4805 + }, + { + "epoch": 0.16878195488721803, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.346, + "step": 4806 + }, + { + "epoch": 0.1688421052631579, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2905, + "step": 4807 + }, + { + "epoch": 0.16890225563909775, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2973, + "step": 4808 + }, + { + "epoch": 0.1689624060150376, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.4302, + "step": 4809 + }, + { + "epoch": 0.16902255639097744, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3347, + "step": 4810 + }, + { + "epoch": 0.1690827067669173, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2742, + "step": 4811 + }, + { + "epoch": 0.16914285714285715, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3921, + "step": 4812 + }, + { + "epoch": 0.169203007518797, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2563, + "step": 4813 + }, + { + "epoch": 0.16926315789473684, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3054, + "step": 4814 + }, + { + "epoch": 0.16932330827067668, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2397, + "step": 4815 + }, + { + "epoch": 0.16938345864661655, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.4082, + "step": 4816 + }, + { + "epoch": 0.1694436090225564, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 4817 + }, + { + "epoch": 0.16950375939849624, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2439, + "step": 4818 + }, + { + "epoch": 0.16956390977443608, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.335, + "step": 4819 + }, + { + "epoch": 0.16962406015037593, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.481, + "step": 4820 + }, + { + "epoch": 0.1696842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3622, + "step": 4821 + }, + { + "epoch": 0.16974436090225564, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2767, + "step": 4822 + }, + { + "epoch": 0.16980451127819549, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.289, + "step": 4823 + }, + { + "epoch": 0.16986466165413533, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.3227, + "step": 4824 + }, + { + "epoch": 0.1699248120300752, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2606, + "step": 4825 + }, + { + "epoch": 0.16998496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3108, + "step": 4826 + }, + { + "epoch": 0.1700451127819549, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2426, + "step": 4827 + }, + { + "epoch": 0.17010526315789473, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.223, + "step": 4828 + }, + { + "epoch": 0.17016541353383458, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2649, + "step": 4829 + }, + { + "epoch": 0.17022556390977445, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3981, + "step": 4830 + }, + { + "epoch": 0.1702857142857143, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2483, + "step": 4831 + }, + { + "epoch": 0.17034586466165413, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3642, + "step": 4832 + }, + { + "epoch": 0.17040601503759398, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3921, + "step": 4833 + }, + { + "epoch": 0.17046616541353382, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3151, + "step": 4834 + }, + { + "epoch": 0.1705263157894737, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2695, + "step": 4835 + }, + { + "epoch": 0.17058646616541354, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.272, + "step": 4836 + }, + { + "epoch": 0.17064661654135338, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3993, + "step": 4837 + }, + { + "epoch": 0.17070676691729322, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2602, + "step": 4838 + }, + { + "epoch": 0.1707669172932331, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3114, + "step": 4839 + }, + { + "epoch": 0.17082706766917294, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2603, + "step": 4840 + }, + { + "epoch": 0.17088721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2243, + "step": 4841 + }, + { + "epoch": 0.17094736842105263, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2681, + "step": 4842 + }, + { + "epoch": 0.17100751879699247, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2654, + "step": 4843 + }, + { + "epoch": 0.17106766917293234, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2892, + "step": 4844 + }, + { + "epoch": 0.17112781954887218, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2772, + "step": 4845 + }, + { + "epoch": 0.17118796992481203, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2794, + "step": 4846 + }, + { + "epoch": 0.17124812030075187, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.322, + "step": 4847 + }, + { + "epoch": 0.17130827067669174, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2663, + "step": 4848 + }, + { + "epoch": 0.1713684210526316, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2791, + "step": 4849 + }, + { + "epoch": 0.17142857142857143, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2517, + "step": 4850 + }, + { + "epoch": 0.17148872180451127, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.315, + "step": 4851 + }, + { + "epoch": 0.17154887218045112, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3057, + "step": 4852 + }, + { + "epoch": 0.171609022556391, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2252, + "step": 4853 + }, + { + "epoch": 0.17166917293233083, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.3024, + "step": 4854 + }, + { + "epoch": 0.17172932330827068, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1972, + "step": 4855 + }, + { + "epoch": 0.17178947368421052, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3357, + "step": 4856 + }, + { + "epoch": 0.17184962406015036, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3142, + "step": 4857 + }, + { + "epoch": 0.17190977443609023, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2765, + "step": 4858 + }, + { + "epoch": 0.17196992481203008, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3211, + "step": 4859 + }, + { + "epoch": 0.17203007518796992, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.4586, + "step": 4860 + }, + { + "epoch": 0.17209022556390977, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3033, + "step": 4861 + }, + { + "epoch": 0.17215037593984964, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 4862 + }, + { + "epoch": 0.17221052631578948, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.3163, + "step": 4863 + }, + { + "epoch": 0.17227067669172932, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3539, + "step": 4864 + }, + { + "epoch": 0.17233082706766917, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 4865 + }, + { + "epoch": 0.172390977443609, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.3047, + "step": 4866 + }, + { + "epoch": 0.17245112781954888, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2554, + "step": 4867 + }, + { + "epoch": 0.17251127819548873, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3638, + "step": 4868 + }, + { + "epoch": 0.17257142857142857, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3543, + "step": 4869 + }, + { + "epoch": 0.1726315789473684, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.321, + "step": 4870 + }, + { + "epoch": 0.17269172932330826, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3556, + "step": 4871 + }, + { + "epoch": 0.17275187969924813, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2587, + "step": 4872 + }, + { + "epoch": 0.17281203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3574, + "step": 4873 + }, + { + "epoch": 0.17287218045112782, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.4675, + "step": 4874 + }, + { + "epoch": 0.17293233082706766, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3364, + "step": 4875 + }, + { + "epoch": 0.17299248120300753, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.4007, + "step": 4876 + }, + { + "epoch": 0.17305263157894737, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2525, + "step": 4877 + }, + { + "epoch": 0.17311278195488722, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.4061, + "step": 4878 + }, + { + "epoch": 0.17317293233082706, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3286, + "step": 4879 + }, + { + "epoch": 0.1732330827067669, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3147, + "step": 4880 + }, + { + "epoch": 0.17329323308270678, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3365, + "step": 4881 + }, + { + "epoch": 0.17335338345864662, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3641, + "step": 4882 + }, + { + "epoch": 0.17341353383458646, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2107, + "step": 4883 + }, + { + "epoch": 0.1734736842105263, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2636, + "step": 4884 + }, + { + "epoch": 0.17353383458646618, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3937, + "step": 4885 + }, + { + "epoch": 0.17359398496240602, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.3804, + "step": 4886 + }, + { + "epoch": 0.17365413533834587, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.3277, + "step": 4887 + }, + { + "epoch": 0.1737142857142857, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2338, + "step": 4888 + }, + { + "epoch": 0.17377443609022555, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3411, + "step": 4889 + }, + { + "epoch": 0.17383458646616542, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 4890 + }, + { + "epoch": 0.17389473684210527, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2878, + "step": 4891 + }, + { + "epoch": 0.1739548872180451, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2174, + "step": 4892 + }, + { + "epoch": 0.17401503759398496, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2952, + "step": 4893 + }, + { + "epoch": 0.1740751879699248, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2713, + "step": 4894 + }, + { + "epoch": 0.17413533834586467, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3521, + "step": 4895 + }, + { + "epoch": 0.17419548872180451, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2808, + "step": 4896 + }, + { + "epoch": 0.17425563909774436, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3739, + "step": 4897 + }, + { + "epoch": 0.1743157894736842, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3193, + "step": 4898 + }, + { + "epoch": 0.17437593984962407, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3116, + "step": 4899 + }, + { + "epoch": 0.17443609022556392, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3323, + "step": 4900 + }, + { + "epoch": 0.17449624060150376, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2649, + "step": 4901 + }, + { + "epoch": 0.1745563909774436, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.306, + "step": 4902 + }, + { + "epoch": 0.17461654135338345, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2741, + "step": 4903 + }, + { + "epoch": 0.17467669172932332, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3069, + "step": 4904 + }, + { + "epoch": 0.17473684210526316, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2493, + "step": 4905 + }, + { + "epoch": 0.174796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2824, + "step": 4906 + }, + { + "epoch": 0.17485714285714285, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2663, + "step": 4907 + }, + { + "epoch": 0.1749172932330827, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2648, + "step": 4908 + }, + { + "epoch": 0.17497744360902256, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3309, + "step": 4909 + }, + { + "epoch": 0.1750375939849624, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2388, + "step": 4910 + }, + { + "epoch": 0.17509774436090225, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 4911 + }, + { + "epoch": 0.1751578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2648, + "step": 4912 + }, + { + "epoch": 0.17521804511278197, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3599, + "step": 4913 + }, + { + "epoch": 0.1752781954887218, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2361, + "step": 4914 + }, + { + "epoch": 0.17533834586466165, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3261, + "step": 4915 + }, + { + "epoch": 0.1753984962406015, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3876, + "step": 4916 + }, + { + "epoch": 0.17545864661654134, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2473, + "step": 4917 + }, + { + "epoch": 0.1755187969924812, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3879, + "step": 4918 + }, + { + "epoch": 0.17557894736842106, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2672, + "step": 4919 + }, + { + "epoch": 0.1756390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3323, + "step": 4920 + }, + { + "epoch": 0.17569924812030074, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2937, + "step": 4921 + }, + { + "epoch": 0.17575939849624062, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3084, + "step": 4922 + }, + { + "epoch": 0.17581954887218046, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2983, + "step": 4923 + }, + { + "epoch": 0.1758796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2975, + "step": 4924 + }, + { + "epoch": 0.17593984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3047, + "step": 4925 + }, + { + "epoch": 0.176, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 4926 + }, + { + "epoch": 0.17606015037593986, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3044, + "step": 4927 + }, + { + "epoch": 0.1761203007518797, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3148, + "step": 4928 + }, + { + "epoch": 0.17618045112781955, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2629, + "step": 4929 + }, + { + "epoch": 0.1762406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3549, + "step": 4930 + }, + { + "epoch": 0.17630075187969924, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2603, + "step": 4931 + }, + { + "epoch": 0.1763609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3009, + "step": 4932 + }, + { + "epoch": 0.17642105263157895, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2634, + "step": 4933 + }, + { + "epoch": 0.1764812030075188, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.312, + "step": 4934 + }, + { + "epoch": 0.17654135338345864, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.3618, + "step": 4935 + }, + { + "epoch": 0.1766015037593985, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 4936 + }, + { + "epoch": 0.17666165413533835, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 4937 + }, + { + "epoch": 0.1767218045112782, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3468, + "step": 4938 + }, + { + "epoch": 0.17678195488721804, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 4939 + }, + { + "epoch": 0.17684210526315788, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3181, + "step": 4940 + }, + { + "epoch": 0.17690225563909776, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2766, + "step": 4941 + }, + { + "epoch": 0.1769624060150376, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2932, + "step": 4942 + }, + { + "epoch": 0.17702255639097744, + "grad_norm": 0.419921875, + "learning_rate": 3e-05, + "loss": 1.3119, + "step": 4943 + }, + { + "epoch": 0.17708270676691729, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2863, + "step": 4944 + }, + { + "epoch": 0.17714285714285713, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3419, + "step": 4945 + }, + { + "epoch": 0.177203007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3146, + "step": 4946 + }, + { + "epoch": 0.17726315789473684, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2164, + "step": 4947 + }, + { + "epoch": 0.1773233082706767, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2869, + "step": 4948 + }, + { + "epoch": 0.17738345864661653, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 4949 + }, + { + "epoch": 0.1774436090225564, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3082, + "step": 4950 + }, + { + "epoch": 0.17750375939849625, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3148, + "step": 4951 + }, + { + "epoch": 0.1775639097744361, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2227, + "step": 4952 + }, + { + "epoch": 0.17762406015037593, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2513, + "step": 4953 + }, + { + "epoch": 0.17768421052631578, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2755, + "step": 4954 + }, + { + "epoch": 0.17774436090225565, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3059, + "step": 4955 + }, + { + "epoch": 0.1778045112781955, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2717, + "step": 4956 + }, + { + "epoch": 0.17786466165413534, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2561, + "step": 4957 + }, + { + "epoch": 0.17792481203007518, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3226, + "step": 4958 + }, + { + "epoch": 0.17798496240601502, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.2683, + "step": 4959 + }, + { + "epoch": 0.1780451127819549, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3555, + "step": 4960 + }, + { + "epoch": 0.17810526315789474, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3075, + "step": 4961 + }, + { + "epoch": 0.17816541353383458, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2375, + "step": 4962 + }, + { + "epoch": 0.17822556390977443, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3724, + "step": 4963 + }, + { + "epoch": 0.1782857142857143, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3033, + "step": 4964 + }, + { + "epoch": 0.17834586466165414, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.3825, + "step": 4965 + }, + { + "epoch": 0.17840601503759398, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3069, + "step": 4966 + }, + { + "epoch": 0.17846616541353383, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2519, + "step": 4967 + }, + { + "epoch": 0.17852631578947367, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3147, + "step": 4968 + }, + { + "epoch": 0.17858646616541354, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.271, + "step": 4969 + }, + { + "epoch": 0.1786466165413534, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2169, + "step": 4970 + }, + { + "epoch": 0.17870676691729323, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.2973, + "step": 4971 + }, + { + "epoch": 0.17876691729323307, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2707, + "step": 4972 + }, + { + "epoch": 0.17882706766917295, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2994, + "step": 4973 + }, + { + "epoch": 0.1788872180451128, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3213, + "step": 4974 + }, + { + "epoch": 0.17894736842105263, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.3806, + "step": 4975 + }, + { + "epoch": 0.17900751879699248, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3234, + "step": 4976 + }, + { + "epoch": 0.17906766917293232, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3052, + "step": 4977 + }, + { + "epoch": 0.1791278195488722, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2726, + "step": 4978 + }, + { + "epoch": 0.17918796992481204, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3763, + "step": 4979 + }, + { + "epoch": 0.17924812030075188, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2994, + "step": 4980 + }, + { + "epoch": 0.17930827067669172, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.287, + "step": 4981 + }, + { + "epoch": 0.17936842105263157, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2338, + "step": 4982 + }, + { + "epoch": 0.17942857142857144, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3061, + "step": 4983 + }, + { + "epoch": 0.17948872180451128, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.3588, + "step": 4984 + }, + { + "epoch": 0.17954887218045112, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3038, + "step": 4985 + }, + { + "epoch": 0.17960902255639097, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3886, + "step": 4986 + }, + { + "epoch": 0.17966917293233084, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3294, + "step": 4987 + }, + { + "epoch": 0.17972932330827068, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2528, + "step": 4988 + }, + { + "epoch": 0.17978947368421053, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2794, + "step": 4989 + }, + { + "epoch": 0.17984962406015037, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3858, + "step": 4990 + }, + { + "epoch": 0.17990977443609021, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2954, + "step": 4991 + }, + { + "epoch": 0.17996992481203009, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2653, + "step": 4992 + }, + { + "epoch": 0.18003007518796993, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3441, + "step": 4993 + }, + { + "epoch": 0.18009022556390977, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2738, + "step": 4994 + }, + { + "epoch": 0.18015037593984962, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2824, + "step": 4995 + }, + { + "epoch": 0.18021052631578946, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3581, + "step": 4996 + }, + { + "epoch": 0.18027067669172933, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.44, + "step": 4997 + }, + { + "epoch": 0.18033082706766917, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2677, + "step": 4998 + }, + { + "epoch": 0.18039097744360902, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2386, + "step": 4999 + }, + { + "epoch": 0.18045112781954886, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3076, + "step": 5000 + }, + { + "epoch": 6.015037593984962e-05, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2369, + "step": 5001 + }, + { + "epoch": 0.00012030075187969925, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2396, + "step": 5002 + }, + { + "epoch": 0.00018045112781954887, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2034, + "step": 5003 + }, + { + "epoch": 0.0002406015037593985, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3108, + "step": 5004 + }, + { + "epoch": 0.0003007518796992481, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3116, + "step": 5005 + }, + { + "epoch": 0.00036090225563909775, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3073, + "step": 5006 + }, + { + "epoch": 0.0004210526315789474, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2688, + "step": 5007 + }, + { + "epoch": 0.000481203007518797, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2811, + "step": 5008 + }, + { + "epoch": 0.0005413533834586466, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2913, + "step": 5009 + }, + { + "epoch": 0.0006015037593984962, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2786, + "step": 5010 + }, + { + "epoch": 0.0006616541353383459, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 5011 + }, + { + "epoch": 0.0007218045112781955, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.4077, + "step": 5012 + }, + { + "epoch": 0.0007819548872180451, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2759, + "step": 5013 + }, + { + "epoch": 0.0008421052631578948, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.3018, + "step": 5014 + }, + { + "epoch": 0.0009022556390977444, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2869, + "step": 5015 + }, + { + "epoch": 0.000962406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.4194, + "step": 5016 + }, + { + "epoch": 0.0010225563909774435, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2366, + "step": 5017 + }, + { + "epoch": 0.0010827067669172932, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2557, + "step": 5018 + }, + { + "epoch": 0.001142857142857143, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1831, + "step": 5019 + }, + { + "epoch": 0.0012030075187969924, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2911, + "step": 5020 + }, + { + "epoch": 0.0012631578947368421, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3142, + "step": 5021 + }, + { + "epoch": 0.0013233082706766918, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3809, + "step": 5022 + }, + { + "epoch": 0.0013834586466165413, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3786, + "step": 5023 + }, + { + "epoch": 0.001443609022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2251, + "step": 5024 + }, + { + "epoch": 0.0015037593984962407, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2949, + "step": 5025 + }, + { + "epoch": 0.0015639097744360902, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.24, + "step": 5026 + }, + { + "epoch": 0.0016240601503759399, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2851, + "step": 5027 + }, + { + "epoch": 0.0016842105263157896, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.3603, + "step": 5028 + }, + { + "epoch": 0.001744360902255639, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2303, + "step": 5029 + }, + { + "epoch": 0.0018045112781954887, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2787, + "step": 5030 + }, + { + "epoch": 0.0018646616541353384, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 5031 + }, + { + "epoch": 0.001924812030075188, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2482, + "step": 5032 + }, + { + "epoch": 0.0019849624060150374, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2865, + "step": 5033 + }, + { + "epoch": 0.002045112781954887, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2477, + "step": 5034 + }, + { + "epoch": 0.002105263157894737, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2761, + "step": 5035 + }, + { + "epoch": 0.0021654135338345865, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3121, + "step": 5036 + }, + { + "epoch": 0.002225563909774436, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2777, + "step": 5037 + }, + { + "epoch": 0.002285714285714286, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.3, + "step": 5038 + }, + { + "epoch": 0.002345864661654135, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3852, + "step": 5039 + }, + { + "epoch": 0.002406015037593985, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3749, + "step": 5040 + }, + { + "epoch": 0.0024661654135338345, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3223, + "step": 5041 + }, + { + "epoch": 0.0025263157894736842, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.3356, + "step": 5042 + }, + { + "epoch": 0.002586466165413534, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2375, + "step": 5043 + }, + { + "epoch": 0.0026466165413533836, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2987, + "step": 5044 + }, + { + "epoch": 0.002706766917293233, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3723, + "step": 5045 + }, + { + "epoch": 0.0027669172932330826, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2484, + "step": 5046 + }, + { + "epoch": 0.0028270676691729323, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3728, + "step": 5047 + }, + { + "epoch": 0.002887218045112782, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 5048 + }, + { + "epoch": 0.0029473684210526317, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 5049 + }, + { + "epoch": 0.0030075187969924814, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2903, + "step": 5050 + }, + { + "epoch": 0.0030676691729323306, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3091, + "step": 5051 + }, + { + "epoch": 0.0031278195488721803, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3445, + "step": 5052 + }, + { + "epoch": 0.00318796992481203, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3035, + "step": 5053 + }, + { + "epoch": 0.0032481203007518797, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3122, + "step": 5054 + }, + { + "epoch": 0.0033082706766917294, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.4026, + "step": 5055 + }, + { + "epoch": 0.003368421052631579, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 5056 + }, + { + "epoch": 0.0034285714285714284, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.284, + "step": 5057 + }, + { + "epoch": 0.003488721804511278, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2941, + "step": 5058 + }, + { + "epoch": 0.003548872180451128, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2898, + "step": 5059 + }, + { + "epoch": 0.0036090225563909775, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3234, + "step": 5060 + }, + { + "epoch": 0.003669172932330827, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2818, + "step": 5061 + }, + { + "epoch": 0.003729323308270677, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3793, + "step": 5062 + }, + { + "epoch": 0.003789473684210526, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2981, + "step": 5063 + }, + { + "epoch": 0.003849624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3633, + "step": 5064 + }, + { + "epoch": 0.003909774436090226, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2489, + "step": 5065 + }, + { + "epoch": 0.003969924812030075, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.4296, + "step": 5066 + }, + { + "epoch": 0.0040300751879699245, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 5067 + }, + { + "epoch": 0.004090225563909774, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2566, + "step": 5068 + }, + { + "epoch": 0.004150375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.292, + "step": 5069 + }, + { + "epoch": 0.004210526315789474, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2459, + "step": 5070 + }, + { + "epoch": 0.004270676691729323, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 5071 + }, + { + "epoch": 0.004330827067669173, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 5072 + }, + { + "epoch": 0.004390977443609023, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.4077, + "step": 5073 + }, + { + "epoch": 0.004451127819548872, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.249, + "step": 5074 + }, + { + "epoch": 0.004511278195488722, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2113, + "step": 5075 + }, + { + "epoch": 0.004571428571428572, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.285, + "step": 5076 + }, + { + "epoch": 0.0046315789473684215, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3968, + "step": 5077 + }, + { + "epoch": 0.00469172932330827, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3581, + "step": 5078 + }, + { + "epoch": 0.00475187969924812, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2703, + "step": 5079 + }, + { + "epoch": 0.00481203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2959, + "step": 5080 + }, + { + "epoch": 0.004872180451127819, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2348, + "step": 5081 + }, + { + "epoch": 0.004932330827067669, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2905, + "step": 5082 + }, + { + "epoch": 0.004992481203007519, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 5083 + }, + { + "epoch": 0.0050526315789473685, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2583, + "step": 5084 + }, + { + "epoch": 0.005112781954887218, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.3165, + "step": 5085 + }, + { + "epoch": 0.005172932330827068, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.284, + "step": 5086 + }, + { + "epoch": 0.005233082706766918, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.206, + "step": 5087 + }, + { + "epoch": 0.005293233082706767, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3363, + "step": 5088 + }, + { + "epoch": 0.005353383458646617, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 5089 + }, + { + "epoch": 0.005413533834586466, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2889, + "step": 5090 + }, + { + "epoch": 0.0054736842105263155, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3009, + "step": 5091 + }, + { + "epoch": 0.005533834586466165, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3321, + "step": 5092 + }, + { + "epoch": 0.005593984962406015, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2584, + "step": 5093 + }, + { + "epoch": 0.005654135338345865, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3696, + "step": 5094 + }, + { + "epoch": 0.005714285714285714, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2017, + "step": 5095 + }, + { + "epoch": 0.005774436090225564, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2825, + "step": 5096 + }, + { + "epoch": 0.005834586466165414, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2753, + "step": 5097 + }, + { + "epoch": 0.005894736842105263, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3085, + "step": 5098 + }, + { + "epoch": 0.005954887218045113, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2399, + "step": 5099 + }, + { + "epoch": 0.006015037593984963, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2868, + "step": 5100 + }, + { + "epoch": 0.0060751879699248125, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 5101 + }, + { + "epoch": 0.006135338345864661, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2915, + "step": 5102 + }, + { + "epoch": 0.006195488721804511, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1808, + "step": 5103 + }, + { + "epoch": 0.006255639097744361, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.2527, + "step": 5104 + }, + { + "epoch": 0.00631578947368421, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 5105 + }, + { + "epoch": 0.00637593984962406, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3007, + "step": 5106 + }, + { + "epoch": 0.00643609022556391, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.203, + "step": 5107 + }, + { + "epoch": 0.0064962406015037595, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.355, + "step": 5108 + }, + { + "epoch": 0.006556390977443609, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3029, + "step": 5109 + }, + { + "epoch": 0.006616541353383459, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2456, + "step": 5110 + }, + { + "epoch": 0.0066766917293233086, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3556, + "step": 5111 + }, + { + "epoch": 0.006736842105263158, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3379, + "step": 5112 + }, + { + "epoch": 0.006796992481203007, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2518, + "step": 5113 + }, + { + "epoch": 0.006857142857142857, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 5114 + }, + { + "epoch": 0.0069172932330827065, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.331, + "step": 5115 + }, + { + "epoch": 0.006977443609022556, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2706, + "step": 5116 + }, + { + "epoch": 0.007037593984962406, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.4292, + "step": 5117 + }, + { + "epoch": 0.007097744360902256, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2061, + "step": 5118 + }, + { + "epoch": 0.007157894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2127, + "step": 5119 + }, + { + "epoch": 0.007218045112781955, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2789, + "step": 5120 + }, + { + "epoch": 0.007278195488721805, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 5121 + }, + { + "epoch": 0.007338345864661654, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2623, + "step": 5122 + }, + { + "epoch": 0.007398496240601504, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3339, + "step": 5123 + }, + { + "epoch": 0.007458646616541354, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2851, + "step": 5124 + }, + { + "epoch": 0.007518796992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 5125 + }, + { + "epoch": 0.007578947368421052, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2735, + "step": 5126 + }, + { + "epoch": 0.007639097744360902, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 5127 + }, + { + "epoch": 0.007699248120300752, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2417, + "step": 5128 + }, + { + "epoch": 0.007759398496240601, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 5129 + }, + { + "epoch": 0.007819548872180452, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 5130 + }, + { + "epoch": 0.0078796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3302, + "step": 5131 + }, + { + "epoch": 0.00793984962406015, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3479, + "step": 5132 + }, + { + "epoch": 0.008, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2892, + "step": 5133 + }, + { + "epoch": 0.008060150375939849, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2477, + "step": 5134 + }, + { + "epoch": 0.0081203007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2922, + "step": 5135 + }, + { + "epoch": 0.008180451127819548, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 5136 + }, + { + "epoch": 0.008240601503759399, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.295, + "step": 5137 + }, + { + "epoch": 0.008300751879699248, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3332, + "step": 5138 + }, + { + "epoch": 0.008360902255639098, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2518, + "step": 5139 + }, + { + "epoch": 0.008421052631578947, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2712, + "step": 5140 + }, + { + "epoch": 0.008481203007518798, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3884, + "step": 5141 + }, + { + "epoch": 0.008541353383458647, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.2737, + "step": 5142 + }, + { + "epoch": 0.008601503759398495, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3629, + "step": 5143 + }, + { + "epoch": 0.008661654135338346, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2829, + "step": 5144 + }, + { + "epoch": 0.008721804511278195, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3137, + "step": 5145 + }, + { + "epoch": 0.008781954887218045, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.328, + "step": 5146 + }, + { + "epoch": 0.008842105263157894, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2934, + "step": 5147 + }, + { + "epoch": 0.008902255639097745, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 5148 + }, + { + "epoch": 0.008962406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.287, + "step": 5149 + }, + { + "epoch": 0.009022556390977444, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2929, + "step": 5150 + }, + { + "epoch": 0.009082706766917293, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2893, + "step": 5151 + }, + { + "epoch": 0.009142857142857144, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.3367, + "step": 5152 + }, + { + "epoch": 0.009203007518796992, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2634, + "step": 5153 + }, + { + "epoch": 0.009263157894736843, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3283, + "step": 5154 + }, + { + "epoch": 0.009323308270676692, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.3264, + "step": 5155 + }, + { + "epoch": 0.00938345864661654, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3249, + "step": 5156 + }, + { + "epoch": 0.009443609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2699, + "step": 5157 + }, + { + "epoch": 0.00950375939849624, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2323, + "step": 5158 + }, + { + "epoch": 0.00956390977443609, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3488, + "step": 5159 + }, + { + "epoch": 0.00962406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.319, + "step": 5160 + }, + { + "epoch": 0.00968421052631579, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.314, + "step": 5161 + }, + { + "epoch": 0.009744360902255639, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3728, + "step": 5162 + }, + { + "epoch": 0.00980451127819549, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3401, + "step": 5163 + }, + { + "epoch": 0.009864661654135338, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3777, + "step": 5164 + }, + { + "epoch": 0.009924812030075189, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.312, + "step": 5165 + }, + { + "epoch": 0.009984962406015038, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3145, + "step": 5166 + }, + { + "epoch": 0.010045112781954886, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3203, + "step": 5167 + }, + { + "epoch": 0.010105263157894737, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3193, + "step": 5168 + }, + { + "epoch": 0.010165413533834586, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2897, + "step": 5169 + }, + { + "epoch": 0.010225563909774436, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2484, + "step": 5170 + }, + { + "epoch": 0.010285714285714285, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2718, + "step": 5171 + }, + { + "epoch": 0.010345864661654136, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3133, + "step": 5172 + }, + { + "epoch": 0.010406015037593985, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1843, + "step": 5173 + }, + { + "epoch": 0.010466165413533835, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3337, + "step": 5174 + }, + { + "epoch": 0.010526315789473684, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2877, + "step": 5175 + }, + { + "epoch": 0.010586466165413535, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3853, + "step": 5176 + }, + { + "epoch": 0.010646616541353383, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.285, + "step": 5177 + }, + { + "epoch": 0.010706766917293234, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.235, + "step": 5178 + }, + { + "epoch": 0.010766917293233083, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3419, + "step": 5179 + }, + { + "epoch": 0.010827067669172932, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1956, + "step": 5180 + }, + { + "epoch": 0.010887218045112782, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2873, + "step": 5181 + }, + { + "epoch": 0.010947368421052631, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2312, + "step": 5182 + }, + { + "epoch": 0.011007518796992482, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 5183 + }, + { + "epoch": 0.01106766917293233, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2431, + "step": 5184 + }, + { + "epoch": 0.011127819548872181, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3169, + "step": 5185 + }, + { + "epoch": 0.01118796992481203, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3523, + "step": 5186 + }, + { + "epoch": 0.01124812030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2458, + "step": 5187 + }, + { + "epoch": 0.01130827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2386, + "step": 5188 + }, + { + "epoch": 0.01136842105263158, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2815, + "step": 5189 + }, + { + "epoch": 0.011428571428571429, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2611, + "step": 5190 + }, + { + "epoch": 0.011488721804511277, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2826, + "step": 5191 + }, + { + "epoch": 0.011548872180451128, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2171, + "step": 5192 + }, + { + "epoch": 0.011609022556390977, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2493, + "step": 5193 + }, + { + "epoch": 0.011669172932330827, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2402, + "step": 5194 + }, + { + "epoch": 0.011729323308270676, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2691, + "step": 5195 + }, + { + "epoch": 0.011789473684210527, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2597, + "step": 5196 + }, + { + "epoch": 0.011849624060150376, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.358, + "step": 5197 + }, + { + "epoch": 0.011909774436090226, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.386, + "step": 5198 + }, + { + "epoch": 0.011969924812030075, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.218, + "step": 5199 + }, + { + "epoch": 0.012030075187969926, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1937, + "step": 5200 + }, + { + "epoch": 0.012090225563909774, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3189, + "step": 5201 + }, + { + "epoch": 0.012150375939849625, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2914, + "step": 5202 + }, + { + "epoch": 0.012210526315789474, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2754, + "step": 5203 + }, + { + "epoch": 0.012270676691729323, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2521, + "step": 5204 + }, + { + "epoch": 0.012330827067669173, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 5205 + }, + { + "epoch": 0.012390977443609022, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2703, + "step": 5206 + }, + { + "epoch": 0.012451127819548873, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2742, + "step": 5207 + }, + { + "epoch": 0.012511278195488721, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3045, + "step": 5208 + }, + { + "epoch": 0.012571428571428572, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2348, + "step": 5209 + }, + { + "epoch": 0.01263157894736842, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2641, + "step": 5210 + }, + { + "epoch": 0.012691729323308271, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2289, + "step": 5211 + }, + { + "epoch": 0.01275187969924812, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3687, + "step": 5212 + }, + { + "epoch": 0.01281203007518797, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 5213 + }, + { + "epoch": 0.01287218045112782, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3887, + "step": 5214 + }, + { + "epoch": 0.012932330827067668, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2818, + "step": 5215 + }, + { + "epoch": 0.012992481203007519, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4224, + "step": 5216 + }, + { + "epoch": 0.013052631578947368, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2087, + "step": 5217 + }, + { + "epoch": 0.013112781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2334, + "step": 5218 + }, + { + "epoch": 0.013172932330827067, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3451, + "step": 5219 + }, + { + "epoch": 0.013233082706766918, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.2496, + "step": 5220 + }, + { + "epoch": 0.013293233082706767, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2689, + "step": 5221 + }, + { + "epoch": 0.013353383458646617, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2498, + "step": 5222 + }, + { + "epoch": 0.013413533834586466, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2132, + "step": 5223 + }, + { + "epoch": 0.013473684210526317, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 5224 + }, + { + "epoch": 0.013533834586466165, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2993, + "step": 5225 + }, + { + "epoch": 0.013593984962406014, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2181, + "step": 5226 + }, + { + "epoch": 0.013654135338345865, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2046, + "step": 5227 + }, + { + "epoch": 0.013714285714285714, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2605, + "step": 5228 + }, + { + "epoch": 0.013774436090225564, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3105, + "step": 5229 + }, + { + "epoch": 0.013834586466165413, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2729, + "step": 5230 + }, + { + "epoch": 0.013894736842105264, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2631, + "step": 5231 + }, + { + "epoch": 0.013954887218045112, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3189, + "step": 5232 + }, + { + "epoch": 0.014015037593984963, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2726, + "step": 5233 + }, + { + "epoch": 0.014075187969924812, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2592, + "step": 5234 + }, + { + "epoch": 0.014135338345864662, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3078, + "step": 5235 + }, + { + "epoch": 0.014195488721804511, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 5236 + }, + { + "epoch": 0.014255639097744362, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.256, + "step": 5237 + }, + { + "epoch": 0.01431578947368421, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3013, + "step": 5238 + }, + { + "epoch": 0.01437593984962406, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3429, + "step": 5239 + }, + { + "epoch": 0.01443609022556391, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.368, + "step": 5240 + }, + { + "epoch": 0.014496240601503759, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3314, + "step": 5241 + }, + { + "epoch": 0.01455639097744361, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3595, + "step": 5242 + }, + { + "epoch": 0.014616541353383458, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2659, + "step": 5243 + }, + { + "epoch": 0.014676691729323309, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2869, + "step": 5244 + }, + { + "epoch": 0.014736842105263158, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.234, + "step": 5245 + }, + { + "epoch": 0.014796992481203008, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.285, + "step": 5246 + }, + { + "epoch": 0.014857142857142857, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3433, + "step": 5247 + }, + { + "epoch": 0.014917293233082708, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2142, + "step": 5248 + }, + { + "epoch": 0.014977443609022556, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2477, + "step": 5249 + }, + { + "epoch": 0.015037593984962405, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 5250 + }, + { + "epoch": 0.015097744360902256, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2898, + "step": 5251 + }, + { + "epoch": 0.015157894736842105, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3229, + "step": 5252 + }, + { + "epoch": 0.015218045112781955, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2397, + "step": 5253 + }, + { + "epoch": 0.015278195488721804, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3262, + "step": 5254 + }, + { + "epoch": 0.015338345864661655, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2841, + "step": 5255 + }, + { + "epoch": 0.015398496240601503, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.22, + "step": 5256 + }, + { + "epoch": 0.015458646616541354, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3113, + "step": 5257 + }, + { + "epoch": 0.015518796992481203, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3232, + "step": 5258 + }, + { + "epoch": 0.015578947368421053, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2214, + "step": 5259 + }, + { + "epoch": 0.015639097744360904, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3291, + "step": 5260 + }, + { + "epoch": 0.01569924812030075, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3208, + "step": 5261 + }, + { + "epoch": 0.0157593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3453, + "step": 5262 + }, + { + "epoch": 0.015819548872180452, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2817, + "step": 5263 + }, + { + "epoch": 0.0158796992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2703, + "step": 5264 + }, + { + "epoch": 0.01593984962406015, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2524, + "step": 5265 + }, + { + "epoch": 0.016, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2323, + "step": 5266 + }, + { + "epoch": 0.01606015037593985, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3852, + "step": 5267 + }, + { + "epoch": 0.016120300751879698, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3562, + "step": 5268 + }, + { + "epoch": 0.01618045112781955, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2577, + "step": 5269 + }, + { + "epoch": 0.0162406015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 5270 + }, + { + "epoch": 0.01630075187969925, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3139, + "step": 5271 + }, + { + "epoch": 0.016360902255639097, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2479, + "step": 5272 + }, + { + "epoch": 0.016421052631578947, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 5273 + }, + { + "epoch": 0.016481203007518798, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2182, + "step": 5274 + }, + { + "epoch": 0.016541353383458645, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2864, + "step": 5275 + }, + { + "epoch": 0.016601503759398496, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.3018, + "step": 5276 + }, + { + "epoch": 0.016661654135338346, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2441, + "step": 5277 + }, + { + "epoch": 0.016721804511278197, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2849, + "step": 5278 + }, + { + "epoch": 0.016781954887218044, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2459, + "step": 5279 + }, + { + "epoch": 0.016842105263157894, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3479, + "step": 5280 + }, + { + "epoch": 0.016902255639097745, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.36, + "step": 5281 + }, + { + "epoch": 0.016962406015037595, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2186, + "step": 5282 + }, + { + "epoch": 0.017022556390977443, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.316, + "step": 5283 + }, + { + "epoch": 0.017082706766917293, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.1881, + "step": 5284 + }, + { + "epoch": 0.017142857142857144, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3232, + "step": 5285 + }, + { + "epoch": 0.01720300751879699, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2641, + "step": 5286 + }, + { + "epoch": 0.01726315789473684, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2467, + "step": 5287 + }, + { + "epoch": 0.017323308270676692, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.2486, + "step": 5288 + }, + { + "epoch": 0.017383458646616542, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2831, + "step": 5289 + }, + { + "epoch": 0.01744360902255639, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.3517, + "step": 5290 + }, + { + "epoch": 0.01750375939849624, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2787, + "step": 5291 + }, + { + "epoch": 0.01756390977443609, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2434, + "step": 5292 + }, + { + "epoch": 0.01762406015037594, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3433, + "step": 5293 + }, + { + "epoch": 0.01768421052631579, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1876, + "step": 5294 + }, + { + "epoch": 0.01774436090225564, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.2964, + "step": 5295 + }, + { + "epoch": 0.01780451127819549, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2544, + "step": 5296 + }, + { + "epoch": 0.01786466165413534, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2361, + "step": 5297 + }, + { + "epoch": 0.017924812030075187, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3446, + "step": 5298 + }, + { + "epoch": 0.017984962406015038, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3267, + "step": 5299 + }, + { + "epoch": 0.01804511278195489, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 5300 + }, + { + "epoch": 0.018105263157894735, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2338, + "step": 5301 + }, + { + "epoch": 0.018165413533834586, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2724, + "step": 5302 + }, + { + "epoch": 0.018225563909774437, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 5303 + }, + { + "epoch": 0.018285714285714287, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 5304 + }, + { + "epoch": 0.018345864661654134, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3656, + "step": 5305 + }, + { + "epoch": 0.018406015037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2547, + "step": 5306 + }, + { + "epoch": 0.018466165413533835, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2302, + "step": 5307 + }, + { + "epoch": 0.018526315789473686, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 5308 + }, + { + "epoch": 0.018586466165413533, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2332, + "step": 5309 + }, + { + "epoch": 0.018646616541353384, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2876, + "step": 5310 + }, + { + "epoch": 0.018706766917293234, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2795, + "step": 5311 + }, + { + "epoch": 0.01876691729323308, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2677, + "step": 5312 + }, + { + "epoch": 0.018827067669172932, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 5313 + }, + { + "epoch": 0.018887218045112782, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2886, + "step": 5314 + }, + { + "epoch": 0.018947368421052633, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.3, + "step": 5315 + }, + { + "epoch": 0.01900751879699248, + "grad_norm": 0.453125, + "learning_rate": 3e-05, + "loss": 1.2412, + "step": 5316 + }, + { + "epoch": 0.01906766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2267, + "step": 5317 + }, + { + "epoch": 0.01912781954887218, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3279, + "step": 5318 + }, + { + "epoch": 0.01918796992481203, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2812, + "step": 5319 + }, + { + "epoch": 0.01924812030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2893, + "step": 5320 + }, + { + "epoch": 0.01930827067669173, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2544, + "step": 5321 + }, + { + "epoch": 0.01936842105263158, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.262, + "step": 5322 + }, + { + "epoch": 0.019428571428571427, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2579, + "step": 5323 + }, + { + "epoch": 0.019488721804511278, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2753, + "step": 5324 + }, + { + "epoch": 0.019548872180451128, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3862, + "step": 5325 + }, + { + "epoch": 0.01960902255639098, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2138, + "step": 5326 + }, + { + "epoch": 0.019669172932330826, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2278, + "step": 5327 + }, + { + "epoch": 0.019729323308270676, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1817, + "step": 5328 + }, + { + "epoch": 0.019789473684210527, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2726, + "step": 5329 + }, + { + "epoch": 0.019849624060150377, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3265, + "step": 5330 + }, + { + "epoch": 0.019909774436090225, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 5331 + }, + { + "epoch": 0.019969924812030075, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2153, + "step": 5332 + }, + { + "epoch": 0.020030075187969926, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3049, + "step": 5333 + }, + { + "epoch": 0.020090225563909773, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2851, + "step": 5334 + }, + { + "epoch": 0.020150375939849623, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2108, + "step": 5335 + }, + { + "epoch": 0.020210526315789474, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2153, + "step": 5336 + }, + { + "epoch": 0.020270676691729324, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2891, + "step": 5337 + }, + { + "epoch": 0.02033082706766917, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2979, + "step": 5338 + }, + { + "epoch": 0.020390977443609022, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3311, + "step": 5339 + }, + { + "epoch": 0.020451127819548873, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.3342, + "step": 5340 + }, + { + "epoch": 0.020511278195488723, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2443, + "step": 5341 + }, + { + "epoch": 0.02057142857142857, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2685, + "step": 5342 + }, + { + "epoch": 0.02063157894736842, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3648, + "step": 5343 + }, + { + "epoch": 0.02069172932330827, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2265, + "step": 5344 + }, + { + "epoch": 0.02075187969924812, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.3218, + "step": 5345 + }, + { + "epoch": 0.02081203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3117, + "step": 5346 + }, + { + "epoch": 0.02087218045112782, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2291, + "step": 5347 + }, + { + "epoch": 0.02093233082706767, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2627, + "step": 5348 + }, + { + "epoch": 0.020992481203007517, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3107, + "step": 5349 + }, + { + "epoch": 0.021052631578947368, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2764, + "step": 5350 + }, + { + "epoch": 0.02111278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2021, + "step": 5351 + }, + { + "epoch": 0.02117293233082707, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2802, + "step": 5352 + }, + { + "epoch": 0.021233082706766916, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3644, + "step": 5353 + }, + { + "epoch": 0.021293233082706767, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3016, + "step": 5354 + }, + { + "epoch": 0.021353383458646617, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2052, + "step": 5355 + }, + { + "epoch": 0.021413533834586468, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3, + "step": 5356 + }, + { + "epoch": 0.021473684210526315, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2669, + "step": 5357 + }, + { + "epoch": 0.021533834586466166, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2591, + "step": 5358 + }, + { + "epoch": 0.021593984962406016, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.2581, + "step": 5359 + }, + { + "epoch": 0.021654135338345863, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.3155, + "step": 5360 + }, + { + "epoch": 0.021714285714285714, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3063, + "step": 5361 + }, + { + "epoch": 0.021774436090225564, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2872, + "step": 5362 + }, + { + "epoch": 0.021834586466165415, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2767, + "step": 5363 + }, + { + "epoch": 0.021894736842105262, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2354, + "step": 5364 + }, + { + "epoch": 0.021954887218045113, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2202, + "step": 5365 + }, + { + "epoch": 0.022015037593984963, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.3086, + "step": 5366 + }, + { + "epoch": 0.022075187969924814, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.3621, + "step": 5367 + }, + { + "epoch": 0.02213533834586466, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.3495, + "step": 5368 + }, + { + "epoch": 0.02219548872180451, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2591, + "step": 5369 + }, + { + "epoch": 0.022255639097744362, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2692, + "step": 5370 + }, + { + "epoch": 0.02231578947368421, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 5371 + }, + { + "epoch": 0.02237593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.192, + "step": 5372 + }, + { + "epoch": 0.02243609022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2616, + "step": 5373 + }, + { + "epoch": 0.02249624060150376, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.248, + "step": 5374 + }, + { + "epoch": 0.022556390977443608, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2348, + "step": 5375 + }, + { + "epoch": 0.02261654135338346, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2535, + "step": 5376 + }, + { + "epoch": 0.02267669172932331, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2246, + "step": 5377 + }, + { + "epoch": 0.02273684210526316, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2313, + "step": 5378 + }, + { + "epoch": 0.022796992481203007, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2771, + "step": 5379 + }, + { + "epoch": 0.022857142857142857, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2847, + "step": 5380 + }, + { + "epoch": 0.022917293233082708, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2223, + "step": 5381 + }, + { + "epoch": 0.022977443609022555, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2555, + "step": 5382 + }, + { + "epoch": 0.023037593984962405, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3291, + "step": 5383 + }, + { + "epoch": 0.023097744360902256, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2558, + "step": 5384 + }, + { + "epoch": 0.023157894736842106, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.3787, + "step": 5385 + }, + { + "epoch": 0.023218045112781954, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2391, + "step": 5386 + }, + { + "epoch": 0.023278195488721804, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2757, + "step": 5387 + }, + { + "epoch": 0.023338345864661655, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 5388 + }, + { + "epoch": 0.023398496240601505, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2924, + "step": 5389 + }, + { + "epoch": 0.023458646616541352, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2451, + "step": 5390 + }, + { + "epoch": 0.023518796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1932, + "step": 5391 + }, + { + "epoch": 0.023578947368421053, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 5392 + }, + { + "epoch": 0.0236390977443609, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3238, + "step": 5393 + }, + { + "epoch": 0.02369924812030075, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2805, + "step": 5394 + }, + { + "epoch": 0.0237593984962406, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3032, + "step": 5395 + }, + { + "epoch": 0.023819548872180452, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3775, + "step": 5396 + }, + { + "epoch": 0.0238796992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2859, + "step": 5397 + }, + { + "epoch": 0.02393984962406015, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.237, + "step": 5398 + }, + { + "epoch": 0.024, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2379, + "step": 5399 + }, + { + "epoch": 0.02406015037593985, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3257, + "step": 5400 + }, + { + "epoch": 0.024120300751879698, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2723, + "step": 5401 + }, + { + "epoch": 0.02418045112781955, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2298, + "step": 5402 + }, + { + "epoch": 0.0242406015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2815, + "step": 5403 + }, + { + "epoch": 0.02430075187969925, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2919, + "step": 5404 + }, + { + "epoch": 0.024360902255639097, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2153, + "step": 5405 + }, + { + "epoch": 0.024421052631578948, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2673, + "step": 5406 + }, + { + "epoch": 0.024481203007518798, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3308, + "step": 5407 + }, + { + "epoch": 0.024541353383458645, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2512, + "step": 5408 + }, + { + "epoch": 0.024601503759398496, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.2676, + "step": 5409 + }, + { + "epoch": 0.024661654135338346, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.2814, + "step": 5410 + }, + { + "epoch": 0.024721804511278197, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 5411 + }, + { + "epoch": 0.024781954887218044, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3192, + "step": 5412 + }, + { + "epoch": 0.024842105263157895, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2321, + "step": 5413 + }, + { + "epoch": 0.024902255639097745, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2595, + "step": 5414 + }, + { + "epoch": 0.024962406015037596, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.4276, + "step": 5415 + }, + { + "epoch": 0.025022556390977443, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3172, + "step": 5416 + }, + { + "epoch": 0.025082706766917293, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2813, + "step": 5417 + }, + { + "epoch": 0.025142857142857144, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3163, + "step": 5418 + }, + { + "epoch": 0.02520300751879699, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.277, + "step": 5419 + }, + { + "epoch": 0.02526315789473684, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3055, + "step": 5420 + }, + { + "epoch": 0.025323308270676692, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 5421 + }, + { + "epoch": 0.025383458646616543, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2652, + "step": 5422 + }, + { + "epoch": 0.02544360902255639, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2554, + "step": 5423 + }, + { + "epoch": 0.02550375939849624, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.2915, + "step": 5424 + }, + { + "epoch": 0.02556390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3075, + "step": 5425 + }, + { + "epoch": 0.02562406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2791, + "step": 5426 + }, + { + "epoch": 0.02568421052631579, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3012, + "step": 5427 + }, + { + "epoch": 0.02574436090225564, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2811, + "step": 5428 + }, + { + "epoch": 0.02580451127819549, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2151, + "step": 5429 + }, + { + "epoch": 0.025864661654135337, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2733, + "step": 5430 + }, + { + "epoch": 0.025924812030075187, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 5431 + }, + { + "epoch": 0.025984962406015038, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2613, + "step": 5432 + }, + { + "epoch": 0.02604511278195489, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2946, + "step": 5433 + }, + { + "epoch": 0.026105263157894736, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 5434 + }, + { + "epoch": 0.026165413533834586, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3508, + "step": 5435 + }, + { + "epoch": 0.026225563909774437, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.192, + "step": 5436 + }, + { + "epoch": 0.026285714285714287, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2793, + "step": 5437 + }, + { + "epoch": 0.026345864661654134, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2491, + "step": 5438 + }, + { + "epoch": 0.026406015037593985, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.266, + "step": 5439 + }, + { + "epoch": 0.026466165413533835, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2852, + "step": 5440 + }, + { + "epoch": 0.026526315789473683, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.286, + "step": 5441 + }, + { + "epoch": 0.026586466165413533, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3701, + "step": 5442 + }, + { + "epoch": 0.026646616541353384, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2423, + "step": 5443 + }, + { + "epoch": 0.026706766917293234, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.354, + "step": 5444 + }, + { + "epoch": 0.02676691729323308, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2644, + "step": 5445 + }, + { + "epoch": 0.026827067669172932, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2601, + "step": 5446 + }, + { + "epoch": 0.026887218045112782, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2577, + "step": 5447 + }, + { + "epoch": 0.026947368421052633, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2572, + "step": 5448 + }, + { + "epoch": 0.02700751879699248, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1707, + "step": 5449 + }, + { + "epoch": 0.02706766917293233, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3757, + "step": 5450 + }, + { + "epoch": 0.02712781954887218, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3101, + "step": 5451 + }, + { + "epoch": 0.02718796992481203, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2666, + "step": 5452 + }, + { + "epoch": 0.02724812030075188, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.2232, + "step": 5453 + }, + { + "epoch": 0.02730827067669173, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2718, + "step": 5454 + }, + { + "epoch": 0.02736842105263158, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2584, + "step": 5455 + }, + { + "epoch": 0.027428571428571427, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 5456 + }, + { + "epoch": 0.027488721804511278, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2237, + "step": 5457 + }, + { + "epoch": 0.02754887218045113, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2704, + "step": 5458 + }, + { + "epoch": 0.02760902255639098, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.298, + "step": 5459 + }, + { + "epoch": 0.027669172932330826, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3165, + "step": 5460 + }, + { + "epoch": 0.027729323308270677, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2459, + "step": 5461 + }, + { + "epoch": 0.027789473684210527, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.32, + "step": 5462 + }, + { + "epoch": 0.027849624060150378, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.304, + "step": 5463 + }, + { + "epoch": 0.027909774436090225, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3066, + "step": 5464 + }, + { + "epoch": 0.027969924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 5465 + }, + { + "epoch": 0.028030075187969926, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3912, + "step": 5466 + }, + { + "epoch": 0.028090225563909773, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2341, + "step": 5467 + }, + { + "epoch": 0.028150375939849624, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2513, + "step": 5468 + }, + { + "epoch": 0.028210526315789474, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.3333, + "step": 5469 + }, + { + "epoch": 0.028270676691729325, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2807, + "step": 5470 + }, + { + "epoch": 0.028330827067669172, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2928, + "step": 5471 + }, + { + "epoch": 0.028390977443609022, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2203, + "step": 5472 + }, + { + "epoch": 0.028451127819548873, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2778, + "step": 5473 + }, + { + "epoch": 0.028511278195488723, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2209, + "step": 5474 + }, + { + "epoch": 0.02857142857142857, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2587, + "step": 5475 + }, + { + "epoch": 0.02863157894736842, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2382, + "step": 5476 + }, + { + "epoch": 0.02869172932330827, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.341, + "step": 5477 + }, + { + "epoch": 0.02875187969924812, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.36, + "step": 5478 + }, + { + "epoch": 0.02881203007518797, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2443, + "step": 5479 + }, + { + "epoch": 0.02887218045112782, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2511, + "step": 5480 + }, + { + "epoch": 0.02893233082706767, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.3043, + "step": 5481 + }, + { + "epoch": 0.028992481203007518, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2654, + "step": 5482 + }, + { + "epoch": 0.029052631578947368, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2753, + "step": 5483 + }, + { + "epoch": 0.02911278195488722, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2703, + "step": 5484 + }, + { + "epoch": 0.02917293233082707, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2363, + "step": 5485 + }, + { + "epoch": 0.029233082706766916, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3589, + "step": 5486 + }, + { + "epoch": 0.029293233082706767, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.2329, + "step": 5487 + }, + { + "epoch": 0.029353383458646617, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3972, + "step": 5488 + }, + { + "epoch": 0.029413533834586465, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3488, + "step": 5489 + }, + { + "epoch": 0.029473684210526315, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 5490 + }, + { + "epoch": 0.029533834586466166, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.278, + "step": 5491 + }, + { + "epoch": 0.029593984962406016, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2019, + "step": 5492 + }, + { + "epoch": 0.029654135338345863, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2186, + "step": 5493 + }, + { + "epoch": 0.029714285714285714, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2877, + "step": 5494 + }, + { + "epoch": 0.029774436090225564, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2823, + "step": 5495 + }, + { + "epoch": 0.029834586466165415, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1998, + "step": 5496 + }, + { + "epoch": 0.029894736842105262, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3314, + "step": 5497 + }, + { + "epoch": 0.029954887218045113, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2259, + "step": 5498 + }, + { + "epoch": 0.030015037593984963, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2856, + "step": 5499 + }, + { + "epoch": 0.03007518796992481, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2481, + "step": 5500 + }, + { + "epoch": 0.03013533834586466, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2235, + "step": 5501 + }, + { + "epoch": 0.03019548872180451, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2952, + "step": 5502 + }, + { + "epoch": 0.030255639097744362, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1831, + "step": 5503 + }, + { + "epoch": 0.03031578947368421, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2214, + "step": 5504 + }, + { + "epoch": 0.03037593984962406, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.329, + "step": 5505 + }, + { + "epoch": 0.03043609022556391, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 5506 + }, + { + "epoch": 0.03049624060150376, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3059, + "step": 5507 + }, + { + "epoch": 0.030556390977443608, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2475, + "step": 5508 + }, + { + "epoch": 0.03061654135338346, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 5509 + }, + { + "epoch": 0.03067669172932331, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.384, + "step": 5510 + }, + { + "epoch": 0.030736842105263156, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2301, + "step": 5511 + }, + { + "epoch": 0.030796992481203007, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.2768, + "step": 5512 + }, + { + "epoch": 0.030857142857142857, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 5513 + }, + { + "epoch": 0.030917293233082708, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3107, + "step": 5514 + }, + { + "epoch": 0.030977443609022555, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2795, + "step": 5515 + }, + { + "epoch": 0.031037593984962406, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2822, + "step": 5516 + }, + { + "epoch": 0.031097744360902256, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2884, + "step": 5517 + }, + { + "epoch": 0.031157894736842107, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 5518 + }, + { + "epoch": 0.031218045112781954, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2854, + "step": 5519 + }, + { + "epoch": 0.03127819548872181, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3818, + "step": 5520 + }, + { + "epoch": 0.03133834586466165, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2745, + "step": 5521 + }, + { + "epoch": 0.0313984962406015, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2177, + "step": 5522 + }, + { + "epoch": 0.03145864661654135, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3839, + "step": 5523 + }, + { + "epoch": 0.0315187969924812, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2561, + "step": 5524 + }, + { + "epoch": 0.031578947368421054, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.272, + "step": 5525 + }, + { + "epoch": 0.031639097744360904, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2389, + "step": 5526 + }, + { + "epoch": 0.031699248120300755, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3025, + "step": 5527 + }, + { + "epoch": 0.0317593984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2855, + "step": 5528 + }, + { + "epoch": 0.03181954887218045, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.308, + "step": 5529 + }, + { + "epoch": 0.0318796992481203, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2666, + "step": 5530 + }, + { + "epoch": 0.03193984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.288, + "step": 5531 + }, + { + "epoch": 0.032, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2617, + "step": 5532 + }, + { + "epoch": 0.03206015037593985, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.3415, + "step": 5533 + }, + { + "epoch": 0.0321203007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2339, + "step": 5534 + }, + { + "epoch": 0.032180451127819545, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2347, + "step": 5535 + }, + { + "epoch": 0.032240601503759396, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.284, + "step": 5536 + }, + { + "epoch": 0.03230075187969925, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3351, + "step": 5537 + }, + { + "epoch": 0.0323609022556391, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 5538 + }, + { + "epoch": 0.03242105263157895, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2213, + "step": 5539 + }, + { + "epoch": 0.0324812030075188, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.258, + "step": 5540 + }, + { + "epoch": 0.03254135338345865, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.442, + "step": 5541 + }, + { + "epoch": 0.0326015037593985, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2017, + "step": 5542 + }, + { + "epoch": 0.03266165413533834, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2377, + "step": 5543 + }, + { + "epoch": 0.032721804511278194, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2451, + "step": 5544 + }, + { + "epoch": 0.032781954887218044, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3647, + "step": 5545 + }, + { + "epoch": 0.032842105263157895, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3209, + "step": 5546 + }, + { + "epoch": 0.032902255639097745, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 5547 + }, + { + "epoch": 0.032962406015037596, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3559, + "step": 5548 + }, + { + "epoch": 0.033022556390977446, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 5549 + }, + { + "epoch": 0.03308270676691729, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 5550 + }, + { + "epoch": 0.03314285714285714, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2379, + "step": 5551 + }, + { + "epoch": 0.03320300751879699, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2686, + "step": 5552 + }, + { + "epoch": 0.03326315789473684, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2164, + "step": 5553 + }, + { + "epoch": 0.03332330827067669, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2571, + "step": 5554 + }, + { + "epoch": 0.03338345864661654, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2627, + "step": 5555 + }, + { + "epoch": 0.03344360902255639, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2817, + "step": 5556 + }, + { + "epoch": 0.033503759398496244, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2677, + "step": 5557 + }, + { + "epoch": 0.03356390977443609, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2212, + "step": 5558 + }, + { + "epoch": 0.03362406015037594, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3062, + "step": 5559 + }, + { + "epoch": 0.03368421052631579, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2541, + "step": 5560 + }, + { + "epoch": 0.03374436090225564, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2407, + "step": 5561 + }, + { + "epoch": 0.03380451127819549, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.3179, + "step": 5562 + }, + { + "epoch": 0.03386466165413534, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2812, + "step": 5563 + }, + { + "epoch": 0.03392481203007519, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.334, + "step": 5564 + }, + { + "epoch": 0.033984962406015035, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1896, + "step": 5565 + }, + { + "epoch": 0.034045112781954885, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3488, + "step": 5566 + }, + { + "epoch": 0.034105263157894736, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2844, + "step": 5567 + }, + { + "epoch": 0.034165413533834586, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2613, + "step": 5568 + }, + { + "epoch": 0.03422556390977444, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3794, + "step": 5569 + }, + { + "epoch": 0.03428571428571429, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2392, + "step": 5570 + }, + { + "epoch": 0.03434586466165414, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2985, + "step": 5571 + }, + { + "epoch": 0.03440601503759398, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3062, + "step": 5572 + }, + { + "epoch": 0.03446616541353383, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2757, + "step": 5573 + }, + { + "epoch": 0.03452631578947368, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2664, + "step": 5574 + }, + { + "epoch": 0.03458646616541353, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.3718, + "step": 5575 + }, + { + "epoch": 0.034646616541353384, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.329, + "step": 5576 + }, + { + "epoch": 0.034706766917293234, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2312, + "step": 5577 + }, + { + "epoch": 0.034766917293233085, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2564, + "step": 5578 + }, + { + "epoch": 0.034827067669172936, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2377, + "step": 5579 + }, + { + "epoch": 0.03488721804511278, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3345, + "step": 5580 + }, + { + "epoch": 0.03494736842105263, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2395, + "step": 5581 + }, + { + "epoch": 0.03500751879699248, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2633, + "step": 5582 + }, + { + "epoch": 0.03506766917293233, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 5583 + }, + { + "epoch": 0.03512781954887218, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2636, + "step": 5584 + }, + { + "epoch": 0.03518796992481203, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2414, + "step": 5585 + }, + { + "epoch": 0.03524812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2114, + "step": 5586 + }, + { + "epoch": 0.035308270676691726, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2712, + "step": 5587 + }, + { + "epoch": 0.03536842105263158, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2383, + "step": 5588 + }, + { + "epoch": 0.03542857142857143, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2025, + "step": 5589 + }, + { + "epoch": 0.03548872180451128, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3158, + "step": 5590 + }, + { + "epoch": 0.03554887218045113, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 5591 + }, + { + "epoch": 0.03560902255639098, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2371, + "step": 5592 + }, + { + "epoch": 0.03566917293233083, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 5593 + }, + { + "epoch": 0.03572932330827068, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.2651, + "step": 5594 + }, + { + "epoch": 0.035789473684210524, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 5595 + }, + { + "epoch": 0.035849624060150374, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3024, + "step": 5596 + }, + { + "epoch": 0.035909774436090225, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.323, + "step": 5597 + }, + { + "epoch": 0.035969924812030075, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.271, + "step": 5598 + }, + { + "epoch": 0.036030075187969926, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3079, + "step": 5599 + }, + { + "epoch": 0.03609022556390978, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 5600 + }, + { + "epoch": 0.03615037593984963, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3632, + "step": 5601 + }, + { + "epoch": 0.03621052631578947, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2391, + "step": 5602 + }, + { + "epoch": 0.03627067669172932, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2816, + "step": 5603 + }, + { + "epoch": 0.03633082706766917, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2861, + "step": 5604 + }, + { + "epoch": 0.03639097744360902, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2568, + "step": 5605 + }, + { + "epoch": 0.03645112781954887, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2112, + "step": 5606 + }, + { + "epoch": 0.036511278195488724, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3551, + "step": 5607 + }, + { + "epoch": 0.036571428571428574, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2951, + "step": 5608 + }, + { + "epoch": 0.03663157894736842, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2462, + "step": 5609 + }, + { + "epoch": 0.03669172932330827, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3309, + "step": 5610 + }, + { + "epoch": 0.03675187969924812, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3263, + "step": 5611 + }, + { + "epoch": 0.03681203007518797, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2773, + "step": 5612 + }, + { + "epoch": 0.03687218045112782, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3037, + "step": 5613 + }, + { + "epoch": 0.03693233082706767, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1903, + "step": 5614 + }, + { + "epoch": 0.03699248120300752, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2969, + "step": 5615 + }, + { + "epoch": 0.03705263157894737, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3643, + "step": 5616 + }, + { + "epoch": 0.037112781954887215, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3175, + "step": 5617 + }, + { + "epoch": 0.037172932330827066, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3294, + "step": 5618 + }, + { + "epoch": 0.037233082706766917, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3176, + "step": 5619 + }, + { + "epoch": 0.03729323308270677, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 5620 + }, + { + "epoch": 0.03735338345864662, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.3584, + "step": 5621 + }, + { + "epoch": 0.03741353383458647, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2638, + "step": 5622 + }, + { + "epoch": 0.03747368421052632, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2613, + "step": 5623 + }, + { + "epoch": 0.03753383458646616, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2719, + "step": 5624 + }, + { + "epoch": 0.03759398496240601, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.4597, + "step": 5625 + }, + { + "epoch": 0.037654135338345864, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.25, + "step": 5626 + }, + { + "epoch": 0.037714285714285714, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 5627 + }, + { + "epoch": 0.037774436090225565, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3522, + "step": 5628 + }, + { + "epoch": 0.037834586466165415, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2877, + "step": 5629 + }, + { + "epoch": 0.037894736842105266, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2281, + "step": 5630 + }, + { + "epoch": 0.03795488721804511, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 5631 + }, + { + "epoch": 0.03801503759398496, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2148, + "step": 5632 + }, + { + "epoch": 0.03807518796992481, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2925, + "step": 5633 + }, + { + "epoch": 0.03813533834586466, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2141, + "step": 5634 + }, + { + "epoch": 0.03819548872180451, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2551, + "step": 5635 + }, + { + "epoch": 0.03825563909774436, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3133, + "step": 5636 + }, + { + "epoch": 0.03831578947368421, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2615, + "step": 5637 + }, + { + "epoch": 0.03837593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.282, + "step": 5638 + }, + { + "epoch": 0.03843609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2333, + "step": 5639 + }, + { + "epoch": 0.03849624060150376, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2299, + "step": 5640 + }, + { + "epoch": 0.03855639097744361, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3289, + "step": 5641 + }, + { + "epoch": 0.03861654135338346, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3009, + "step": 5642 + }, + { + "epoch": 0.03867669172932331, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.238, + "step": 5643 + }, + { + "epoch": 0.03873684210526316, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2842, + "step": 5644 + }, + { + "epoch": 0.03879699248120301, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2037, + "step": 5645 + }, + { + "epoch": 0.038857142857142854, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2109, + "step": 5646 + }, + { + "epoch": 0.038917293233082705, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 5647 + }, + { + "epoch": 0.038977443609022555, + "grad_norm": 0.416015625, + "learning_rate": 3e-05, + "loss": 1.2696, + "step": 5648 + }, + { + "epoch": 0.039037593984962406, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 5649 + }, + { + "epoch": 0.039097744360902256, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 5650 + }, + { + "epoch": 0.03915789473684211, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3633, + "step": 5651 + }, + { + "epoch": 0.03921804511278196, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.2941, + "step": 5652 + }, + { + "epoch": 0.03927819548872181, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2567, + "step": 5653 + }, + { + "epoch": 0.03933834586466165, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2895, + "step": 5654 + }, + { + "epoch": 0.0393984962406015, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.3206, + "step": 5655 + }, + { + "epoch": 0.03945864661654135, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2845, + "step": 5656 + }, + { + "epoch": 0.0395187969924812, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2576, + "step": 5657 + }, + { + "epoch": 0.039578947368421054, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 5658 + }, + { + "epoch": 0.039639097744360904, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3409, + "step": 5659 + }, + { + "epoch": 0.039699248120300755, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2476, + "step": 5660 + }, + { + "epoch": 0.0397593984962406, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3256, + "step": 5661 + }, + { + "epoch": 0.03981954887218045, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.2171, + "step": 5662 + }, + { + "epoch": 0.0398796992481203, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3427, + "step": 5663 + }, + { + "epoch": 0.03993984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2631, + "step": 5664 + }, + { + "epoch": 0.04, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3446, + "step": 5665 + }, + { + "epoch": 0.04006015037593985, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3967, + "step": 5666 + }, + { + "epoch": 0.0401203007518797, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2145, + "step": 5667 + }, + { + "epoch": 0.040180451127819546, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2785, + "step": 5668 + }, + { + "epoch": 0.040240601503759396, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2407, + "step": 5669 + }, + { + "epoch": 0.04030075187969925, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2211, + "step": 5670 + }, + { + "epoch": 0.0403609022556391, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2993, + "step": 5671 + }, + { + "epoch": 0.04042105263157895, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 5672 + }, + { + "epoch": 0.0404812030075188, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.3756, + "step": 5673 + }, + { + "epoch": 0.04054135338345865, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3246, + "step": 5674 + }, + { + "epoch": 0.0406015037593985, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 5675 + }, + { + "epoch": 0.04066165413533834, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3467, + "step": 5676 + }, + { + "epoch": 0.040721804511278194, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2176, + "step": 5677 + }, + { + "epoch": 0.040781954887218044, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2912, + "step": 5678 + }, + { + "epoch": 0.040842105263157895, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 5679 + }, + { + "epoch": 0.040902255639097745, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.318, + "step": 5680 + }, + { + "epoch": 0.040962406015037596, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1785, + "step": 5681 + }, + { + "epoch": 0.04102255639097745, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2915, + "step": 5682 + }, + { + "epoch": 0.04108270676691729, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2553, + "step": 5683 + }, + { + "epoch": 0.04114285714285714, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2295, + "step": 5684 + }, + { + "epoch": 0.04120300751879699, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2933, + "step": 5685 + }, + { + "epoch": 0.04126315789473684, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2613, + "step": 5686 + }, + { + "epoch": 0.04132330827067669, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.354, + "step": 5687 + }, + { + "epoch": 0.04138345864661654, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2735, + "step": 5688 + }, + { + "epoch": 0.041443609022556394, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2186, + "step": 5689 + }, + { + "epoch": 0.04150375939849624, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 5690 + }, + { + "epoch": 0.04156390977443609, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2761, + "step": 5691 + }, + { + "epoch": 0.04162406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 5692 + }, + { + "epoch": 0.04168421052631579, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3167, + "step": 5693 + }, + { + "epoch": 0.04174436090225564, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2719, + "step": 5694 + }, + { + "epoch": 0.04180451127819549, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.4009, + "step": 5695 + }, + { + "epoch": 0.04186466165413534, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2967, + "step": 5696 + }, + { + "epoch": 0.04192481203007519, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2729, + "step": 5697 + }, + { + "epoch": 0.041984962406015035, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3214, + "step": 5698 + }, + { + "epoch": 0.042045112781954885, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3484, + "step": 5699 + }, + { + "epoch": 0.042105263157894736, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2682, + "step": 5700 + }, + { + "epoch": 0.042165413533834586, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2955, + "step": 5701 + }, + { + "epoch": 0.04222556390977444, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 5702 + }, + { + "epoch": 0.04228571428571429, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2228, + "step": 5703 + }, + { + "epoch": 0.04234586466165414, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2308, + "step": 5704 + }, + { + "epoch": 0.04240601503759398, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3373, + "step": 5705 + }, + { + "epoch": 0.04246616541353383, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2992, + "step": 5706 + }, + { + "epoch": 0.04252631578947368, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2217, + "step": 5707 + }, + { + "epoch": 0.04258646616541353, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2466, + "step": 5708 + }, + { + "epoch": 0.042646616541353384, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3167, + "step": 5709 + }, + { + "epoch": 0.042706766917293235, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2953, + "step": 5710 + }, + { + "epoch": 0.042766917293233085, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2664, + "step": 5711 + }, + { + "epoch": 0.042827067669172936, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3049, + "step": 5712 + }, + { + "epoch": 0.04288721804511278, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2534, + "step": 5713 + }, + { + "epoch": 0.04294736842105263, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2945, + "step": 5714 + }, + { + "epoch": 0.04300751879699248, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.319, + "step": 5715 + }, + { + "epoch": 0.04306766917293233, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3373, + "step": 5716 + }, + { + "epoch": 0.04312781954887218, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2983, + "step": 5717 + }, + { + "epoch": 0.04318796992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2284, + "step": 5718 + }, + { + "epoch": 0.04324812030075188, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2021, + "step": 5719 + }, + { + "epoch": 0.043308270676691726, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2288, + "step": 5720 + }, + { + "epoch": 0.04336842105263158, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2408, + "step": 5721 + }, + { + "epoch": 0.04342857142857143, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2674, + "step": 5722 + }, + { + "epoch": 0.04348872180451128, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3181, + "step": 5723 + }, + { + "epoch": 0.04354887218045113, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2678, + "step": 5724 + }, + { + "epoch": 0.04360902255639098, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 5725 + }, + { + "epoch": 0.04366917293233083, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3181, + "step": 5726 + }, + { + "epoch": 0.04372932330827067, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2236, + "step": 5727 + }, + { + "epoch": 0.043789473684210524, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 5728 + }, + { + "epoch": 0.043849624060150375, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2994, + "step": 5729 + }, + { + "epoch": 0.043909774436090225, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2802, + "step": 5730 + }, + { + "epoch": 0.043969924812030076, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.371, + "step": 5731 + }, + { + "epoch": 0.044030075187969926, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3168, + "step": 5732 + }, + { + "epoch": 0.04409022556390978, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2547, + "step": 5733 + }, + { + "epoch": 0.04415037593984963, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.4054, + "step": 5734 + }, + { + "epoch": 0.04421052631578947, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 5735 + }, + { + "epoch": 0.04427067669172932, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.2861, + "step": 5736 + }, + { + "epoch": 0.04433082706766917, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3451, + "step": 5737 + }, + { + "epoch": 0.04439097744360902, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.26, + "step": 5738 + }, + { + "epoch": 0.04445112781954887, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3141, + "step": 5739 + }, + { + "epoch": 0.044511278195488724, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3715, + "step": 5740 + }, + { + "epoch": 0.044571428571428574, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2741, + "step": 5741 + }, + { + "epoch": 0.04463157894736842, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2245, + "step": 5742 + }, + { + "epoch": 0.04469172932330827, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2852, + "step": 5743 + }, + { + "epoch": 0.04475187969924812, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 5744 + }, + { + "epoch": 0.04481203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 5745 + }, + { + "epoch": 0.04487218045112782, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 5746 + }, + { + "epoch": 0.04493233082706767, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3464, + "step": 5747 + }, + { + "epoch": 0.04499248120300752, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.2552, + "step": 5748 + }, + { + "epoch": 0.045052631578947365, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2725, + "step": 5749 + }, + { + "epoch": 0.045112781954887216, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 5750 + }, + { + "epoch": 0.045172932330827066, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 5751 + }, + { + "epoch": 0.04523308270676692, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2599, + "step": 5752 + }, + { + "epoch": 0.04529323308270677, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2142, + "step": 5753 + }, + { + "epoch": 0.04535338345864662, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.4986, + "step": 5754 + }, + { + "epoch": 0.04541353383458647, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3039, + "step": 5755 + }, + { + "epoch": 0.04547368421052632, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 5756 + }, + { + "epoch": 0.04553383458646616, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2843, + "step": 5757 + }, + { + "epoch": 0.04559398496240601, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2794, + "step": 5758 + }, + { + "epoch": 0.045654135338345864, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.258, + "step": 5759 + }, + { + "epoch": 0.045714285714285714, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2225, + "step": 5760 + }, + { + "epoch": 0.045774436090225565, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3134, + "step": 5761 + }, + { + "epoch": 0.045834586466165415, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.258, + "step": 5762 + }, + { + "epoch": 0.045894736842105266, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2165, + "step": 5763 + }, + { + "epoch": 0.04595488721804511, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2443, + "step": 5764 + }, + { + "epoch": 0.04601503759398496, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2597, + "step": 5765 + }, + { + "epoch": 0.04607518796992481, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2662, + "step": 5766 + }, + { + "epoch": 0.04613533834586466, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.285, + "step": 5767 + }, + { + "epoch": 0.04619548872180451, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3361, + "step": 5768 + }, + { + "epoch": 0.04625563909774436, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3516, + "step": 5769 + }, + { + "epoch": 0.04631578947368421, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.286, + "step": 5770 + }, + { + "epoch": 0.046375939849624064, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2866, + "step": 5771 + }, + { + "epoch": 0.04643609022556391, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2066, + "step": 5772 + }, + { + "epoch": 0.04649624060150376, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2509, + "step": 5773 + }, + { + "epoch": 0.04655639097744361, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2602, + "step": 5774 + }, + { + "epoch": 0.04661654135338346, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2784, + "step": 5775 + }, + { + "epoch": 0.04667669172932331, + "grad_norm": 0.46484375, + "learning_rate": 3e-05, + "loss": 1.2939, + "step": 5776 + }, + { + "epoch": 0.04673684210526316, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 5777 + }, + { + "epoch": 0.04679699248120301, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2689, + "step": 5778 + }, + { + "epoch": 0.046857142857142854, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2733, + "step": 5779 + }, + { + "epoch": 0.046917293233082705, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2636, + "step": 5780 + }, + { + "epoch": 0.046977443609022555, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3259, + "step": 5781 + }, + { + "epoch": 0.047037593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3155, + "step": 5782 + }, + { + "epoch": 0.047097744360902256, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1888, + "step": 5783 + }, + { + "epoch": 0.04715789473684211, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.2831, + "step": 5784 + }, + { + "epoch": 0.04721804511278196, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 5785 + }, + { + "epoch": 0.0472781954887218, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3754, + "step": 5786 + }, + { + "epoch": 0.04733834586466165, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1925, + "step": 5787 + }, + { + "epoch": 0.0473984962406015, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3745, + "step": 5788 + }, + { + "epoch": 0.04745864661654135, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.273, + "step": 5789 + }, + { + "epoch": 0.0475187969924812, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3209, + "step": 5790 + }, + { + "epoch": 0.047578947368421054, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2696, + "step": 5791 + }, + { + "epoch": 0.047639097744360905, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.3897, + "step": 5792 + }, + { + "epoch": 0.047699248120300755, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2188, + "step": 5793 + }, + { + "epoch": 0.0477593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1841, + "step": 5794 + }, + { + "epoch": 0.04781954887218045, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2465, + "step": 5795 + }, + { + "epoch": 0.0478796992481203, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2481, + "step": 5796 + }, + { + "epoch": 0.04793984962406015, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2736, + "step": 5797 + }, + { + "epoch": 0.048, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 5798 + }, + { + "epoch": 0.04806015037593985, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2593, + "step": 5799 + }, + { + "epoch": 0.0481203007518797, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2305, + "step": 5800 + }, + { + "epoch": 0.048180451127819546, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3161, + "step": 5801 + }, + { + "epoch": 0.048240601503759396, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2856, + "step": 5802 + }, + { + "epoch": 0.04830075187969925, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 5803 + }, + { + "epoch": 0.0483609022556391, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2279, + "step": 5804 + }, + { + "epoch": 0.04842105263157895, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2823, + "step": 5805 + }, + { + "epoch": 0.0484812030075188, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 5806 + }, + { + "epoch": 0.04854135338345865, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3482, + "step": 5807 + }, + { + "epoch": 0.0486015037593985, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.29, + "step": 5808 + }, + { + "epoch": 0.04866165413533834, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2999, + "step": 5809 + }, + { + "epoch": 0.048721804511278194, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.285, + "step": 5810 + }, + { + "epoch": 0.048781954887218044, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2468, + "step": 5811 + }, + { + "epoch": 0.048842105263157895, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1764, + "step": 5812 + }, + { + "epoch": 0.048902255639097746, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3563, + "step": 5813 + }, + { + "epoch": 0.048962406015037596, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2807, + "step": 5814 + }, + { + "epoch": 0.04902255639097745, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2627, + "step": 5815 + }, + { + "epoch": 0.04908270676691729, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2351, + "step": 5816 + }, + { + "epoch": 0.04914285714285714, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1833, + "step": 5817 + }, + { + "epoch": 0.04920300751879699, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.3323, + "step": 5818 + }, + { + "epoch": 0.04926315789473684, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3149, + "step": 5819 + }, + { + "epoch": 0.04932330827067669, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2633, + "step": 5820 + }, + { + "epoch": 0.04938345864661654, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2072, + "step": 5821 + }, + { + "epoch": 0.049443609022556394, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.16, + "step": 5822 + }, + { + "epoch": 0.04950375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 5823 + }, + { + "epoch": 0.04956390977443609, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.253, + "step": 5824 + }, + { + "epoch": 0.04962406015037594, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2783, + "step": 5825 + }, + { + "epoch": 0.04968421052631579, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2029, + "step": 5826 + }, + { + "epoch": 0.04974436090225564, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3204, + "step": 5827 + }, + { + "epoch": 0.04980451127819549, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 5828 + }, + { + "epoch": 0.04986466165413534, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 5829 + }, + { + "epoch": 0.04992481203007519, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 5830 + }, + { + "epoch": 0.049984962406015035, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2878, + "step": 5831 + }, + { + "epoch": 0.050045112781954885, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.27, + "step": 5832 + }, + { + "epoch": 0.050105263157894736, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2514, + "step": 5833 + }, + { + "epoch": 0.05016541353383459, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3294, + "step": 5834 + }, + { + "epoch": 0.05022556390977444, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2834, + "step": 5835 + }, + { + "epoch": 0.05028571428571429, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.309, + "step": 5836 + }, + { + "epoch": 0.05034586466165414, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 5837 + }, + { + "epoch": 0.05040601503759398, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2868, + "step": 5838 + }, + { + "epoch": 0.05046616541353383, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 5839 + }, + { + "epoch": 0.05052631578947368, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3578, + "step": 5840 + }, + { + "epoch": 0.050586466165413534, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2318, + "step": 5841 + }, + { + "epoch": 0.050646616541353384, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.224, + "step": 5842 + }, + { + "epoch": 0.050706766917293235, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3197, + "step": 5843 + }, + { + "epoch": 0.050766917293233085, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3417, + "step": 5844 + }, + { + "epoch": 0.05082706766917293, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2336, + "step": 5845 + }, + { + "epoch": 0.05088721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3162, + "step": 5846 + }, + { + "epoch": 0.05094736842105263, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1966, + "step": 5847 + }, + { + "epoch": 0.05100751879699248, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.325, + "step": 5848 + }, + { + "epoch": 0.05106766917293233, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 5849 + }, + { + "epoch": 0.05112781954887218, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2653, + "step": 5850 + }, + { + "epoch": 0.05118796992481203, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3146, + "step": 5851 + }, + { + "epoch": 0.05124812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 5852 + }, + { + "epoch": 0.05130827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3269, + "step": 5853 + }, + { + "epoch": 0.05136842105263158, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.3716, + "step": 5854 + }, + { + "epoch": 0.05142857142857143, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2771, + "step": 5855 + }, + { + "epoch": 0.05148872180451128, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 5856 + }, + { + "epoch": 0.05154887218045113, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 5857 + }, + { + "epoch": 0.05160902255639098, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2661, + "step": 5858 + }, + { + "epoch": 0.05166917293233083, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3034, + "step": 5859 + }, + { + "epoch": 0.051729323308270674, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3378, + "step": 5860 + }, + { + "epoch": 0.051789473684210524, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2796, + "step": 5861 + }, + { + "epoch": 0.051849624060150375, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2447, + "step": 5862 + }, + { + "epoch": 0.051909774436090225, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2634, + "step": 5863 + }, + { + "epoch": 0.051969924812030076, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.262, + "step": 5864 + }, + { + "epoch": 0.052030075187969926, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.318, + "step": 5865 + }, + { + "epoch": 0.05209022556390978, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2229, + "step": 5866 + }, + { + "epoch": 0.05215037593984963, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2391, + "step": 5867 + }, + { + "epoch": 0.05221052631578947, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2329, + "step": 5868 + }, + { + "epoch": 0.05227067669172932, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 5869 + }, + { + "epoch": 0.05233082706766917, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2605, + "step": 5870 + }, + { + "epoch": 0.05239097744360902, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3036, + "step": 5871 + }, + { + "epoch": 0.05245112781954887, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 5872 + }, + { + "epoch": 0.052511278195488724, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.236, + "step": 5873 + }, + { + "epoch": 0.052571428571428575, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.284, + "step": 5874 + }, + { + "epoch": 0.05263157894736842, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2156, + "step": 5875 + }, + { + "epoch": 0.05269172932330827, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.4496, + "step": 5876 + }, + { + "epoch": 0.05275187969924812, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 5877 + }, + { + "epoch": 0.05281203007518797, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.236, + "step": 5878 + }, + { + "epoch": 0.05287218045112782, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2405, + "step": 5879 + }, + { + "epoch": 0.05293233082706767, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.309, + "step": 5880 + }, + { + "epoch": 0.05299248120300752, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3881, + "step": 5881 + }, + { + "epoch": 0.053052631578947365, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 5882 + }, + { + "epoch": 0.053112781954887216, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2352, + "step": 5883 + }, + { + "epoch": 0.053172932330827066, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3276, + "step": 5884 + }, + { + "epoch": 0.05323308270676692, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.286, + "step": 5885 + }, + { + "epoch": 0.05329323308270677, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.311, + "step": 5886 + }, + { + "epoch": 0.05335338345864662, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2692, + "step": 5887 + }, + { + "epoch": 0.05341353383458647, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.4146, + "step": 5888 + }, + { + "epoch": 0.05347368421052632, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.2756, + "step": 5889 + }, + { + "epoch": 0.05353383458646616, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2586, + "step": 5890 + }, + { + "epoch": 0.05359398496240601, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2867, + "step": 5891 + }, + { + "epoch": 0.053654135338345864, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3659, + "step": 5892 + }, + { + "epoch": 0.053714285714285714, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3309, + "step": 5893 + }, + { + "epoch": 0.053774436090225565, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2175, + "step": 5894 + }, + { + "epoch": 0.053834586466165416, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3526, + "step": 5895 + }, + { + "epoch": 0.053894736842105266, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2452, + "step": 5896 + }, + { + "epoch": 0.05395488721804511, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2152, + "step": 5897 + }, + { + "epoch": 0.05401503759398496, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 5898 + }, + { + "epoch": 0.05407518796992481, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.255, + "step": 5899 + }, + { + "epoch": 0.05413533834586466, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.317, + "step": 5900 + }, + { + "epoch": 0.05419548872180451, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2662, + "step": 5901 + }, + { + "epoch": 0.05425563909774436, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2372, + "step": 5902 + }, + { + "epoch": 0.05431578947368421, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 5903 + }, + { + "epoch": 0.05437593984962406, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2777, + "step": 5904 + }, + { + "epoch": 0.05443609022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2683, + "step": 5905 + }, + { + "epoch": 0.05449624060150376, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2802, + "step": 5906 + }, + { + "epoch": 0.05455639097744361, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 5907 + }, + { + "epoch": 0.05461654135338346, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2666, + "step": 5908 + }, + { + "epoch": 0.05467669172932331, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2441, + "step": 5909 + }, + { + "epoch": 0.05473684210526316, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3182, + "step": 5910 + }, + { + "epoch": 0.05479699248120301, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2622, + "step": 5911 + }, + { + "epoch": 0.054857142857142854, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2229, + "step": 5912 + }, + { + "epoch": 0.054917293233082705, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2885, + "step": 5913 + }, + { + "epoch": 0.054977443609022555, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 5914 + }, + { + "epoch": 0.055037593984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2352, + "step": 5915 + }, + { + "epoch": 0.05509774436090226, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 5916 + }, + { + "epoch": 0.05515789473684211, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 5917 + }, + { + "epoch": 0.05521804511278196, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1925, + "step": 5918 + }, + { + "epoch": 0.0552781954887218, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2826, + "step": 5919 + }, + { + "epoch": 0.05533834586466165, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2446, + "step": 5920 + }, + { + "epoch": 0.0553984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1954, + "step": 5921 + }, + { + "epoch": 0.05545864661654135, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.296, + "step": 5922 + }, + { + "epoch": 0.055518796992481204, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3244, + "step": 5923 + }, + { + "epoch": 0.055578947368421054, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.3354, + "step": 5924 + }, + { + "epoch": 0.055639097744360905, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 5925 + }, + { + "epoch": 0.055699248120300755, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.299, + "step": 5926 + }, + { + "epoch": 0.0557593984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2394, + "step": 5927 + }, + { + "epoch": 0.05581954887218045, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2631, + "step": 5928 + }, + { + "epoch": 0.0558796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3022, + "step": 5929 + }, + { + "epoch": 0.05593984962406015, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3881, + "step": 5930 + }, + { + "epoch": 0.056, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2819, + "step": 5931 + }, + { + "epoch": 0.05606015037593985, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2258, + "step": 5932 + }, + { + "epoch": 0.0561203007518797, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2827, + "step": 5933 + }, + { + "epoch": 0.056180451127819546, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2392, + "step": 5934 + }, + { + "epoch": 0.056240601503759396, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1934, + "step": 5935 + }, + { + "epoch": 0.05630075187969925, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 5936 + }, + { + "epoch": 0.0563609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2907, + "step": 5937 + }, + { + "epoch": 0.05642105263157895, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2823, + "step": 5938 + }, + { + "epoch": 0.0564812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2679, + "step": 5939 + }, + { + "epoch": 0.05654135338345865, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3296, + "step": 5940 + }, + { + "epoch": 0.05660150375939849, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1688, + "step": 5941 + }, + { + "epoch": 0.056661654135338343, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2436, + "step": 5942 + }, + { + "epoch": 0.056721804511278194, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 5943 + }, + { + "epoch": 0.056781954887218045, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2859, + "step": 5944 + }, + { + "epoch": 0.056842105263157895, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2923, + "step": 5945 + }, + { + "epoch": 0.056902255639097746, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 5946 + }, + { + "epoch": 0.056962406015037596, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.4157, + "step": 5947 + }, + { + "epoch": 0.05702255639097745, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2062, + "step": 5948 + }, + { + "epoch": 0.05708270676691729, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.3653, + "step": 5949 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2849, + "step": 5950 + }, + { + "epoch": 0.05720300751879699, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2981, + "step": 5951 + }, + { + "epoch": 0.05726315789473684, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 5952 + }, + { + "epoch": 0.05732330827067669, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3042, + "step": 5953 + }, + { + "epoch": 0.05738345864661654, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.21, + "step": 5954 + }, + { + "epoch": 0.057443609022556394, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 5955 + }, + { + "epoch": 0.05750375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 5956 + }, + { + "epoch": 0.05756390977443609, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2543, + "step": 5957 + }, + { + "epoch": 0.05762406015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2474, + "step": 5958 + }, + { + "epoch": 0.05768421052631579, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 5959 + }, + { + "epoch": 0.05774436090225564, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2868, + "step": 5960 + }, + { + "epoch": 0.05780451127819549, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 5961 + }, + { + "epoch": 0.05786466165413534, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.3383, + "step": 5962 + }, + { + "epoch": 0.057924812030075185, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2768, + "step": 5963 + }, + { + "epoch": 0.057984962406015035, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 5964 + }, + { + "epoch": 0.058045112781954886, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2136, + "step": 5965 + }, + { + "epoch": 0.058105263157894736, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3172, + "step": 5966 + }, + { + "epoch": 0.05816541353383459, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.223, + "step": 5967 + }, + { + "epoch": 0.05822556390977444, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2759, + "step": 5968 + }, + { + "epoch": 0.05828571428571429, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3288, + "step": 5969 + }, + { + "epoch": 0.05834586466165414, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2495, + "step": 5970 + }, + { + "epoch": 0.05840601503759398, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 5971 + }, + { + "epoch": 0.05846616541353383, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 5972 + }, + { + "epoch": 0.05852631578947368, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2476, + "step": 5973 + }, + { + "epoch": 0.058586466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 5974 + }, + { + "epoch": 0.058646616541353384, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1766, + "step": 5975 + }, + { + "epoch": 0.058706766917293235, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2037, + "step": 5976 + }, + { + "epoch": 0.058766917293233085, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2056, + "step": 5977 + }, + { + "epoch": 0.05882706766917293, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2479, + "step": 5978 + }, + { + "epoch": 0.05888721804511278, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.202, + "step": 5979 + }, + { + "epoch": 0.05894736842105263, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.28, + "step": 5980 + }, + { + "epoch": 0.05900751879699248, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3446, + "step": 5981 + }, + { + "epoch": 0.05906766917293233, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2369, + "step": 5982 + }, + { + "epoch": 0.05912781954887218, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2574, + "step": 5983 + }, + { + "epoch": 0.05918796992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.215, + "step": 5984 + }, + { + "epoch": 0.05924812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2041, + "step": 5985 + }, + { + "epoch": 0.05930827067669173, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2485, + "step": 5986 + }, + { + "epoch": 0.05936842105263158, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2885, + "step": 5987 + }, + { + "epoch": 0.05942857142857143, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3037, + "step": 5988 + }, + { + "epoch": 0.05948872180451128, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2164, + "step": 5989 + }, + { + "epoch": 0.05954887218045113, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 5990 + }, + { + "epoch": 0.05960902255639098, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2488, + "step": 5991 + }, + { + "epoch": 0.05966917293233083, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2223, + "step": 5992 + }, + { + "epoch": 0.059729323308270674, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2601, + "step": 5993 + }, + { + "epoch": 0.059789473684210524, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.229, + "step": 5994 + }, + { + "epoch": 0.059849624060150375, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2819, + "step": 5995 + }, + { + "epoch": 0.059909774436090225, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2387, + "step": 5996 + }, + { + "epoch": 0.059969924812030076, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2321, + "step": 5997 + }, + { + "epoch": 0.06003007518796993, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2004, + "step": 5998 + }, + { + "epoch": 0.06009022556390978, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2541, + "step": 5999 + }, + { + "epoch": 0.06015037593984962, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.273, + "step": 6000 + }, + { + "epoch": 0.06021052631578947, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2586, + "step": 6001 + }, + { + "epoch": 0.06027067669172932, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1696, + "step": 6002 + }, + { + "epoch": 0.06033082706766917, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1872, + "step": 6003 + }, + { + "epoch": 0.06039097744360902, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.302, + "step": 6004 + }, + { + "epoch": 0.060451127819548874, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2356, + "step": 6005 + }, + { + "epoch": 0.060511278195488724, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 6006 + }, + { + "epoch": 0.060571428571428575, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2948, + "step": 6007 + }, + { + "epoch": 0.06063157894736842, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.16, + "step": 6008 + }, + { + "epoch": 0.06069172932330827, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 6009 + }, + { + "epoch": 0.06075187969924812, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 6010 + }, + { + "epoch": 0.06081203007518797, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 6011 + }, + { + "epoch": 0.06087218045112782, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2251, + "step": 6012 + }, + { + "epoch": 0.06093233082706767, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2729, + "step": 6013 + }, + { + "epoch": 0.06099248120300752, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2632, + "step": 6014 + }, + { + "epoch": 0.061052631578947365, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.245, + "step": 6015 + }, + { + "epoch": 0.061112781954887216, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.232, + "step": 6016 + }, + { + "epoch": 0.061172932330827066, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3092, + "step": 6017 + }, + { + "epoch": 0.06123308270676692, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2987, + "step": 6018 + }, + { + "epoch": 0.06129323308270677, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2928, + "step": 6019 + }, + { + "epoch": 0.06135338345864662, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.304, + "step": 6020 + }, + { + "epoch": 0.06141353383458647, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 6021 + }, + { + "epoch": 0.06147368421052631, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2097, + "step": 6022 + }, + { + "epoch": 0.06153383458646616, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2743, + "step": 6023 + }, + { + "epoch": 0.06159398496240601, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 6024 + }, + { + "epoch": 0.061654135338345864, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 6025 + }, + { + "epoch": 0.061714285714285715, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3017, + "step": 6026 + }, + { + "epoch": 0.061774436090225565, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3242, + "step": 6027 + }, + { + "epoch": 0.061834586466165416, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 6028 + }, + { + "epoch": 0.061894736842105266, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 6029 + }, + { + "epoch": 0.06195488721804511, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2163, + "step": 6030 + }, + { + "epoch": 0.06201503759398496, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2565, + "step": 6031 + }, + { + "epoch": 0.06207518796992481, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2445, + "step": 6032 + }, + { + "epoch": 0.06213533834586466, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.245, + "step": 6033 + }, + { + "epoch": 0.06219548872180451, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.3063, + "step": 6034 + }, + { + "epoch": 0.06225563909774436, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2463, + "step": 6035 + }, + { + "epoch": 0.06231578947368421, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2744, + "step": 6036 + }, + { + "epoch": 0.06237593984962406, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2395, + "step": 6037 + }, + { + "epoch": 0.06243609022556391, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2063, + "step": 6038 + }, + { + "epoch": 0.06249624060150376, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3546, + "step": 6039 + }, + { + "epoch": 0.06255639097744362, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2328, + "step": 6040 + }, + { + "epoch": 0.06261654135338346, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2352, + "step": 6041 + }, + { + "epoch": 0.0626766917293233, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2884, + "step": 6042 + }, + { + "epoch": 0.06273684210526316, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2065, + "step": 6043 + }, + { + "epoch": 0.062796992481203, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1918, + "step": 6044 + }, + { + "epoch": 0.06285714285714286, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2604, + "step": 6045 + }, + { + "epoch": 0.0629172932330827, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2456, + "step": 6046 + }, + { + "epoch": 0.06297744360902256, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 6047 + }, + { + "epoch": 0.0630375939849624, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3402, + "step": 6048 + }, + { + "epoch": 0.06309774436090225, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2513, + "step": 6049 + }, + { + "epoch": 0.06315789473684211, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2202, + "step": 6050 + }, + { + "epoch": 0.06321804511278195, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3104, + "step": 6051 + }, + { + "epoch": 0.06327819548872181, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3441, + "step": 6052 + }, + { + "epoch": 0.06333834586466165, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2916, + "step": 6053 + }, + { + "epoch": 0.06339849624060151, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2429, + "step": 6054 + }, + { + "epoch": 0.06345864661654135, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2772, + "step": 6055 + }, + { + "epoch": 0.0635187969924812, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2922, + "step": 6056 + }, + { + "epoch": 0.06357894736842105, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3035, + "step": 6057 + }, + { + "epoch": 0.0636390977443609, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.2494, + "step": 6058 + }, + { + "epoch": 0.06369924812030076, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 6059 + }, + { + "epoch": 0.0637593984962406, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2342, + "step": 6060 + }, + { + "epoch": 0.06381954887218046, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2583, + "step": 6061 + }, + { + "epoch": 0.0638796992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2571, + "step": 6062 + }, + { + "epoch": 0.06393984962406014, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3278, + "step": 6063 + }, + { + "epoch": 0.064, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 6064 + }, + { + "epoch": 0.06406015037593984, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.252, + "step": 6065 + }, + { + "epoch": 0.0641203007518797, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2284, + "step": 6066 + }, + { + "epoch": 0.06418045112781955, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.267, + "step": 6067 + }, + { + "epoch": 0.0642406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3776, + "step": 6068 + }, + { + "epoch": 0.06430075187969925, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 6069 + }, + { + "epoch": 0.06436090225563909, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2371, + "step": 6070 + }, + { + "epoch": 0.06442105263157895, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2805, + "step": 6071 + }, + { + "epoch": 0.06448120300751879, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 6072 + }, + { + "epoch": 0.06454135338345865, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2085, + "step": 6073 + }, + { + "epoch": 0.0646015037593985, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 6074 + }, + { + "epoch": 0.06466165413533835, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3897, + "step": 6075 + }, + { + "epoch": 0.0647218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 6076 + }, + { + "epoch": 0.06478195488721805, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 6077 + }, + { + "epoch": 0.0648421052631579, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2455, + "step": 6078 + }, + { + "epoch": 0.06490225563909774, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2113, + "step": 6079 + }, + { + "epoch": 0.0649624060150376, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2509, + "step": 6080 + }, + { + "epoch": 0.06502255639097744, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 6081 + }, + { + "epoch": 0.0650827067669173, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2436, + "step": 6082 + }, + { + "epoch": 0.06514285714285714, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2439, + "step": 6083 + }, + { + "epoch": 0.065203007518797, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2578, + "step": 6084 + }, + { + "epoch": 0.06526315789473684, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2763, + "step": 6085 + }, + { + "epoch": 0.06532330827067669, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2112, + "step": 6086 + }, + { + "epoch": 0.06538345864661654, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3381, + "step": 6087 + }, + { + "epoch": 0.06544360902255639, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2544, + "step": 6088 + }, + { + "epoch": 0.06550375939849624, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1986, + "step": 6089 + }, + { + "epoch": 0.06556390977443609, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2318, + "step": 6090 + }, + { + "epoch": 0.06562406015037595, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3429, + "step": 6091 + }, + { + "epoch": 0.06568421052631579, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.278, + "step": 6092 + }, + { + "epoch": 0.06574436090225563, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3367, + "step": 6093 + }, + { + "epoch": 0.06580451127819549, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2835, + "step": 6094 + }, + { + "epoch": 0.06586466165413533, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2461, + "step": 6095 + }, + { + "epoch": 0.06592481203007519, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2573, + "step": 6096 + }, + { + "epoch": 0.06598496240601504, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.1946, + "step": 6097 + }, + { + "epoch": 0.06604511278195489, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.296, + "step": 6098 + }, + { + "epoch": 0.06610526315789474, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.266, + "step": 6099 + }, + { + "epoch": 0.06616541353383458, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2354, + "step": 6100 + }, + { + "epoch": 0.06622556390977444, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2696, + "step": 6101 + }, + { + "epoch": 0.06628571428571428, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2409, + "step": 6102 + }, + { + "epoch": 0.06634586466165414, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3744, + "step": 6103 + }, + { + "epoch": 0.06640601503759398, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.2884, + "step": 6104 + }, + { + "epoch": 0.06646616541353384, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3343, + "step": 6105 + }, + { + "epoch": 0.06652631578947368, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 6106 + }, + { + "epoch": 0.06658646616541353, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 6107 + }, + { + "epoch": 0.06664661654135338, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2715, + "step": 6108 + }, + { + "epoch": 0.06670676691729323, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 6109 + }, + { + "epoch": 0.06676691729323309, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2741, + "step": 6110 + }, + { + "epoch": 0.06682706766917293, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2965, + "step": 6111 + }, + { + "epoch": 0.06688721804511279, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 6112 + }, + { + "epoch": 0.06694736842105263, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 6113 + }, + { + "epoch": 0.06700751879699249, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3326, + "step": 6114 + }, + { + "epoch": 0.06706766917293233, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2024, + "step": 6115 + }, + { + "epoch": 0.06712781954887218, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3072, + "step": 6116 + }, + { + "epoch": 0.06718796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2411, + "step": 6117 + }, + { + "epoch": 0.06724812030075188, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2473, + "step": 6118 + }, + { + "epoch": 0.06730827067669173, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2624, + "step": 6119 + }, + { + "epoch": 0.06736842105263158, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3332, + "step": 6120 + }, + { + "epoch": 0.06742857142857143, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2893, + "step": 6121 + }, + { + "epoch": 0.06748872180451128, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 6122 + }, + { + "epoch": 0.06754887218045112, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.3492, + "step": 6123 + }, + { + "epoch": 0.06760902255639098, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1722, + "step": 6124 + }, + { + "epoch": 0.06766917293233082, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2313, + "step": 6125 + }, + { + "epoch": 0.06772932330827068, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3517, + "step": 6126 + }, + { + "epoch": 0.06778947368421052, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2476, + "step": 6127 + }, + { + "epoch": 0.06784962406015038, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2433, + "step": 6128 + }, + { + "epoch": 0.06790977443609023, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 6129 + }, + { + "epoch": 0.06796992481203007, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 6130 + }, + { + "epoch": 0.06803007518796993, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.4383, + "step": 6131 + }, + { + "epoch": 0.06809022556390977, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2024, + "step": 6132 + }, + { + "epoch": 0.06815037593984963, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2119, + "step": 6133 + }, + { + "epoch": 0.06821052631578947, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.262, + "step": 6134 + }, + { + "epoch": 0.06827067669172933, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3394, + "step": 6135 + }, + { + "epoch": 0.06833082706766917, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2514, + "step": 6136 + }, + { + "epoch": 0.06839097744360902, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2814, + "step": 6137 + }, + { + "epoch": 0.06845112781954887, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 6138 + }, + { + "epoch": 0.06851127819548872, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2963, + "step": 6139 + }, + { + "epoch": 0.06857142857142857, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2823, + "step": 6140 + }, + { + "epoch": 0.06863157894736842, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.22, + "step": 6141 + }, + { + "epoch": 0.06869172932330828, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2253, + "step": 6142 + }, + { + "epoch": 0.06875187969924812, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3056, + "step": 6143 + }, + { + "epoch": 0.06881203007518796, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2674, + "step": 6144 + }, + { + "epoch": 0.06887218045112782, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2748, + "step": 6145 + }, + { + "epoch": 0.06893233082706766, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2138, + "step": 6146 + }, + { + "epoch": 0.06899248120300752, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.2606, + "step": 6147 + }, + { + "epoch": 0.06905263157894737, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2802, + "step": 6148 + }, + { + "epoch": 0.06911278195488722, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 6149 + }, + { + "epoch": 0.06917293233082707, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2611, + "step": 6150 + }, + { + "epoch": 0.06923308270676692, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.243, + "step": 6151 + }, + { + "epoch": 0.06929323308270677, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.3023, + "step": 6152 + }, + { + "epoch": 0.06935338345864661, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3055, + "step": 6153 + }, + { + "epoch": 0.06941353383458647, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1713, + "step": 6154 + }, + { + "epoch": 0.06947368421052631, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.3243, + "step": 6155 + }, + { + "epoch": 0.06953383458646617, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2141, + "step": 6156 + }, + { + "epoch": 0.06959398496240601, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 6157 + }, + { + "epoch": 0.06965413533834587, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1951, + "step": 6158 + }, + { + "epoch": 0.06971428571428571, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2995, + "step": 6159 + }, + { + "epoch": 0.06977443609022556, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.34, + "step": 6160 + }, + { + "epoch": 0.06983458646616542, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2286, + "step": 6161 + }, + { + "epoch": 0.06989473684210526, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2942, + "step": 6162 + }, + { + "epoch": 0.06995488721804512, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2769, + "step": 6163 + }, + { + "epoch": 0.07001503759398496, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2858, + "step": 6164 + }, + { + "epoch": 0.07007518796992482, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.3517, + "step": 6165 + }, + { + "epoch": 0.07013533834586466, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 6166 + }, + { + "epoch": 0.0701954887218045, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2094, + "step": 6167 + }, + { + "epoch": 0.07025563909774436, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 6168 + }, + { + "epoch": 0.0703157894736842, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.3515, + "step": 6169 + }, + { + "epoch": 0.07037593984962406, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.21, + "step": 6170 + }, + { + "epoch": 0.07043609022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 6171 + }, + { + "epoch": 0.07049624060150377, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 6172 + }, + { + "epoch": 0.07055639097744361, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1658, + "step": 6173 + }, + { + "epoch": 0.07061654135338345, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3163, + "step": 6174 + }, + { + "epoch": 0.07067669172932331, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 6175 + }, + { + "epoch": 0.07073684210526315, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2538, + "step": 6176 + }, + { + "epoch": 0.07079699248120301, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2319, + "step": 6177 + }, + { + "epoch": 0.07085714285714285, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2586, + "step": 6178 + }, + { + "epoch": 0.07091729323308271, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2405, + "step": 6179 + }, + { + "epoch": 0.07097744360902256, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 6180 + }, + { + "epoch": 0.0710375939849624, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.3108, + "step": 6181 + }, + { + "epoch": 0.07109774436090226, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2206, + "step": 6182 + }, + { + "epoch": 0.0711578947368421, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2345, + "step": 6183 + }, + { + "epoch": 0.07121804511278196, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1643, + "step": 6184 + }, + { + "epoch": 0.0712781954887218, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 6185 + }, + { + "epoch": 0.07133834586466166, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2533, + "step": 6186 + }, + { + "epoch": 0.0713984962406015, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3109, + "step": 6187 + }, + { + "epoch": 0.07145864661654136, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2342, + "step": 6188 + }, + { + "epoch": 0.0715187969924812, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 6189 + }, + { + "epoch": 0.07157894736842105, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.3158, + "step": 6190 + }, + { + "epoch": 0.0716390977443609, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 6191 + }, + { + "epoch": 0.07169924812030075, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2331, + "step": 6192 + }, + { + "epoch": 0.0717593984962406, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2381, + "step": 6193 + }, + { + "epoch": 0.07181954887218045, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2139, + "step": 6194 + }, + { + "epoch": 0.07187969924812031, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3019, + "step": 6195 + }, + { + "epoch": 0.07193984962406015, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2875, + "step": 6196 + }, + { + "epoch": 0.072, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1692, + "step": 6197 + }, + { + "epoch": 0.07206015037593985, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3032, + "step": 6198 + }, + { + "epoch": 0.0721203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2643, + "step": 6199 + }, + { + "epoch": 0.07218045112781955, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.3063, + "step": 6200 + }, + { + "epoch": 0.0722406015037594, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 6201 + }, + { + "epoch": 0.07230075187969925, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 1.3236, + "step": 6202 + }, + { + "epoch": 0.0723609022556391, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.3222, + "step": 6203 + }, + { + "epoch": 0.07242105263157894, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 6204 + }, + { + "epoch": 0.0724812030075188, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2189, + "step": 6205 + }, + { + "epoch": 0.07254135338345864, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2648, + "step": 6206 + }, + { + "epoch": 0.0726015037593985, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 6207 + }, + { + "epoch": 0.07266165413533834, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3089, + "step": 6208 + }, + { + "epoch": 0.0727218045112782, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.3306, + "step": 6209 + }, + { + "epoch": 0.07278195488721804, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2935, + "step": 6210 + }, + { + "epoch": 0.07284210526315789, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3442, + "step": 6211 + }, + { + "epoch": 0.07290225563909775, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2935, + "step": 6212 + }, + { + "epoch": 0.07296240601503759, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3285, + "step": 6213 + }, + { + "epoch": 0.07302255639097745, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2392, + "step": 6214 + }, + { + "epoch": 0.07308270676691729, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3185, + "step": 6215 + }, + { + "epoch": 0.07314285714285715, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 6216 + }, + { + "epoch": 0.07320300751879699, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1997, + "step": 6217 + }, + { + "epoch": 0.07326315789473684, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.314, + "step": 6218 + }, + { + "epoch": 0.0733233082706767, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2636, + "step": 6219 + }, + { + "epoch": 0.07338345864661654, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 6220 + }, + { + "epoch": 0.0734436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2183, + "step": 6221 + }, + { + "epoch": 0.07350375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2805, + "step": 6222 + }, + { + "epoch": 0.0735639097744361, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2659, + "step": 6223 + }, + { + "epoch": 0.07362406015037594, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2675, + "step": 6224 + }, + { + "epoch": 0.07368421052631578, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 6225 + }, + { + "epoch": 0.07374436090225564, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 6226 + }, + { + "epoch": 0.07380451127819548, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2435, + "step": 6227 + }, + { + "epoch": 0.07386466165413534, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3238, + "step": 6228 + }, + { + "epoch": 0.07392481203007518, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3275, + "step": 6229 + }, + { + "epoch": 0.07398496240601504, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2126, + "step": 6230 + }, + { + "epoch": 0.07404511278195489, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2556, + "step": 6231 + }, + { + "epoch": 0.07410526315789474, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2208, + "step": 6232 + }, + { + "epoch": 0.07416541353383459, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2621, + "step": 6233 + }, + { + "epoch": 0.07422556390977443, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2467, + "step": 6234 + }, + { + "epoch": 0.07428571428571429, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2537, + "step": 6235 + }, + { + "epoch": 0.07434586466165413, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2677, + "step": 6236 + }, + { + "epoch": 0.07440601503759399, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2604, + "step": 6237 + }, + { + "epoch": 0.07446616541353383, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2504, + "step": 6238 + }, + { + "epoch": 0.07452631578947369, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 6239 + }, + { + "epoch": 0.07458646616541353, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2223, + "step": 6240 + }, + { + "epoch": 0.07464661654135338, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2859, + "step": 6241 + }, + { + "epoch": 0.07470676691729324, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2473, + "step": 6242 + }, + { + "epoch": 0.07476691729323308, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 6243 + }, + { + "epoch": 0.07482706766917294, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2752, + "step": 6244 + }, + { + "epoch": 0.07488721804511278, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2373, + "step": 6245 + }, + { + "epoch": 0.07494736842105264, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2731, + "step": 6246 + }, + { + "epoch": 0.07500751879699248, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 6247 + }, + { + "epoch": 0.07506766917293232, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2319, + "step": 6248 + }, + { + "epoch": 0.07512781954887218, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1891, + "step": 6249 + }, + { + "epoch": 0.07518796992481203, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.3899, + "step": 6250 + }, + { + "epoch": 0.07524812030075188, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3694, + "step": 6251 + }, + { + "epoch": 0.07530827067669173, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2574, + "step": 6252 + }, + { + "epoch": 0.07536842105263158, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2765, + "step": 6253 + }, + { + "epoch": 0.07542857142857143, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.3148, + "step": 6254 + }, + { + "epoch": 0.07548872180451127, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 6255 + }, + { + "epoch": 0.07554887218045113, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2413, + "step": 6256 + }, + { + "epoch": 0.07560902255639097, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2883, + "step": 6257 + }, + { + "epoch": 0.07566917293233083, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2409, + "step": 6258 + }, + { + "epoch": 0.07572932330827067, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.3728, + "step": 6259 + }, + { + "epoch": 0.07578947368421053, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 6260 + }, + { + "epoch": 0.07584962406015038, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3535, + "step": 6261 + }, + { + "epoch": 0.07590977443609022, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1881, + "step": 6262 + }, + { + "epoch": 0.07596992481203008, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 6263 + }, + { + "epoch": 0.07603007518796992, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2386, + "step": 6264 + }, + { + "epoch": 0.07609022556390978, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 6265 + }, + { + "epoch": 0.07615037593984962, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 6266 + }, + { + "epoch": 0.07621052631578948, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.234, + "step": 6267 + }, + { + "epoch": 0.07627067669172932, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2523, + "step": 6268 + }, + { + "epoch": 0.07633082706766918, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 6269 + }, + { + "epoch": 0.07639097744360902, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2612, + "step": 6270 + }, + { + "epoch": 0.07645112781954887, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2542, + "step": 6271 + }, + { + "epoch": 0.07651127819548872, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3683, + "step": 6272 + }, + { + "epoch": 0.07657142857142857, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 6273 + }, + { + "epoch": 0.07663157894736843, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1837, + "step": 6274 + }, + { + "epoch": 0.07669172932330827, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 6275 + }, + { + "epoch": 0.07675187969924813, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2556, + "step": 6276 + }, + { + "epoch": 0.07681203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 6277 + }, + { + "epoch": 0.07687218045112781, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2498, + "step": 6278 + }, + { + "epoch": 0.07693233082706767, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3053, + "step": 6279 + }, + { + "epoch": 0.07699248120300752, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2478, + "step": 6280 + }, + { + "epoch": 0.07705263157894737, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 6281 + }, + { + "epoch": 0.07711278195488722, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2748, + "step": 6282 + }, + { + "epoch": 0.07717293233082707, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 6283 + }, + { + "epoch": 0.07723308270676692, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2809, + "step": 6284 + }, + { + "epoch": 0.07729323308270676, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 6285 + }, + { + "epoch": 0.07735338345864662, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3936, + "step": 6286 + }, + { + "epoch": 0.07741353383458646, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3165, + "step": 6287 + }, + { + "epoch": 0.07747368421052632, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 6288 + }, + { + "epoch": 0.07753383458646616, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3219, + "step": 6289 + }, + { + "epoch": 0.07759398496240602, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2552, + "step": 6290 + }, + { + "epoch": 0.07765413533834586, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 6291 + }, + { + "epoch": 0.07771428571428571, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2398, + "step": 6292 + }, + { + "epoch": 0.07777443609022557, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 6293 + }, + { + "epoch": 0.07783458646616541, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2043, + "step": 6294 + }, + { + "epoch": 0.07789473684210527, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2077, + "step": 6295 + }, + { + "epoch": 0.07795488721804511, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2263, + "step": 6296 + }, + { + "epoch": 0.07801503759398497, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 6297 + }, + { + "epoch": 0.07807518796992481, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1982, + "step": 6298 + }, + { + "epoch": 0.07813533834586465, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2573, + "step": 6299 + }, + { + "epoch": 0.07819548872180451, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3008, + "step": 6300 + }, + { + "epoch": 0.07825563909774436, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2852, + "step": 6301 + }, + { + "epoch": 0.07831578947368421, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.187, + "step": 6302 + }, + { + "epoch": 0.07837593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 6303 + }, + { + "epoch": 0.07843609022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 6304 + }, + { + "epoch": 0.07849624060150376, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 6305 + }, + { + "epoch": 0.07855639097744362, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3057, + "step": 6306 + }, + { + "epoch": 0.07861654135338346, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2535, + "step": 6307 + }, + { + "epoch": 0.0786766917293233, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 6308 + }, + { + "epoch": 0.07873684210526316, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3162, + "step": 6309 + }, + { + "epoch": 0.078796992481203, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 6310 + }, + { + "epoch": 0.07885714285714286, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 6311 + }, + { + "epoch": 0.0789172932330827, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.3076, + "step": 6312 + }, + { + "epoch": 0.07897744360902256, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2657, + "step": 6313 + }, + { + "epoch": 0.0790375939849624, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 6314 + }, + { + "epoch": 0.07909774436090225, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.304, + "step": 6315 + }, + { + "epoch": 0.07915789473684211, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2623, + "step": 6316 + }, + { + "epoch": 0.07921804511278195, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2864, + "step": 6317 + }, + { + "epoch": 0.07927819548872181, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2652, + "step": 6318 + }, + { + "epoch": 0.07933834586466165, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2296, + "step": 6319 + }, + { + "epoch": 0.07939849624060151, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2591, + "step": 6320 + }, + { + "epoch": 0.07945864661654135, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 6321 + }, + { + "epoch": 0.0795187969924812, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2493, + "step": 6322 + }, + { + "epoch": 0.07957894736842105, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 6323 + }, + { + "epoch": 0.0796390977443609, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.2367, + "step": 6324 + }, + { + "epoch": 0.07969924812030076, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2348, + "step": 6325 + }, + { + "epoch": 0.0797593984962406, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2462, + "step": 6326 + }, + { + "epoch": 0.07981954887218046, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2067, + "step": 6327 + }, + { + "epoch": 0.0798796992481203, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2117, + "step": 6328 + }, + { + "epoch": 0.07993984962406014, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 6329 + }, + { + "epoch": 0.08, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3135, + "step": 6330 + }, + { + "epoch": 0.08006015037593985, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2695, + "step": 6331 + }, + { + "epoch": 0.0801203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2569, + "step": 6332 + }, + { + "epoch": 0.08018045112781955, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2715, + "step": 6333 + }, + { + "epoch": 0.0802406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2279, + "step": 6334 + }, + { + "epoch": 0.08030075187969925, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2396, + "step": 6335 + }, + { + "epoch": 0.08036090225563909, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.3422, + "step": 6336 + }, + { + "epoch": 0.08042105263157895, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2134, + "step": 6337 + }, + { + "epoch": 0.08048120300751879, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2423, + "step": 6338 + }, + { + "epoch": 0.08054135338345865, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3295, + "step": 6339 + }, + { + "epoch": 0.0806015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2147, + "step": 6340 + }, + { + "epoch": 0.08066165413533835, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.334, + "step": 6341 + }, + { + "epoch": 0.0807218045112782, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 6342 + }, + { + "epoch": 0.08078195488721804, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2667, + "step": 6343 + }, + { + "epoch": 0.0808421052631579, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3067, + "step": 6344 + }, + { + "epoch": 0.08090225563909774, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2595, + "step": 6345 + }, + { + "epoch": 0.0809624060150376, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2821, + "step": 6346 + }, + { + "epoch": 0.08102255639097744, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 6347 + }, + { + "epoch": 0.0810827067669173, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.275, + "step": 6348 + }, + { + "epoch": 0.08114285714285714, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 6349 + }, + { + "epoch": 0.081203007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1904, + "step": 6350 + }, + { + "epoch": 0.08126315789473684, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2781, + "step": 6351 + }, + { + "epoch": 0.08132330827067669, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.227, + "step": 6352 + }, + { + "epoch": 0.08138345864661654, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3045, + "step": 6353 + }, + { + "epoch": 0.08144360902255639, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2653, + "step": 6354 + }, + { + "epoch": 0.08150375939849624, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.256, + "step": 6355 + }, + { + "epoch": 0.08156390977443609, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.2224, + "step": 6356 + }, + { + "epoch": 0.08162406015037595, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2686, + "step": 6357 + }, + { + "epoch": 0.08168421052631579, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 6358 + }, + { + "epoch": 0.08174436090225563, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2296, + "step": 6359 + }, + { + "epoch": 0.08180451127819549, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2618, + "step": 6360 + }, + { + "epoch": 0.08186466165413533, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2423, + "step": 6361 + }, + { + "epoch": 0.08192481203007519, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2661, + "step": 6362 + }, + { + "epoch": 0.08198496240601504, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2478, + "step": 6363 + }, + { + "epoch": 0.0820451127819549, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2389, + "step": 6364 + }, + { + "epoch": 0.08210526315789474, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3014, + "step": 6365 + }, + { + "epoch": 0.08216541353383458, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2661, + "step": 6366 + }, + { + "epoch": 0.08222556390977444, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2636, + "step": 6367 + }, + { + "epoch": 0.08228571428571428, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.5436, + "step": 6368 + }, + { + "epoch": 0.08234586466165414, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2172, + "step": 6369 + }, + { + "epoch": 0.08240601503759398, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2047, + "step": 6370 + }, + { + "epoch": 0.08246616541353384, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 6371 + }, + { + "epoch": 0.08252631578947368, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2212, + "step": 6372 + }, + { + "epoch": 0.08258646616541353, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 6373 + }, + { + "epoch": 0.08264661654135338, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2677, + "step": 6374 + }, + { + "epoch": 0.08270676691729323, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2283, + "step": 6375 + }, + { + "epoch": 0.08276691729323309, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3191, + "step": 6376 + }, + { + "epoch": 0.08282706766917293, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2514, + "step": 6377 + }, + { + "epoch": 0.08288721804511279, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2749, + "step": 6378 + }, + { + "epoch": 0.08294736842105263, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.329, + "step": 6379 + }, + { + "epoch": 0.08300751879699247, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2709, + "step": 6380 + }, + { + "epoch": 0.08306766917293233, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 6381 + }, + { + "epoch": 0.08312781954887218, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.249, + "step": 6382 + }, + { + "epoch": 0.08318796992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2994, + "step": 6383 + }, + { + "epoch": 0.08324812030075188, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2747, + "step": 6384 + }, + { + "epoch": 0.08330827067669173, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2805, + "step": 6385 + }, + { + "epoch": 0.08336842105263158, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.261, + "step": 6386 + }, + { + "epoch": 0.08342857142857144, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3089, + "step": 6387 + }, + { + "epoch": 0.08348872180451128, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2799, + "step": 6388 + }, + { + "epoch": 0.08354887218045112, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.2153, + "step": 6389 + }, + { + "epoch": 0.08360902255639098, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3295, + "step": 6390 + }, + { + "epoch": 0.08366917293233082, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2338, + "step": 6391 + }, + { + "epoch": 0.08372932330827068, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3108, + "step": 6392 + }, + { + "epoch": 0.08378947368421052, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2731, + "step": 6393 + }, + { + "epoch": 0.08384962406015038, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 6394 + }, + { + "epoch": 0.08390977443609023, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2524, + "step": 6395 + }, + { + "epoch": 0.08396992481203007, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 6396 + }, + { + "epoch": 0.08403007518796993, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2943, + "step": 6397 + }, + { + "epoch": 0.08409022556390977, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2759, + "step": 6398 + }, + { + "epoch": 0.08415037593984963, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 6399 + }, + { + "epoch": 0.08421052631578947, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2668, + "step": 6400 + }, + { + "epoch": 0.08427067669172933, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.226, + "step": 6401 + }, + { + "epoch": 0.08433082706766917, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.2907, + "step": 6402 + }, + { + "epoch": 0.08439097744360902, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.4053, + "step": 6403 + }, + { + "epoch": 0.08445112781954887, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 6404 + }, + { + "epoch": 0.08451127819548872, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.244, + "step": 6405 + }, + { + "epoch": 0.08457142857142858, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 6406 + }, + { + "epoch": 0.08463157894736842, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1839, + "step": 6407 + }, + { + "epoch": 0.08469172932330828, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 6408 + }, + { + "epoch": 0.08475187969924812, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.348, + "step": 6409 + }, + { + "epoch": 0.08481203007518796, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3669, + "step": 6410 + }, + { + "epoch": 0.08487218045112782, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.3588, + "step": 6411 + }, + { + "epoch": 0.08493233082706766, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1896, + "step": 6412 + }, + { + "epoch": 0.08499248120300752, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2279, + "step": 6413 + }, + { + "epoch": 0.08505263157894737, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.218, + "step": 6414 + }, + { + "epoch": 0.08511278195488722, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2624, + "step": 6415 + }, + { + "epoch": 0.08517293233082707, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2908, + "step": 6416 + }, + { + "epoch": 0.08523308270676691, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3357, + "step": 6417 + }, + { + "epoch": 0.08529323308270677, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2365, + "step": 6418 + }, + { + "epoch": 0.08535338345864661, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2297, + "step": 6419 + }, + { + "epoch": 0.08541353383458647, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.2791, + "step": 6420 + }, + { + "epoch": 0.08547368421052631, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 6421 + }, + { + "epoch": 0.08553383458646617, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1965, + "step": 6422 + }, + { + "epoch": 0.08559398496240601, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 6423 + }, + { + "epoch": 0.08565413533834587, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2626, + "step": 6424 + }, + { + "epoch": 0.08571428571428572, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.27, + "step": 6425 + }, + { + "epoch": 0.08577443609022556, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.333, + "step": 6426 + }, + { + "epoch": 0.08583458646616542, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2894, + "step": 6427 + }, + { + "epoch": 0.08589473684210526, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 6428 + }, + { + "epoch": 0.08595488721804512, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2257, + "step": 6429 + }, + { + "epoch": 0.08601503759398496, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3028, + "step": 6430 + }, + { + "epoch": 0.08607518796992482, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2044, + "step": 6431 + }, + { + "epoch": 0.08613533834586466, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2855, + "step": 6432 + }, + { + "epoch": 0.0861954887218045, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3485, + "step": 6433 + }, + { + "epoch": 0.08625563909774436, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2379, + "step": 6434 + }, + { + "epoch": 0.0863157894736842, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1836, + "step": 6435 + }, + { + "epoch": 0.08637593984962406, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2684, + "step": 6436 + }, + { + "epoch": 0.08643609022556391, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.261, + "step": 6437 + }, + { + "epoch": 0.08649624060150377, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 6438 + }, + { + "epoch": 0.08655639097744361, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1808, + "step": 6439 + }, + { + "epoch": 0.08661654135338345, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2737, + "step": 6440 + }, + { + "epoch": 0.08667669172932331, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.3605, + "step": 6441 + }, + { + "epoch": 0.08673684210526315, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 6442 + }, + { + "epoch": 0.08679699248120301, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 6443 + }, + { + "epoch": 0.08685714285714285, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 6444 + }, + { + "epoch": 0.08691729323308271, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2761, + "step": 6445 + }, + { + "epoch": 0.08697744360902256, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2519, + "step": 6446 + }, + { + "epoch": 0.0870375939849624, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1994, + "step": 6447 + }, + { + "epoch": 0.08709774436090226, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 6448 + }, + { + "epoch": 0.0871578947368421, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2526, + "step": 6449 + }, + { + "epoch": 0.08721804511278196, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2025, + "step": 6450 + }, + { + "epoch": 0.0872781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3693, + "step": 6451 + }, + { + "epoch": 0.08733834586466166, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3552, + "step": 6452 + }, + { + "epoch": 0.0873984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2414, + "step": 6453 + }, + { + "epoch": 0.08745864661654135, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 6454 + }, + { + "epoch": 0.0875187969924812, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2382, + "step": 6455 + }, + { + "epoch": 0.08757894736842105, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.3536, + "step": 6456 + }, + { + "epoch": 0.0876390977443609, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 6457 + }, + { + "epoch": 0.08769924812030075, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2289, + "step": 6458 + }, + { + "epoch": 0.0877593984962406, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2776, + "step": 6459 + }, + { + "epoch": 0.08781954887218045, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2244, + "step": 6460 + }, + { + "epoch": 0.08787969924812031, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2673, + "step": 6461 + }, + { + "epoch": 0.08793984962406015, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2937, + "step": 6462 + }, + { + "epoch": 0.088, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2602, + "step": 6463 + }, + { + "epoch": 0.08806015037593985, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2935, + "step": 6464 + }, + { + "epoch": 0.0881203007518797, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2039, + "step": 6465 + }, + { + "epoch": 0.08818045112781955, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3412, + "step": 6466 + }, + { + "epoch": 0.0882406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2153, + "step": 6467 + }, + { + "epoch": 0.08830075187969925, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2671, + "step": 6468 + }, + { + "epoch": 0.0883609022556391, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.263, + "step": 6469 + }, + { + "epoch": 0.08842105263157894, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2579, + "step": 6470 + }, + { + "epoch": 0.0884812030075188, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2045, + "step": 6471 + }, + { + "epoch": 0.08854135338345864, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2116, + "step": 6472 + }, + { + "epoch": 0.0886015037593985, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1907, + "step": 6473 + }, + { + "epoch": 0.08866165413533834, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2559, + "step": 6474 + }, + { + "epoch": 0.0887218045112782, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 6475 + }, + { + "epoch": 0.08878195488721805, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 6476 + }, + { + "epoch": 0.08884210526315789, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2621, + "step": 6477 + }, + { + "epoch": 0.08890225563909775, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3024, + "step": 6478 + }, + { + "epoch": 0.08896240601503759, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2173, + "step": 6479 + }, + { + "epoch": 0.08902255639097745, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2, + "step": 6480 + }, + { + "epoch": 0.08908270676691729, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 6481 + }, + { + "epoch": 0.08914285714285715, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.293, + "step": 6482 + }, + { + "epoch": 0.08920300751879699, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1925, + "step": 6483 + }, + { + "epoch": 0.08926315789473684, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2326, + "step": 6484 + }, + { + "epoch": 0.0893233082706767, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2182, + "step": 6485 + }, + { + "epoch": 0.08938345864661654, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.244, + "step": 6486 + }, + { + "epoch": 0.0894436090225564, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2587, + "step": 6487 + }, + { + "epoch": 0.08950375939849624, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 6488 + }, + { + "epoch": 0.0895639097744361, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2395, + "step": 6489 + }, + { + "epoch": 0.08962406015037594, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.252, + "step": 6490 + }, + { + "epoch": 0.08968421052631578, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2913, + "step": 6491 + }, + { + "epoch": 0.08974436090225564, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2697, + "step": 6492 + }, + { + "epoch": 0.08980451127819548, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1997, + "step": 6493 + }, + { + "epoch": 0.08986466165413534, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2728, + "step": 6494 + }, + { + "epoch": 0.08992481203007519, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.233, + "step": 6495 + }, + { + "epoch": 0.08998496240601504, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1927, + "step": 6496 + }, + { + "epoch": 0.09004511278195489, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2163, + "step": 6497 + }, + { + "epoch": 0.09010526315789473, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2139, + "step": 6498 + }, + { + "epoch": 0.09016541353383459, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2066, + "step": 6499 + }, + { + "epoch": 0.09022556390977443, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2755, + "step": 6500 + }, + { + "epoch": 0.09028571428571429, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 6501 + }, + { + "epoch": 0.09034586466165413, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3537, + "step": 6502 + }, + { + "epoch": 0.09040601503759399, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.2532, + "step": 6503 + }, + { + "epoch": 0.09046616541353383, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2367, + "step": 6504 + }, + { + "epoch": 0.09052631578947369, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2875, + "step": 6505 + }, + { + "epoch": 0.09058646616541353, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2982, + "step": 6506 + }, + { + "epoch": 0.09064661654135338, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3234, + "step": 6507 + }, + { + "epoch": 0.09070676691729324, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 6508 + }, + { + "epoch": 0.09076691729323308, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2759, + "step": 6509 + }, + { + "epoch": 0.09082706766917294, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2138, + "step": 6510 + }, + { + "epoch": 0.09088721804511278, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1623, + "step": 6511 + }, + { + "epoch": 0.09094736842105264, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.424, + "step": 6512 + }, + { + "epoch": 0.09100751879699248, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3211, + "step": 6513 + }, + { + "epoch": 0.09106766917293233, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2657, + "step": 6514 + }, + { + "epoch": 0.09112781954887218, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2358, + "step": 6515 + }, + { + "epoch": 0.09118796992481203, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2417, + "step": 6516 + }, + { + "epoch": 0.09124812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2709, + "step": 6517 + }, + { + "epoch": 0.09130827067669173, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3019, + "step": 6518 + }, + { + "epoch": 0.09136842105263158, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 6519 + }, + { + "epoch": 0.09142857142857143, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2876, + "step": 6520 + }, + { + "epoch": 0.09148872180451127, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.256, + "step": 6521 + }, + { + "epoch": 0.09154887218045113, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 6522 + }, + { + "epoch": 0.09160902255639097, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 6523 + }, + { + "epoch": 0.09166917293233083, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2391, + "step": 6524 + }, + { + "epoch": 0.09172932330827067, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 6525 + }, + { + "epoch": 0.09178947368421053, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2869, + "step": 6526 + }, + { + "epoch": 0.09184962406015038, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.279, + "step": 6527 + }, + { + "epoch": 0.09190977443609022, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.24, + "step": 6528 + }, + { + "epoch": 0.09196992481203008, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 6529 + }, + { + "epoch": 0.09203007518796992, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 6530 + }, + { + "epoch": 0.09209022556390978, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2423, + "step": 6531 + }, + { + "epoch": 0.09215037593984962, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2682, + "step": 6532 + }, + { + "epoch": 0.09221052631578948, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.275, + "step": 6533 + }, + { + "epoch": 0.09227067669172932, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2824, + "step": 6534 + }, + { + "epoch": 0.09233082706766917, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 6535 + }, + { + "epoch": 0.09239097744360902, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2829, + "step": 6536 + }, + { + "epoch": 0.09245112781954887, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.3102, + "step": 6537 + }, + { + "epoch": 0.09251127819548872, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2628, + "step": 6538 + }, + { + "epoch": 0.09257142857142857, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 6539 + }, + { + "epoch": 0.09263157894736843, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2736, + "step": 6540 + }, + { + "epoch": 0.09269172932330827, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2079, + "step": 6541 + }, + { + "epoch": 0.09275187969924813, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2479, + "step": 6542 + }, + { + "epoch": 0.09281203007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2257, + "step": 6543 + }, + { + "epoch": 0.09287218045112781, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 6544 + }, + { + "epoch": 0.09293233082706767, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 6545 + }, + { + "epoch": 0.09299248120300752, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2852, + "step": 6546 + }, + { + "epoch": 0.09305263157894737, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.2176, + "step": 6547 + }, + { + "epoch": 0.09311278195488722, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3001, + "step": 6548 + }, + { + "epoch": 0.09317293233082707, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2875, + "step": 6549 + }, + { + "epoch": 0.09323308270676692, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.2332, + "step": 6550 + }, + { + "epoch": 0.09329323308270676, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2457, + "step": 6551 + }, + { + "epoch": 0.09335338345864662, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 6552 + }, + { + "epoch": 0.09341353383458646, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3175, + "step": 6553 + }, + { + "epoch": 0.09347368421052632, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2251, + "step": 6554 + }, + { + "epoch": 0.09353383458646616, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.236, + "step": 6555 + }, + { + "epoch": 0.09359398496240602, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2134, + "step": 6556 + }, + { + "epoch": 0.09365413533834586, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.4124, + "step": 6557 + }, + { + "epoch": 0.09371428571428571, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2002, + "step": 6558 + }, + { + "epoch": 0.09377443609022557, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1718, + "step": 6559 + }, + { + "epoch": 0.09383458646616541, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2548, + "step": 6560 + }, + { + "epoch": 0.09389473684210527, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2206, + "step": 6561 + }, + { + "epoch": 0.09395488721804511, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2434, + "step": 6562 + }, + { + "epoch": 0.09401503759398497, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 6563 + }, + { + "epoch": 0.09407518796992481, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3324, + "step": 6564 + }, + { + "epoch": 0.09413533834586466, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.279, + "step": 6565 + }, + { + "epoch": 0.09419548872180451, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 6566 + }, + { + "epoch": 0.09425563909774436, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2281, + "step": 6567 + }, + { + "epoch": 0.09431578947368421, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2597, + "step": 6568 + }, + { + "epoch": 0.09437593984962406, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2398, + "step": 6569 + }, + { + "epoch": 0.09443609022556392, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 6570 + }, + { + "epoch": 0.09449624060150376, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 6571 + }, + { + "epoch": 0.0945563909774436, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3068, + "step": 6572 + }, + { + "epoch": 0.09461654135338346, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 6573 + }, + { + "epoch": 0.0946766917293233, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2779, + "step": 6574 + }, + { + "epoch": 0.09473684210526316, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2069, + "step": 6575 + }, + { + "epoch": 0.094796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2863, + "step": 6576 + }, + { + "epoch": 0.09485714285714286, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.261, + "step": 6577 + }, + { + "epoch": 0.0949172932330827, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2237, + "step": 6578 + }, + { + "epoch": 0.09497744360902256, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.2914, + "step": 6579 + }, + { + "epoch": 0.0950375939849624, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2966, + "step": 6580 + }, + { + "epoch": 0.09509774436090225, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3115, + "step": 6581 + }, + { + "epoch": 0.09515789473684211, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 6582 + }, + { + "epoch": 0.09521804511278195, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3405, + "step": 6583 + }, + { + "epoch": 0.09527819548872181, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 6584 + }, + { + "epoch": 0.09533834586466165, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2351, + "step": 6585 + }, + { + "epoch": 0.09539849624060151, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3469, + "step": 6586 + }, + { + "epoch": 0.09545864661654135, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.34, + "step": 6587 + }, + { + "epoch": 0.0955187969924812, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3424, + "step": 6588 + }, + { + "epoch": 0.09557894736842105, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2024, + "step": 6589 + }, + { + "epoch": 0.0956390977443609, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2267, + "step": 6590 + }, + { + "epoch": 0.09569924812030076, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2896, + "step": 6591 + }, + { + "epoch": 0.0957593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3041, + "step": 6592 + }, + { + "epoch": 0.09581954887218046, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.296, + "step": 6593 + }, + { + "epoch": 0.0958796992481203, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 6594 + }, + { + "epoch": 0.09593984962406014, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 6595 + }, + { + "epoch": 0.096, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2642, + "step": 6596 + }, + { + "epoch": 0.09606015037593985, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2382, + "step": 6597 + }, + { + "epoch": 0.0961203007518797, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2328, + "step": 6598 + }, + { + "epoch": 0.09618045112781955, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2252, + "step": 6599 + }, + { + "epoch": 0.0962406015037594, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2766, + "step": 6600 + }, + { + "epoch": 0.09630075187969925, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 6601 + }, + { + "epoch": 0.09636090225563909, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2349, + "step": 6602 + }, + { + "epoch": 0.09642105263157895, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2953, + "step": 6603 + }, + { + "epoch": 0.09648120300751879, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2635, + "step": 6604 + }, + { + "epoch": 0.09654135338345865, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2507, + "step": 6605 + }, + { + "epoch": 0.0966015037593985, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1728, + "step": 6606 + }, + { + "epoch": 0.09666165413533835, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2451, + "step": 6607 + }, + { + "epoch": 0.0967218045112782, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1763, + "step": 6608 + }, + { + "epoch": 0.09678195488721804, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 6609 + }, + { + "epoch": 0.0968421052631579, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 6610 + }, + { + "epoch": 0.09690225563909774, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2312, + "step": 6611 + }, + { + "epoch": 0.0969624060150376, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 6612 + }, + { + "epoch": 0.09702255639097744, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 6613 + }, + { + "epoch": 0.0970827067669173, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2619, + "step": 6614 + }, + { + "epoch": 0.09714285714285714, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3234, + "step": 6615 + }, + { + "epoch": 0.097203007518797, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.3346, + "step": 6616 + }, + { + "epoch": 0.09726315789473684, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.4646, + "step": 6617 + }, + { + "epoch": 0.09732330827067669, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2163, + "step": 6618 + }, + { + "epoch": 0.09738345864661654, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2444, + "step": 6619 + }, + { + "epoch": 0.09744360902255639, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 6620 + }, + { + "epoch": 0.09750375939849625, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 6621 + }, + { + "epoch": 0.09756390977443609, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2663, + "step": 6622 + }, + { + "epoch": 0.09762406015037595, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 6623 + }, + { + "epoch": 0.09768421052631579, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2703, + "step": 6624 + }, + { + "epoch": 0.09774436090225563, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 6625 + }, + { + "epoch": 0.09780451127819549, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2843, + "step": 6626 + }, + { + "epoch": 0.09786466165413533, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.228, + "step": 6627 + }, + { + "epoch": 0.09792481203007519, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 6628 + }, + { + "epoch": 0.09798496240601504, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2913, + "step": 6629 + }, + { + "epoch": 0.0980451127819549, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3385, + "step": 6630 + }, + { + "epoch": 0.09810526315789474, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 6631 + }, + { + "epoch": 0.09816541353383458, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2375, + "step": 6632 + }, + { + "epoch": 0.09822556390977444, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3277, + "step": 6633 + }, + { + "epoch": 0.09828571428571428, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 6634 + }, + { + "epoch": 0.09834586466165414, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 6635 + }, + { + "epoch": 0.09840601503759398, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.3148, + "step": 6636 + }, + { + "epoch": 0.09846616541353384, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3777, + "step": 6637 + }, + { + "epoch": 0.09852631578947368, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2315, + "step": 6638 + }, + { + "epoch": 0.09858646616541353, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2234, + "step": 6639 + }, + { + "epoch": 0.09864661654135339, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2343, + "step": 6640 + }, + { + "epoch": 0.09870676691729323, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2715, + "step": 6641 + }, + { + "epoch": 0.09876691729323309, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3119, + "step": 6642 + }, + { + "epoch": 0.09882706766917293, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2743, + "step": 6643 + }, + { + "epoch": 0.09888721804511279, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2911, + "step": 6644 + }, + { + "epoch": 0.09894736842105263, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2539, + "step": 6645 + }, + { + "epoch": 0.09900751879699247, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3909, + "step": 6646 + }, + { + "epoch": 0.09906766917293233, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.4861, + "step": 6647 + }, + { + "epoch": 0.09912781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2108, + "step": 6648 + }, + { + "epoch": 0.09918796992481203, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2745, + "step": 6649 + }, + { + "epoch": 0.09924812030075188, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1643, + "step": 6650 + }, + { + "epoch": 0.09930827067669173, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2516, + "step": 6651 + }, + { + "epoch": 0.09936842105263158, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2452, + "step": 6652 + }, + { + "epoch": 0.09942857142857142, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 6653 + }, + { + "epoch": 0.09948872180451128, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1965, + "step": 6654 + }, + { + "epoch": 0.09954887218045112, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1951, + "step": 6655 + }, + { + "epoch": 0.09960902255639098, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.227, + "step": 6656 + }, + { + "epoch": 0.09966917293233082, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2494, + "step": 6657 + }, + { + "epoch": 0.09972932330827068, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2079, + "step": 6658 + }, + { + "epoch": 0.09978947368421053, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2828, + "step": 6659 + }, + { + "epoch": 0.09984962406015038, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 6660 + }, + { + "epoch": 0.09990977443609023, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 6661 + }, + { + "epoch": 0.09996992481203007, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2151, + "step": 6662 + }, + { + "epoch": 0.10003007518796993, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2981, + "step": 6663 + }, + { + "epoch": 0.10009022556390977, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2431, + "step": 6664 + }, + { + "epoch": 0.10015037593984963, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1936, + "step": 6665 + }, + { + "epoch": 0.10021052631578947, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2757, + "step": 6666 + }, + { + "epoch": 0.10027067669172933, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 6667 + }, + { + "epoch": 0.10033082706766917, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3225, + "step": 6668 + }, + { + "epoch": 0.10039097744360902, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2024, + "step": 6669 + }, + { + "epoch": 0.10045112781954887, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2369, + "step": 6670 + }, + { + "epoch": 0.10051127819548872, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2852, + "step": 6671 + }, + { + "epoch": 0.10057142857142858, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2172, + "step": 6672 + }, + { + "epoch": 0.10063157894736842, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 6673 + }, + { + "epoch": 0.10069172932330828, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.226, + "step": 6674 + }, + { + "epoch": 0.10075187969924812, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 6675 + }, + { + "epoch": 0.10081203007518796, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 6676 + }, + { + "epoch": 0.10087218045112782, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2575, + "step": 6677 + }, + { + "epoch": 0.10093233082706767, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.3097, + "step": 6678 + }, + { + "epoch": 0.10099248120300752, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2027, + "step": 6679 + }, + { + "epoch": 0.10105263157894737, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2241, + "step": 6680 + }, + { + "epoch": 0.10111278195488722, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2246, + "step": 6681 + }, + { + "epoch": 0.10117293233082707, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2768, + "step": 6682 + }, + { + "epoch": 0.10123308270676691, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3113, + "step": 6683 + }, + { + "epoch": 0.10129323308270677, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2381, + "step": 6684 + }, + { + "epoch": 0.10135338345864661, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3472, + "step": 6685 + }, + { + "epoch": 0.10141353383458647, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2194, + "step": 6686 + }, + { + "epoch": 0.10147368421052631, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1852, + "step": 6687 + }, + { + "epoch": 0.10153383458646617, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2089, + "step": 6688 + }, + { + "epoch": 0.10159398496240601, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2393, + "step": 6689 + }, + { + "epoch": 0.10165413533834586, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1929, + "step": 6690 + }, + { + "epoch": 0.10171428571428572, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 6691 + }, + { + "epoch": 0.10177443609022556, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 6692 + }, + { + "epoch": 0.10183458646616542, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2734, + "step": 6693 + }, + { + "epoch": 0.10189473684210526, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3063, + "step": 6694 + }, + { + "epoch": 0.10195488721804512, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 6695 + }, + { + "epoch": 0.10201503759398496, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2383, + "step": 6696 + }, + { + "epoch": 0.10207518796992482, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3146, + "step": 6697 + }, + { + "epoch": 0.10213533834586466, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2363, + "step": 6698 + }, + { + "epoch": 0.1021954887218045, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2721, + "step": 6699 + }, + { + "epoch": 0.10225563909774436, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1839, + "step": 6700 + }, + { + "epoch": 0.10231578947368421, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 6701 + }, + { + "epoch": 0.10237593984962406, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2159, + "step": 6702 + }, + { + "epoch": 0.10243609022556391, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 6703 + }, + { + "epoch": 0.10249624060150377, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2245, + "step": 6704 + }, + { + "epoch": 0.10255639097744361, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.386, + "step": 6705 + }, + { + "epoch": 0.10261654135338345, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2908, + "step": 6706 + }, + { + "epoch": 0.10267669172932331, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.202, + "step": 6707 + }, + { + "epoch": 0.10273684210526315, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2013, + "step": 6708 + }, + { + "epoch": 0.10279699248120301, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3247, + "step": 6709 + }, + { + "epoch": 0.10285714285714286, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2668, + "step": 6710 + }, + { + "epoch": 0.10291729323308271, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3215, + "step": 6711 + }, + { + "epoch": 0.10297744360902256, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3066, + "step": 6712 + }, + { + "epoch": 0.1030375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2167, + "step": 6713 + }, + { + "epoch": 0.10309774436090226, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3932, + "step": 6714 + }, + { + "epoch": 0.1031578947368421, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2672, + "step": 6715 + }, + { + "epoch": 0.10321804511278196, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 6716 + }, + { + "epoch": 0.1032781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.5521, + "step": 6717 + }, + { + "epoch": 0.10333834586466166, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.2422, + "step": 6718 + }, + { + "epoch": 0.1033984962406015, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2929, + "step": 6719 + }, + { + "epoch": 0.10345864661654135, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2974, + "step": 6720 + }, + { + "epoch": 0.1035187969924812, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2156, + "step": 6721 + }, + { + "epoch": 0.10357894736842105, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2379, + "step": 6722 + }, + { + "epoch": 0.1036390977443609, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 6723 + }, + { + "epoch": 0.10369924812030075, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.2097, + "step": 6724 + }, + { + "epoch": 0.1037593984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2365, + "step": 6725 + }, + { + "epoch": 0.10381954887218045, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2729, + "step": 6726 + }, + { + "epoch": 0.1038796992481203, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2243, + "step": 6727 + }, + { + "epoch": 0.10393984962406015, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 6728 + }, + { + "epoch": 0.104, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2411, + "step": 6729 + }, + { + "epoch": 0.10406015037593985, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2233, + "step": 6730 + }, + { + "epoch": 0.1041203007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2827, + "step": 6731 + }, + { + "epoch": 0.10418045112781955, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2129, + "step": 6732 + }, + { + "epoch": 0.1042406015037594, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2952, + "step": 6733 + }, + { + "epoch": 0.10430075187969925, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1721, + "step": 6734 + }, + { + "epoch": 0.1043609022556391, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 6735 + }, + { + "epoch": 0.10442105263157894, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.297, + "step": 6736 + }, + { + "epoch": 0.1044812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2417, + "step": 6737 + }, + { + "epoch": 0.10454135338345864, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2753, + "step": 6738 + }, + { + "epoch": 0.1046015037593985, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2434, + "step": 6739 + }, + { + "epoch": 0.10466165413533834, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2452, + "step": 6740 + }, + { + "epoch": 0.1047218045112782, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1733, + "step": 6741 + }, + { + "epoch": 0.10478195488721805, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3325, + "step": 6742 + }, + { + "epoch": 0.10484210526315789, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2318, + "step": 6743 + }, + { + "epoch": 0.10490225563909775, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1991, + "step": 6744 + }, + { + "epoch": 0.10496240601503759, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3174, + "step": 6745 + }, + { + "epoch": 0.10502255639097745, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2464, + "step": 6746 + }, + { + "epoch": 0.10508270676691729, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3352, + "step": 6747 + }, + { + "epoch": 0.10514285714285715, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2202, + "step": 6748 + }, + { + "epoch": 0.10520300751879699, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 6749 + }, + { + "epoch": 0.10526315789473684, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 6750 + }, + { + "epoch": 0.1053233082706767, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2251, + "step": 6751 + }, + { + "epoch": 0.10538345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 6752 + }, + { + "epoch": 0.1054436090225564, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 6753 + }, + { + "epoch": 0.10550375939849624, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2201, + "step": 6754 + }, + { + "epoch": 0.1055639097744361, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 6755 + }, + { + "epoch": 0.10562406015037594, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1903, + "step": 6756 + }, + { + "epoch": 0.10568421052631578, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2867, + "step": 6757 + }, + { + "epoch": 0.10574436090225564, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 6758 + }, + { + "epoch": 0.10580451127819548, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2044, + "step": 6759 + }, + { + "epoch": 0.10586466165413534, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2007, + "step": 6760 + }, + { + "epoch": 0.10592481203007519, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2165, + "step": 6761 + }, + { + "epoch": 0.10598496240601504, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2503, + "step": 6762 + }, + { + "epoch": 0.10604511278195489, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2593, + "step": 6763 + }, + { + "epoch": 0.10610526315789473, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2582, + "step": 6764 + }, + { + "epoch": 0.10616541353383459, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3647, + "step": 6765 + }, + { + "epoch": 0.10622556390977443, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.276, + "step": 6766 + }, + { + "epoch": 0.10628571428571429, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2936, + "step": 6767 + }, + { + "epoch": 0.10634586466165413, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.292, + "step": 6768 + }, + { + "epoch": 0.10640601503759399, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 6769 + }, + { + "epoch": 0.10646616541353383, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2548, + "step": 6770 + }, + { + "epoch": 0.10652631578947368, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3339, + "step": 6771 + }, + { + "epoch": 0.10658646616541353, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2523, + "step": 6772 + }, + { + "epoch": 0.10664661654135338, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3487, + "step": 6773 + }, + { + "epoch": 0.10670676691729324, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2248, + "step": 6774 + }, + { + "epoch": 0.10676691729323308, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 6775 + }, + { + "epoch": 0.10682706766917294, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1911, + "step": 6776 + }, + { + "epoch": 0.10688721804511278, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 6777 + }, + { + "epoch": 0.10694736842105264, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1949, + "step": 6778 + }, + { + "epoch": 0.10700751879699248, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 6779 + }, + { + "epoch": 0.10706766917293233, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.219, + "step": 6780 + }, + { + "epoch": 0.10712781954887218, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 6781 + }, + { + "epoch": 0.10718796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4509, + "step": 6782 + }, + { + "epoch": 0.10724812030075188, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 6783 + }, + { + "epoch": 0.10730827067669173, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2612, + "step": 6784 + }, + { + "epoch": 0.10736842105263159, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2807, + "step": 6785 + }, + { + "epoch": 0.10742857142857143, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.248, + "step": 6786 + }, + { + "epoch": 0.10748872180451127, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 6787 + }, + { + "epoch": 0.10754887218045113, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2685, + "step": 6788 + }, + { + "epoch": 0.10760902255639097, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2564, + "step": 6789 + }, + { + "epoch": 0.10766917293233083, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.3108, + "step": 6790 + }, + { + "epoch": 0.10772932330827067, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2655, + "step": 6791 + }, + { + "epoch": 0.10778947368421053, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.224, + "step": 6792 + }, + { + "epoch": 0.10784962406015038, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3315, + "step": 6793 + }, + { + "epoch": 0.10790977443609022, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2865, + "step": 6794 + }, + { + "epoch": 0.10796992481203008, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2418, + "step": 6795 + }, + { + "epoch": 0.10803007518796992, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2781, + "step": 6796 + }, + { + "epoch": 0.10809022556390978, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2456, + "step": 6797 + }, + { + "epoch": 0.10815037593984962, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.31, + "step": 6798 + }, + { + "epoch": 0.10821052631578948, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 6799 + }, + { + "epoch": 0.10827067669172932, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 6800 + }, + { + "epoch": 0.10833082706766917, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2267, + "step": 6801 + }, + { + "epoch": 0.10839097744360902, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2212, + "step": 6802 + }, + { + "epoch": 0.10845112781954887, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3581, + "step": 6803 + }, + { + "epoch": 0.10851127819548873, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 6804 + }, + { + "epoch": 0.10857142857142857, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.3066, + "step": 6805 + }, + { + "epoch": 0.10863157894736843, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2786, + "step": 6806 + }, + { + "epoch": 0.10869172932330827, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2028, + "step": 6807 + }, + { + "epoch": 0.10875187969924811, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3404, + "step": 6808 + }, + { + "epoch": 0.10881203007518797, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2353, + "step": 6809 + }, + { + "epoch": 0.10887218045112781, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1805, + "step": 6810 + }, + { + "epoch": 0.10893233082706767, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2179, + "step": 6811 + }, + { + "epoch": 0.10899248120300752, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 1.276, + "step": 6812 + }, + { + "epoch": 0.10905263157894737, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2352, + "step": 6813 + }, + { + "epoch": 0.10911278195488722, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2939, + "step": 6814 + }, + { + "epoch": 0.10917293233082707, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2582, + "step": 6815 + }, + { + "epoch": 0.10923308270676692, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2099, + "step": 6816 + }, + { + "epoch": 0.10929323308270676, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 6817 + }, + { + "epoch": 0.10935338345864662, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3158, + "step": 6818 + }, + { + "epoch": 0.10941353383458646, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 6819 + }, + { + "epoch": 0.10947368421052632, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2302, + "step": 6820 + }, + { + "epoch": 0.10953383458646616, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.189, + "step": 6821 + }, + { + "epoch": 0.10959398496240602, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2462, + "step": 6822 + }, + { + "epoch": 0.10965413533834587, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.4624, + "step": 6823 + }, + { + "epoch": 0.10971428571428571, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2094, + "step": 6824 + }, + { + "epoch": 0.10977443609022557, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1986, + "step": 6825 + }, + { + "epoch": 0.10983458646616541, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3376, + "step": 6826 + }, + { + "epoch": 0.10989473684210527, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2526, + "step": 6827 + }, + { + "epoch": 0.10995488721804511, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2956, + "step": 6828 + }, + { + "epoch": 0.11001503759398497, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1711, + "step": 6829 + }, + { + "epoch": 0.11007518796992481, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2889, + "step": 6830 + }, + { + "epoch": 0.11013533834586466, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 6831 + }, + { + "epoch": 0.11019548872180451, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.227, + "step": 6832 + }, + { + "epoch": 0.11025563909774436, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2367, + "step": 6833 + }, + { + "epoch": 0.11031578947368421, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2284, + "step": 6834 + }, + { + "epoch": 0.11037593984962406, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.3261, + "step": 6835 + }, + { + "epoch": 0.11043609022556392, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2197, + "step": 6836 + }, + { + "epoch": 0.11049624060150376, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3047, + "step": 6837 + }, + { + "epoch": 0.1105563909774436, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 6838 + }, + { + "epoch": 0.11061654135338346, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 6839 + }, + { + "epoch": 0.1106766917293233, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 6840 + }, + { + "epoch": 0.11073684210526316, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3216, + "step": 6841 + }, + { + "epoch": 0.110796992481203, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.228, + "step": 6842 + }, + { + "epoch": 0.11085714285714286, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.293, + "step": 6843 + }, + { + "epoch": 0.1109172932330827, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 6844 + }, + { + "epoch": 0.11097744360902255, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3454, + "step": 6845 + }, + { + "epoch": 0.11103759398496241, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 6846 + }, + { + "epoch": 0.11109774436090225, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 6847 + }, + { + "epoch": 0.11115789473684211, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.211, + "step": 6848 + }, + { + "epoch": 0.11121804511278195, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2538, + "step": 6849 + }, + { + "epoch": 0.11127819548872181, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2982, + "step": 6850 + }, + { + "epoch": 0.11133834586466165, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2288, + "step": 6851 + }, + { + "epoch": 0.11139849624060151, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2519, + "step": 6852 + }, + { + "epoch": 0.11145864661654135, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 6853 + }, + { + "epoch": 0.1115187969924812, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.189, + "step": 6854 + }, + { + "epoch": 0.11157894736842106, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.308, + "step": 6855 + }, + { + "epoch": 0.1116390977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3783, + "step": 6856 + }, + { + "epoch": 0.11169924812030076, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2237, + "step": 6857 + }, + { + "epoch": 0.1117593984962406, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 6858 + }, + { + "epoch": 0.11181954887218046, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3718, + "step": 6859 + }, + { + "epoch": 0.1118796992481203, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.344, + "step": 6860 + }, + { + "epoch": 0.11193984962406014, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2287, + "step": 6861 + }, + { + "epoch": 0.112, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.2824, + "step": 6862 + }, + { + "epoch": 0.11206015037593985, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 6863 + }, + { + "epoch": 0.1121203007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2244, + "step": 6864 + }, + { + "epoch": 0.11218045112781955, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2806, + "step": 6865 + }, + { + "epoch": 0.1122406015037594, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2123, + "step": 6866 + }, + { + "epoch": 0.11230075187969925, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2858, + "step": 6867 + }, + { + "epoch": 0.11236090225563909, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.4164, + "step": 6868 + }, + { + "epoch": 0.11242105263157895, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2819, + "step": 6869 + }, + { + "epoch": 0.11248120300751879, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2705, + "step": 6870 + }, + { + "epoch": 0.11254135338345865, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.274, + "step": 6871 + }, + { + "epoch": 0.1126015037593985, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2097, + "step": 6872 + }, + { + "epoch": 0.11266165413533835, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 6873 + }, + { + "epoch": 0.1127218045112782, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 6874 + }, + { + "epoch": 0.11278195488721804, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2119, + "step": 6875 + }, + { + "epoch": 0.1128421052631579, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3272, + "step": 6876 + }, + { + "epoch": 0.11290225563909774, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 6877 + }, + { + "epoch": 0.1129624060150376, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.312, + "step": 6878 + }, + { + "epoch": 0.11302255639097744, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 6879 + }, + { + "epoch": 0.1130827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 6880 + }, + { + "epoch": 0.11314285714285714, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 6881 + }, + { + "epoch": 0.11320300751879699, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 6882 + }, + { + "epoch": 0.11326315789473684, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2729, + "step": 6883 + }, + { + "epoch": 0.11332330827067669, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.329, + "step": 6884 + }, + { + "epoch": 0.11338345864661654, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1836, + "step": 6885 + }, + { + "epoch": 0.11344360902255639, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3121, + "step": 6886 + }, + { + "epoch": 0.11350375939849625, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2429, + "step": 6887 + }, + { + "epoch": 0.11356390977443609, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 6888 + }, + { + "epoch": 0.11362406015037595, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 6889 + }, + { + "epoch": 0.11368421052631579, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2838, + "step": 6890 + }, + { + "epoch": 0.11374436090225563, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3221, + "step": 6891 + }, + { + "epoch": 0.11380451127819549, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2822, + "step": 6892 + }, + { + "epoch": 0.11386466165413534, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2167, + "step": 6893 + }, + { + "epoch": 0.11392481203007519, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2363, + "step": 6894 + }, + { + "epoch": 0.11398496240601504, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1776, + "step": 6895 + }, + { + "epoch": 0.1140451127819549, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2543, + "step": 6896 + }, + { + "epoch": 0.11410526315789474, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2485, + "step": 6897 + }, + { + "epoch": 0.11416541353383458, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.2177, + "step": 6898 + }, + { + "epoch": 0.11422556390977444, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1868, + "step": 6899 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2304, + "step": 6900 + }, + { + "epoch": 0.11434586466165414, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.298, + "step": 6901 + }, + { + "epoch": 0.11440601503759398, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2251, + "step": 6902 + }, + { + "epoch": 0.11446616541353384, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3038, + "step": 6903 + }, + { + "epoch": 0.11452631578947368, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.4632, + "step": 6904 + }, + { + "epoch": 0.11458646616541353, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2394, + "step": 6905 + }, + { + "epoch": 0.11464661654135339, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2025, + "step": 6906 + }, + { + "epoch": 0.11470676691729323, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2212, + "step": 6907 + }, + { + "epoch": 0.11476691729323309, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 6908 + }, + { + "epoch": 0.11482706766917293, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2514, + "step": 6909 + }, + { + "epoch": 0.11488721804511279, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.237, + "step": 6910 + }, + { + "epoch": 0.11494736842105263, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2653, + "step": 6911 + }, + { + "epoch": 0.11500751879699248, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2369, + "step": 6912 + }, + { + "epoch": 0.11506766917293233, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 6913 + }, + { + "epoch": 0.11512781954887218, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2377, + "step": 6914 + }, + { + "epoch": 0.11518796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 6915 + }, + { + "epoch": 0.11524812030075188, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 6916 + }, + { + "epoch": 0.11530827067669173, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2607, + "step": 6917 + }, + { + "epoch": 0.11536842105263158, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3402, + "step": 6918 + }, + { + "epoch": 0.11542857142857142, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.3611, + "step": 6919 + }, + { + "epoch": 0.11548872180451128, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.271, + "step": 6920 + }, + { + "epoch": 0.11554887218045112, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3193, + "step": 6921 + }, + { + "epoch": 0.11560902255639098, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2187, + "step": 6922 + }, + { + "epoch": 0.11566917293233082, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 6923 + }, + { + "epoch": 0.11572932330827068, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 6924 + }, + { + "epoch": 0.11578947368421053, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.235, + "step": 6925 + }, + { + "epoch": 0.11584962406015037, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 6926 + }, + { + "epoch": 0.11590977443609023, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.202, + "step": 6927 + }, + { + "epoch": 0.11596992481203007, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.3039, + "step": 6928 + }, + { + "epoch": 0.11603007518796993, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2907, + "step": 6929 + }, + { + "epoch": 0.11609022556390977, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2893, + "step": 6930 + }, + { + "epoch": 0.11615037593984963, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.3321, + "step": 6931 + }, + { + "epoch": 0.11621052631578947, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.327, + "step": 6932 + }, + { + "epoch": 0.11627067669172933, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2769, + "step": 6933 + }, + { + "epoch": 0.11633082706766917, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2578, + "step": 6934 + }, + { + "epoch": 0.11639097744360902, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2112, + "step": 6935 + }, + { + "epoch": 0.11645112781954887, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2692, + "step": 6936 + }, + { + "epoch": 0.11651127819548872, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2976, + "step": 6937 + }, + { + "epoch": 0.11657142857142858, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.3915, + "step": 6938 + }, + { + "epoch": 0.11663157894736842, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 6939 + }, + { + "epoch": 0.11669172932330828, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.2549, + "step": 6940 + }, + { + "epoch": 0.11675187969924812, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2893, + "step": 6941 + }, + { + "epoch": 0.11681203007518796, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.2289, + "step": 6942 + }, + { + "epoch": 0.11687218045112782, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 6943 + }, + { + "epoch": 0.11693233082706767, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2498, + "step": 6944 + }, + { + "epoch": 0.11699248120300752, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2797, + "step": 6945 + }, + { + "epoch": 0.11705263157894737, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3007, + "step": 6946 + }, + { + "epoch": 0.11711278195488722, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2082, + "step": 6947 + }, + { + "epoch": 0.11717293233082707, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.203, + "step": 6948 + }, + { + "epoch": 0.11723308270676691, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 6949 + }, + { + "epoch": 0.11729323308270677, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 6950 + }, + { + "epoch": 0.11735338345864661, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 6951 + }, + { + "epoch": 0.11741353383458647, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 6952 + }, + { + "epoch": 0.11747368421052631, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 6953 + }, + { + "epoch": 0.11753383458646617, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 6954 + }, + { + "epoch": 0.11759398496240601, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2011, + "step": 6955 + }, + { + "epoch": 0.11765413533834586, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2389, + "step": 6956 + }, + { + "epoch": 0.11771428571428572, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 6957 + }, + { + "epoch": 0.11777443609022556, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.245, + "step": 6958 + }, + { + "epoch": 0.11783458646616542, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2765, + "step": 6959 + }, + { + "epoch": 0.11789473684210526, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.3199, + "step": 6960 + }, + { + "epoch": 0.11795488721804512, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2856, + "step": 6961 + }, + { + "epoch": 0.11801503759398496, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2112, + "step": 6962 + }, + { + "epoch": 0.1180751879699248, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2493, + "step": 6963 + }, + { + "epoch": 0.11813533834586466, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3257, + "step": 6964 + }, + { + "epoch": 0.1181954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3166, + "step": 6965 + }, + { + "epoch": 0.11825563909774436, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 6966 + }, + { + "epoch": 0.11831578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2079, + "step": 6967 + }, + { + "epoch": 0.11837593984962407, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1972, + "step": 6968 + }, + { + "epoch": 0.11843609022556391, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2329, + "step": 6969 + }, + { + "epoch": 0.11849624060150377, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 6970 + }, + { + "epoch": 0.11855639097744361, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2459, + "step": 6971 + }, + { + "epoch": 0.11861654135338345, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 6972 + }, + { + "epoch": 0.11867669172932331, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 6973 + }, + { + "epoch": 0.11873684210526315, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2433, + "step": 6974 + }, + { + "epoch": 0.11879699248120301, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2516, + "step": 6975 + }, + { + "epoch": 0.11885714285714286, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 6976 + }, + { + "epoch": 0.11891729323308271, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.295, + "step": 6977 + }, + { + "epoch": 0.11897744360902256, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 6978 + }, + { + "epoch": 0.1190375939849624, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2537, + "step": 6979 + }, + { + "epoch": 0.11909774436090226, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2559, + "step": 6980 + }, + { + "epoch": 0.1191578947368421, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 6981 + }, + { + "epoch": 0.11921804511278196, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 6982 + }, + { + "epoch": 0.1192781954887218, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2333, + "step": 6983 + }, + { + "epoch": 0.11933834586466166, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3196, + "step": 6984 + }, + { + "epoch": 0.1193984962406015, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2839, + "step": 6985 + }, + { + "epoch": 0.11945864661654135, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 6986 + }, + { + "epoch": 0.1195187969924812, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 6987 + }, + { + "epoch": 0.11957894736842105, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 6988 + }, + { + "epoch": 0.1196390977443609, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 6989 + }, + { + "epoch": 0.11969924812030075, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2989, + "step": 6990 + }, + { + "epoch": 0.11975939849624061, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 6991 + }, + { + "epoch": 0.11981954887218045, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2692, + "step": 6992 + }, + { + "epoch": 0.1198796992481203, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3454, + "step": 6993 + }, + { + "epoch": 0.11993984962406015, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 6994 + }, + { + "epoch": 0.12, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 6995 + }, + { + "epoch": 0.12006015037593985, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2209, + "step": 6996 + }, + { + "epoch": 0.1201203007518797, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2744, + "step": 6997 + }, + { + "epoch": 0.12018045112781955, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2745, + "step": 6998 + }, + { + "epoch": 0.1202406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 6999 + }, + { + "epoch": 0.12030075187969924, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.3609, + "step": 7000 + }, + { + "epoch": 0.1203609022556391, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2046, + "step": 7001 + }, + { + "epoch": 0.12042105263157894, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2108, + "step": 7002 + }, + { + "epoch": 0.1204812030075188, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 7003 + }, + { + "epoch": 0.12054135338345864, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 7004 + }, + { + "epoch": 0.1206015037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2164, + "step": 7005 + }, + { + "epoch": 0.12066165413533834, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2451, + "step": 7006 + }, + { + "epoch": 0.1207218045112782, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2107, + "step": 7007 + }, + { + "epoch": 0.12078195488721805, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2582, + "step": 7008 + }, + { + "epoch": 0.12084210526315789, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2576, + "step": 7009 + }, + { + "epoch": 0.12090225563909775, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3092, + "step": 7010 + }, + { + "epoch": 0.12096240601503759, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2032, + "step": 7011 + }, + { + "epoch": 0.12102255639097745, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 7012 + }, + { + "epoch": 0.12108270676691729, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 7013 + }, + { + "epoch": 0.12114285714285715, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2335, + "step": 7014 + }, + { + "epoch": 0.12120300751879699, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 7015 + }, + { + "epoch": 0.12126315789473684, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.3576, + "step": 7016 + }, + { + "epoch": 0.1213233082706767, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3188, + "step": 7017 + }, + { + "epoch": 0.12138345864661654, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2804, + "step": 7018 + }, + { + "epoch": 0.1214436090225564, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.315, + "step": 7019 + }, + { + "epoch": 0.12150375939849624, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 7020 + }, + { + "epoch": 0.1215639097744361, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 7021 + }, + { + "epoch": 0.12162406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 7022 + }, + { + "epoch": 0.12168421052631578, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2446, + "step": 7023 + }, + { + "epoch": 0.12174436090225564, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 7024 + }, + { + "epoch": 0.12180451127819548, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3069, + "step": 7025 + }, + { + "epoch": 0.12186466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 7026 + }, + { + "epoch": 0.12192481203007519, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2502, + "step": 7027 + }, + { + "epoch": 0.12198496240601504, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2021, + "step": 7028 + }, + { + "epoch": 0.12204511278195489, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2636, + "step": 7029 + }, + { + "epoch": 0.12210526315789473, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 7030 + }, + { + "epoch": 0.12216541353383459, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 7031 + }, + { + "epoch": 0.12222556390977443, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1936, + "step": 7032 + }, + { + "epoch": 0.12228571428571429, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 7033 + }, + { + "epoch": 0.12234586466165413, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3868, + "step": 7034 + }, + { + "epoch": 0.12240601503759399, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1845, + "step": 7035 + }, + { + "epoch": 0.12246616541353383, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2514, + "step": 7036 + }, + { + "epoch": 0.12252631578947368, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 7037 + }, + { + "epoch": 0.12258646616541354, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 7038 + }, + { + "epoch": 0.12264661654135338, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 7039 + }, + { + "epoch": 0.12270676691729324, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2864, + "step": 7040 + }, + { + "epoch": 0.12276691729323308, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 7041 + }, + { + "epoch": 0.12282706766917294, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.3604, + "step": 7042 + }, + { + "epoch": 0.12288721804511278, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 7043 + }, + { + "epoch": 0.12294736842105262, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3038, + "step": 7044 + }, + { + "epoch": 0.12300751879699248, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 7045 + }, + { + "epoch": 0.12306766917293233, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 7046 + }, + { + "epoch": 0.12312781954887218, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3004, + "step": 7047 + }, + { + "epoch": 0.12318796992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2114, + "step": 7048 + }, + { + "epoch": 0.12324812030075188, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 7049 + }, + { + "epoch": 0.12330827067669173, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2802, + "step": 7050 + }, + { + "epoch": 0.12336842105263159, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 7051 + }, + { + "epoch": 0.12342857142857143, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2635, + "step": 7052 + }, + { + "epoch": 0.12348872180451127, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2561, + "step": 7053 + }, + { + "epoch": 0.12354887218045113, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 7054 + }, + { + "epoch": 0.12360902255639097, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2554, + "step": 7055 + }, + { + "epoch": 0.12366917293233083, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2301, + "step": 7056 + }, + { + "epoch": 0.12372932330827068, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2185, + "step": 7057 + }, + { + "epoch": 0.12378947368421053, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2518, + "step": 7058 + }, + { + "epoch": 0.12384962406015038, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 7059 + }, + { + "epoch": 0.12390977443609022, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2333, + "step": 7060 + }, + { + "epoch": 0.12396992481203008, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2063, + "step": 7061 + }, + { + "epoch": 0.12403007518796992, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 7062 + }, + { + "epoch": 0.12409022556390978, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3029, + "step": 7063 + }, + { + "epoch": 0.12415037593984962, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1953, + "step": 7064 + }, + { + "epoch": 0.12421052631578948, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2449, + "step": 7065 + }, + { + "epoch": 0.12427067669172932, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2166, + "step": 7066 + }, + { + "epoch": 0.12433082706766917, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 7067 + }, + { + "epoch": 0.12439097744360902, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.197, + "step": 7068 + }, + { + "epoch": 0.12445112781954887, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 7069 + }, + { + "epoch": 0.12451127819548873, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 7070 + }, + { + "epoch": 0.12457142857142857, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3483, + "step": 7071 + }, + { + "epoch": 0.12463157894736843, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2798, + "step": 7072 + }, + { + "epoch": 0.12469172932330827, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2848, + "step": 7073 + }, + { + "epoch": 0.12475187969924811, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2486, + "step": 7074 + }, + { + "epoch": 0.12481203007518797, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 7075 + }, + { + "epoch": 0.12487218045112781, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 7076 + }, + { + "epoch": 0.12493233082706767, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2583, + "step": 7077 + }, + { + "epoch": 0.12499248120300752, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2144, + "step": 7078 + }, + { + "epoch": 0.12505263157894736, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3304, + "step": 7079 + }, + { + "epoch": 0.12511278195488723, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2084, + "step": 7080 + }, + { + "epoch": 0.12517293233082707, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 7081 + }, + { + "epoch": 0.12523308270676692, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.3425, + "step": 7082 + }, + { + "epoch": 0.12529323308270676, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2661, + "step": 7083 + }, + { + "epoch": 0.1253533834586466, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 7084 + }, + { + "epoch": 0.12541353383458648, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2486, + "step": 7085 + }, + { + "epoch": 0.12547368421052632, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2242, + "step": 7086 + }, + { + "epoch": 0.12553383458646616, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.3789, + "step": 7087 + }, + { + "epoch": 0.125593984962406, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.309, + "step": 7088 + }, + { + "epoch": 0.12565413533834585, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1651, + "step": 7089 + }, + { + "epoch": 0.12571428571428572, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 7090 + }, + { + "epoch": 0.12577443609022557, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2508, + "step": 7091 + }, + { + "epoch": 0.1258345864661654, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3279, + "step": 7092 + }, + { + "epoch": 0.12589473684210525, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3942, + "step": 7093 + }, + { + "epoch": 0.12595488721804513, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 7094 + }, + { + "epoch": 0.12601503759398497, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2301, + "step": 7095 + }, + { + "epoch": 0.1260751879699248, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2881, + "step": 7096 + }, + { + "epoch": 0.12613533834586466, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 7097 + }, + { + "epoch": 0.1261954887218045, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2436, + "step": 7098 + }, + { + "epoch": 0.12625563909774437, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2772, + "step": 7099 + }, + { + "epoch": 0.12631578947368421, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2035, + "step": 7100 + }, + { + "epoch": 0.12637593984962406, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 7101 + }, + { + "epoch": 0.1264360902255639, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2585, + "step": 7102 + }, + { + "epoch": 0.12649624060150377, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2357, + "step": 7103 + }, + { + "epoch": 0.12655639097744362, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 7104 + }, + { + "epoch": 0.12661654135338346, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2638, + "step": 7105 + }, + { + "epoch": 0.1266766917293233, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3228, + "step": 7106 + }, + { + "epoch": 0.12673684210526315, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2388, + "step": 7107 + }, + { + "epoch": 0.12679699248120302, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2604, + "step": 7108 + }, + { + "epoch": 0.12685714285714286, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2434, + "step": 7109 + }, + { + "epoch": 0.1269172932330827, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.1763, + "step": 7110 + }, + { + "epoch": 0.12697744360902255, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2066, + "step": 7111 + }, + { + "epoch": 0.1270375939849624, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2435, + "step": 7112 + }, + { + "epoch": 0.12709774436090227, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 7113 + }, + { + "epoch": 0.1271578947368421, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2466, + "step": 7114 + }, + { + "epoch": 0.12721804511278195, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 7115 + }, + { + "epoch": 0.1272781954887218, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2641, + "step": 7116 + }, + { + "epoch": 0.12733834586466167, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.3581, + "step": 7117 + }, + { + "epoch": 0.1273984962406015, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2714, + "step": 7118 + }, + { + "epoch": 0.12745864661654135, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2925, + "step": 7119 + }, + { + "epoch": 0.1275187969924812, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2614, + "step": 7120 + }, + { + "epoch": 0.12757894736842104, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1834, + "step": 7121 + }, + { + "epoch": 0.1276390977443609, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2459, + "step": 7122 + }, + { + "epoch": 0.12769924812030076, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2926, + "step": 7123 + }, + { + "epoch": 0.1277593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2607, + "step": 7124 + }, + { + "epoch": 0.12781954887218044, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 7125 + }, + { + "epoch": 0.1278796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2835, + "step": 7126 + }, + { + "epoch": 0.12793984962406016, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3073, + "step": 7127 + }, + { + "epoch": 0.128, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1808, + "step": 7128 + }, + { + "epoch": 0.12806015037593985, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1954, + "step": 7129 + }, + { + "epoch": 0.1281203007518797, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 7130 + }, + { + "epoch": 0.12818045112781956, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2945, + "step": 7131 + }, + { + "epoch": 0.1282406015037594, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 7132 + }, + { + "epoch": 0.12830075187969925, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.2468, + "step": 7133 + }, + { + "epoch": 0.1283609022556391, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.2521, + "step": 7134 + }, + { + "epoch": 0.12842105263157894, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2942, + "step": 7135 + }, + { + "epoch": 0.1284812030075188, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 7136 + }, + { + "epoch": 0.12854135338345865, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 7137 + }, + { + "epoch": 0.1286015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2061, + "step": 7138 + }, + { + "epoch": 0.12866165413533834, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 7139 + }, + { + "epoch": 0.12872180451127818, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 7140 + }, + { + "epoch": 0.12878195488721805, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 7141 + }, + { + "epoch": 0.1288421052631579, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.333, + "step": 7142 + }, + { + "epoch": 0.12890225563909774, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2661, + "step": 7143 + }, + { + "epoch": 0.12896240601503758, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2781, + "step": 7144 + }, + { + "epoch": 0.12902255639097746, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.2456, + "step": 7145 + }, + { + "epoch": 0.1290827067669173, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 7146 + }, + { + "epoch": 0.12914285714285714, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.243, + "step": 7147 + }, + { + "epoch": 0.129203007518797, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2561, + "step": 7148 + }, + { + "epoch": 0.12926315789473683, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2987, + "step": 7149 + }, + { + "epoch": 0.1293233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2001, + "step": 7150 + }, + { + "epoch": 0.12938345864661654, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2669, + "step": 7151 + }, + { + "epoch": 0.1294436090225564, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2013, + "step": 7152 + }, + { + "epoch": 0.12950375939849623, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2253, + "step": 7153 + }, + { + "epoch": 0.1295639097744361, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2655, + "step": 7154 + }, + { + "epoch": 0.12962406015037595, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 7155 + }, + { + "epoch": 0.1296842105263158, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2192, + "step": 7156 + }, + { + "epoch": 0.12974436090225563, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 7157 + }, + { + "epoch": 0.12980451127819548, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3119, + "step": 7158 + }, + { + "epoch": 0.12986466165413535, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 7159 + }, + { + "epoch": 0.1299248120300752, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 7160 + }, + { + "epoch": 0.12998496240601504, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.2241, + "step": 7161 + }, + { + "epoch": 0.13004511278195488, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2828, + "step": 7162 + }, + { + "epoch": 0.13010526315789472, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2196, + "step": 7163 + }, + { + "epoch": 0.1301654135338346, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2486, + "step": 7164 + }, + { + "epoch": 0.13022556390977444, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2872, + "step": 7165 + }, + { + "epoch": 0.13028571428571428, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1865, + "step": 7166 + }, + { + "epoch": 0.13034586466165413, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.2578, + "step": 7167 + }, + { + "epoch": 0.130406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 7168 + }, + { + "epoch": 0.13046616541353384, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1653, + "step": 7169 + }, + { + "epoch": 0.13052631578947368, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3152, + "step": 7170 + }, + { + "epoch": 0.13058646616541353, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2511, + "step": 7171 + }, + { + "epoch": 0.13064661654135337, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.2476, + "step": 7172 + }, + { + "epoch": 0.13070676691729324, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2299, + "step": 7173 + }, + { + "epoch": 0.1307669172932331, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3038, + "step": 7174 + }, + { + "epoch": 0.13082706766917293, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2208, + "step": 7175 + }, + { + "epoch": 0.13088721804511277, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1728, + "step": 7176 + }, + { + "epoch": 0.13094736842105262, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1652, + "step": 7177 + }, + { + "epoch": 0.1310075187969925, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2791, + "step": 7178 + }, + { + "epoch": 0.13106766917293233, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3, + "step": 7179 + }, + { + "epoch": 0.13112781954887218, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3205, + "step": 7180 + }, + { + "epoch": 0.13118796992481202, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 7181 + }, + { + "epoch": 0.1312481203007519, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2341, + "step": 7182 + }, + { + "epoch": 0.13130827067669174, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2427, + "step": 7183 + }, + { + "epoch": 0.13136842105263158, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2053, + "step": 7184 + }, + { + "epoch": 0.13142857142857142, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2382, + "step": 7185 + }, + { + "epoch": 0.13148872180451127, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 7186 + }, + { + "epoch": 0.13154887218045114, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2496, + "step": 7187 + }, + { + "epoch": 0.13160902255639098, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2508, + "step": 7188 + }, + { + "epoch": 0.13166917293233082, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2494, + "step": 7189 + }, + { + "epoch": 0.13172932330827067, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3275, + "step": 7190 + }, + { + "epoch": 0.13178947368421054, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2482, + "step": 7191 + }, + { + "epoch": 0.13184962406015038, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2086, + "step": 7192 + }, + { + "epoch": 0.13190977443609023, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.2256, + "step": 7193 + }, + { + "epoch": 0.13196992481203007, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.316, + "step": 7194 + }, + { + "epoch": 0.13203007518796991, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2683, + "step": 7195 + }, + { + "epoch": 0.13209022556390979, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.2184, + "step": 7196 + }, + { + "epoch": 0.13215037593984963, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2819, + "step": 7197 + }, + { + "epoch": 0.13221052631578947, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2366, + "step": 7198 + }, + { + "epoch": 0.13227067669172932, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2591, + "step": 7199 + }, + { + "epoch": 0.13233082706766916, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.3455, + "step": 7200 + }, + { + "epoch": 0.13239097744360903, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3675, + "step": 7201 + }, + { + "epoch": 0.13245112781954888, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 7202 + }, + { + "epoch": 0.13251127819548872, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2831, + "step": 7203 + }, + { + "epoch": 0.13257142857142856, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2467, + "step": 7204 + }, + { + "epoch": 0.13263157894736843, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3245, + "step": 7205 + }, + { + "epoch": 0.13269172932330828, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2334, + "step": 7206 + }, + { + "epoch": 0.13275187969924812, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2092, + "step": 7207 + }, + { + "epoch": 0.13281203007518796, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2667, + "step": 7208 + }, + { + "epoch": 0.1328721804511278, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 7209 + }, + { + "epoch": 0.13293233082706768, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 7210 + }, + { + "epoch": 0.13299248120300752, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3138, + "step": 7211 + }, + { + "epoch": 0.13305263157894737, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2208, + "step": 7212 + }, + { + "epoch": 0.1331127819548872, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 7213 + }, + { + "epoch": 0.13317293233082705, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2865, + "step": 7214 + }, + { + "epoch": 0.13323308270676693, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2632, + "step": 7215 + }, + { + "epoch": 0.13329323308270677, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 7216 + }, + { + "epoch": 0.1333533834586466, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 7217 + }, + { + "epoch": 0.13341353383458646, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2363, + "step": 7218 + }, + { + "epoch": 0.13347368421052633, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3055, + "step": 7219 + }, + { + "epoch": 0.13353383458646617, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2102, + "step": 7220 + }, + { + "epoch": 0.13359398496240601, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 7221 + }, + { + "epoch": 0.13365413533834586, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1897, + "step": 7222 + }, + { + "epoch": 0.1337142857142857, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2902, + "step": 7223 + }, + { + "epoch": 0.13377443609022557, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2163, + "step": 7224 + }, + { + "epoch": 0.13383458646616542, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2358, + "step": 7225 + }, + { + "epoch": 0.13389473684210526, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.224, + "step": 7226 + }, + { + "epoch": 0.1339548872180451, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 7227 + }, + { + "epoch": 0.13401503759398498, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1917, + "step": 7228 + }, + { + "epoch": 0.13407518796992482, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2597, + "step": 7229 + }, + { + "epoch": 0.13413533834586466, + "grad_norm": 0.451171875, + "learning_rate": 3e-05, + "loss": 1.2253, + "step": 7230 + }, + { + "epoch": 0.1341954887218045, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 7231 + }, + { + "epoch": 0.13425563909774435, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.455, + "step": 7232 + }, + { + "epoch": 0.13431578947368422, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.22, + "step": 7233 + }, + { + "epoch": 0.13437593984962407, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2784, + "step": 7234 + }, + { + "epoch": 0.1344360902255639, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2977, + "step": 7235 + }, + { + "epoch": 0.13449624060150375, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3228, + "step": 7236 + }, + { + "epoch": 0.1345563909774436, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2552, + "step": 7237 + }, + { + "epoch": 0.13461654135338347, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 7238 + }, + { + "epoch": 0.1346766917293233, + "grad_norm": 1.9140625, + "learning_rate": 3e-05, + "loss": 1.2492, + "step": 7239 + }, + { + "epoch": 0.13473684210526315, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 7240 + }, + { + "epoch": 0.134796992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3107, + "step": 7241 + }, + { + "epoch": 0.13485714285714287, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 7242 + }, + { + "epoch": 0.1349172932330827, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 7243 + }, + { + "epoch": 0.13497744360902256, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 7244 + }, + { + "epoch": 0.1350375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2252, + "step": 7245 + }, + { + "epoch": 0.13509774436090224, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2929, + "step": 7246 + }, + { + "epoch": 0.13515789473684212, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2489, + "step": 7247 + }, + { + "epoch": 0.13521804511278196, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.3393, + "step": 7248 + }, + { + "epoch": 0.1352781954887218, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.251, + "step": 7249 + }, + { + "epoch": 0.13533834586466165, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.2329, + "step": 7250 + }, + { + "epoch": 0.1353984962406015, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2737, + "step": 7251 + }, + { + "epoch": 0.13545864661654136, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2505, + "step": 7252 + }, + { + "epoch": 0.1355187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2679, + "step": 7253 + }, + { + "epoch": 0.13557894736842105, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2494, + "step": 7254 + }, + { + "epoch": 0.1356390977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 7255 + }, + { + "epoch": 0.13569924812030076, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 7256 + }, + { + "epoch": 0.1357593984962406, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 7257 + }, + { + "epoch": 0.13581954887218045, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2667, + "step": 7258 + }, + { + "epoch": 0.1358796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 7259 + }, + { + "epoch": 0.13593984962406014, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1671, + "step": 7260 + }, + { + "epoch": 0.136, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2451, + "step": 7261 + }, + { + "epoch": 0.13606015037593985, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3253, + "step": 7262 + }, + { + "epoch": 0.1361203007518797, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2151, + "step": 7263 + }, + { + "epoch": 0.13618045112781954, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 7264 + }, + { + "epoch": 0.1362406015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 7265 + }, + { + "epoch": 0.13630075187969926, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 7266 + }, + { + "epoch": 0.1363609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2558, + "step": 7267 + }, + { + "epoch": 0.13642105263157894, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 7268 + }, + { + "epoch": 0.1364812030075188, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2056, + "step": 7269 + }, + { + "epoch": 0.13654135338345866, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2764, + "step": 7270 + }, + { + "epoch": 0.1366015037593985, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2178, + "step": 7271 + }, + { + "epoch": 0.13666165413533835, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2301, + "step": 7272 + }, + { + "epoch": 0.1367218045112782, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 7273 + }, + { + "epoch": 0.13678195488721803, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.259, + "step": 7274 + }, + { + "epoch": 0.1368421052631579, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 7275 + }, + { + "epoch": 0.13690225563909775, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 7276 + }, + { + "epoch": 0.1369624060150376, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 7277 + }, + { + "epoch": 0.13702255639097743, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2417, + "step": 7278 + }, + { + "epoch": 0.1370827067669173, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2505, + "step": 7279 + }, + { + "epoch": 0.13714285714285715, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.288, + "step": 7280 + }, + { + "epoch": 0.137203007518797, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1997, + "step": 7281 + }, + { + "epoch": 0.13726315789473684, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 7282 + }, + { + "epoch": 0.13732330827067668, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 7283 + }, + { + "epoch": 0.13738345864661655, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2194, + "step": 7284 + }, + { + "epoch": 0.1374436090225564, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 7285 + }, + { + "epoch": 0.13750375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2981, + "step": 7286 + }, + { + "epoch": 0.13756390977443608, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 1.2908, + "step": 7287 + }, + { + "epoch": 0.13762406015037593, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 7288 + }, + { + "epoch": 0.1376842105263158, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.3623, + "step": 7289 + }, + { + "epoch": 0.13774436090225564, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2701, + "step": 7290 + }, + { + "epoch": 0.13780451127819549, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 7291 + }, + { + "epoch": 0.13786466165413533, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2236, + "step": 7292 + }, + { + "epoch": 0.1379248120300752, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2488, + "step": 7293 + }, + { + "epoch": 0.13798496240601504, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.3082, + "step": 7294 + }, + { + "epoch": 0.1380451127819549, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.285, + "step": 7295 + }, + { + "epoch": 0.13810526315789473, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.218, + "step": 7296 + }, + { + "epoch": 0.13816541353383457, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2417, + "step": 7297 + }, + { + "epoch": 0.13822556390977445, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2626, + "step": 7298 + }, + { + "epoch": 0.1382857142857143, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2101, + "step": 7299 + }, + { + "epoch": 0.13834586466165413, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 7300 + }, + { + "epoch": 0.13840601503759398, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 7301 + }, + { + "epoch": 0.13846616541353385, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 7302 + }, + { + "epoch": 0.1385263157894737, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2226, + "step": 7303 + }, + { + "epoch": 0.13858646616541354, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.2332, + "step": 7304 + }, + { + "epoch": 0.13864661654135338, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2062, + "step": 7305 + }, + { + "epoch": 0.13870676691729322, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2445, + "step": 7306 + }, + { + "epoch": 0.1387669172932331, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2647, + "step": 7307 + }, + { + "epoch": 0.13882706766917294, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2739, + "step": 7308 + }, + { + "epoch": 0.13888721804511278, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2334, + "step": 7309 + }, + { + "epoch": 0.13894736842105262, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2938, + "step": 7310 + }, + { + "epoch": 0.13900751879699247, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3313, + "step": 7311 + }, + { + "epoch": 0.13906766917293234, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2629, + "step": 7312 + }, + { + "epoch": 0.13912781954887218, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1989, + "step": 7313 + }, + { + "epoch": 0.13918796992481203, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.4617, + "step": 7314 + }, + { + "epoch": 0.13924812030075187, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2724, + "step": 7315 + }, + { + "epoch": 0.13930827067669174, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2376, + "step": 7316 + }, + { + "epoch": 0.13936842105263159, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2211, + "step": 7317 + }, + { + "epoch": 0.13942857142857143, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.203, + "step": 7318 + }, + { + "epoch": 0.13948872180451127, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 7319 + }, + { + "epoch": 0.13954887218045112, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2903, + "step": 7320 + }, + { + "epoch": 0.139609022556391, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 7321 + }, + { + "epoch": 0.13966917293233083, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2937, + "step": 7322 + }, + { + "epoch": 0.13972932330827068, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 7323 + }, + { + "epoch": 0.13978947368421052, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2515, + "step": 7324 + }, + { + "epoch": 0.13984962406015036, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 7325 + }, + { + "epoch": 0.13990977443609023, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.3325, + "step": 7326 + }, + { + "epoch": 0.13996992481203008, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.3536, + "step": 7327 + }, + { + "epoch": 0.14003007518796992, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 7328 + }, + { + "epoch": 0.14009022556390976, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3074, + "step": 7329 + }, + { + "epoch": 0.14015037593984964, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2212, + "step": 7330 + }, + { + "epoch": 0.14021052631578948, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2779, + "step": 7331 + }, + { + "epoch": 0.14027067669172932, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2129, + "step": 7332 + }, + { + "epoch": 0.14033082706766917, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2488, + "step": 7333 + }, + { + "epoch": 0.140390977443609, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3127, + "step": 7334 + }, + { + "epoch": 0.14045112781954888, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2335, + "step": 7335 + }, + { + "epoch": 0.14051127819548873, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.3195, + "step": 7336 + }, + { + "epoch": 0.14057142857142857, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 7337 + }, + { + "epoch": 0.1406315789473684, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 7338 + }, + { + "epoch": 0.14069172932330828, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2195, + "step": 7339 + }, + { + "epoch": 0.14075187969924813, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2029, + "step": 7340 + }, + { + "epoch": 0.14081203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2768, + "step": 7341 + }, + { + "epoch": 0.14087218045112782, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2651, + "step": 7342 + }, + { + "epoch": 0.14093233082706766, + "grad_norm": 0.447265625, + "learning_rate": 3e-05, + "loss": 1.2228, + "step": 7343 + }, + { + "epoch": 0.14099248120300753, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2706, + "step": 7344 + }, + { + "epoch": 0.14105263157894737, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2347, + "step": 7345 + }, + { + "epoch": 0.14111278195488722, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2876, + "step": 7346 + }, + { + "epoch": 0.14117293233082706, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2791, + "step": 7347 + }, + { + "epoch": 0.1412330827067669, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2317, + "step": 7348 + }, + { + "epoch": 0.14129323308270678, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 7349 + }, + { + "epoch": 0.14135338345864662, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 7350 + }, + { + "epoch": 0.14141353383458646, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.238, + "step": 7351 + }, + { + "epoch": 0.1414736842105263, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2711, + "step": 7352 + }, + { + "epoch": 0.14153383458646618, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 7353 + }, + { + "epoch": 0.14159398496240602, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 7354 + }, + { + "epoch": 0.14165413533834587, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2582, + "step": 7355 + }, + { + "epoch": 0.1417142857142857, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2176, + "step": 7356 + }, + { + "epoch": 0.14177443609022555, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 7357 + }, + { + "epoch": 0.14183458646616542, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 7358 + }, + { + "epoch": 0.14189473684210527, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2376, + "step": 7359 + }, + { + "epoch": 0.1419548872180451, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.3237, + "step": 7360 + }, + { + "epoch": 0.14201503759398496, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2205, + "step": 7361 + }, + { + "epoch": 0.1420751879699248, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3067, + "step": 7362 + }, + { + "epoch": 0.14213533834586467, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2151, + "step": 7363 + }, + { + "epoch": 0.1421954887218045, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2868, + "step": 7364 + }, + { + "epoch": 0.14225563909774436, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.267, + "step": 7365 + }, + { + "epoch": 0.1423157894736842, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 7366 + }, + { + "epoch": 0.14237593984962407, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2953, + "step": 7367 + }, + { + "epoch": 0.14243609022556392, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3298, + "step": 7368 + }, + { + "epoch": 0.14249624060150376, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2678, + "step": 7369 + }, + { + "epoch": 0.1425563909774436, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2568, + "step": 7370 + }, + { + "epoch": 0.14261654135338345, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.1843, + "step": 7371 + }, + { + "epoch": 0.14267669172932332, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 7372 + }, + { + "epoch": 0.14273684210526316, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2795, + "step": 7373 + }, + { + "epoch": 0.142796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3084, + "step": 7374 + }, + { + "epoch": 0.14285714285714285, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2662, + "step": 7375 + }, + { + "epoch": 0.14291729323308272, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 7376 + }, + { + "epoch": 0.14297744360902256, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 7377 + }, + { + "epoch": 0.1430375939849624, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 7378 + }, + { + "epoch": 0.14309774436090225, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3121, + "step": 7379 + }, + { + "epoch": 0.1431578947368421, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.223, + "step": 7380 + }, + { + "epoch": 0.14321804511278197, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2398, + "step": 7381 + }, + { + "epoch": 0.1432781954887218, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1955, + "step": 7382 + }, + { + "epoch": 0.14333834586466165, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2234, + "step": 7383 + }, + { + "epoch": 0.1433984962406015, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2456, + "step": 7384 + }, + { + "epoch": 0.14345864661654134, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2665, + "step": 7385 + }, + { + "epoch": 0.1435187969924812, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 7386 + }, + { + "epoch": 0.14357894736842106, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 7387 + }, + { + "epoch": 0.1436390977443609, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2698, + "step": 7388 + }, + { + "epoch": 0.14369924812030074, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 7389 + }, + { + "epoch": 0.14375939849624061, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.224, + "step": 7390 + }, + { + "epoch": 0.14381954887218046, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.4712, + "step": 7391 + }, + { + "epoch": 0.1438796992481203, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2413, + "step": 7392 + }, + { + "epoch": 0.14393984962406015, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 7393 + }, + { + "epoch": 0.144, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.203, + "step": 7394 + }, + { + "epoch": 0.14406015037593986, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2228, + "step": 7395 + }, + { + "epoch": 0.1441203007518797, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3381, + "step": 7396 + }, + { + "epoch": 0.14418045112781955, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 7397 + }, + { + "epoch": 0.1442406015037594, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 7398 + }, + { + "epoch": 0.14430075187969924, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2638, + "step": 7399 + }, + { + "epoch": 0.1443609022556391, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.4038, + "step": 7400 + }, + { + "epoch": 0.14442105263157895, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 7401 + }, + { + "epoch": 0.1444812030075188, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2369, + "step": 7402 + }, + { + "epoch": 0.14454135338345864, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 7403 + }, + { + "epoch": 0.1446015037593985, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1997, + "step": 7404 + }, + { + "epoch": 0.14466165413533835, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 7405 + }, + { + "epoch": 0.1447218045112782, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2302, + "step": 7406 + }, + { + "epoch": 0.14478195488721804, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 7407 + }, + { + "epoch": 0.14484210526315788, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1959, + "step": 7408 + }, + { + "epoch": 0.14490225563909775, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 7409 + }, + { + "epoch": 0.1449624060150376, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 7410 + }, + { + "epoch": 0.14502255639097744, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2162, + "step": 7411 + }, + { + "epoch": 0.14508270676691729, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2961, + "step": 7412 + }, + { + "epoch": 0.14514285714285713, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2707, + "step": 7413 + }, + { + "epoch": 0.145203007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2586, + "step": 7414 + }, + { + "epoch": 0.14526315789473684, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.299, + "step": 7415 + }, + { + "epoch": 0.1453233082706767, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2978, + "step": 7416 + }, + { + "epoch": 0.14538345864661653, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1908, + "step": 7417 + }, + { + "epoch": 0.1454436090225564, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1888, + "step": 7418 + }, + { + "epoch": 0.14550375939849625, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 7419 + }, + { + "epoch": 0.1455639097744361, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 7420 + }, + { + "epoch": 0.14562406015037593, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3101, + "step": 7421 + }, + { + "epoch": 0.14568421052631578, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2736, + "step": 7422 + }, + { + "epoch": 0.14574436090225565, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2065, + "step": 7423 + }, + { + "epoch": 0.1458045112781955, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2096, + "step": 7424 + }, + { + "epoch": 0.14586466165413534, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3279, + "step": 7425 + }, + { + "epoch": 0.14592481203007518, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2819, + "step": 7426 + }, + { + "epoch": 0.14598496240601505, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 7427 + }, + { + "epoch": 0.1460451127819549, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2278, + "step": 7428 + }, + { + "epoch": 0.14610526315789474, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2719, + "step": 7429 + }, + { + "epoch": 0.14616541353383458, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3346, + "step": 7430 + }, + { + "epoch": 0.14622556390977443, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2367, + "step": 7431 + }, + { + "epoch": 0.1462857142857143, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 7432 + }, + { + "epoch": 0.14634586466165414, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.284, + "step": 7433 + }, + { + "epoch": 0.14640601503759398, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2463, + "step": 7434 + }, + { + "epoch": 0.14646616541353383, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2314, + "step": 7435 + }, + { + "epoch": 0.14652631578947367, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 7436 + }, + { + "epoch": 0.14658646616541354, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 7437 + }, + { + "epoch": 0.1466466165413534, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2717, + "step": 7438 + }, + { + "epoch": 0.14670676691729323, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2301, + "step": 7439 + }, + { + "epoch": 0.14676691729323307, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 7440 + }, + { + "epoch": 0.14682706766917294, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2664, + "step": 7441 + }, + { + "epoch": 0.1468872180451128, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1832, + "step": 7442 + }, + { + "epoch": 0.14694736842105263, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2516, + "step": 7443 + }, + { + "epoch": 0.14700751879699248, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2191, + "step": 7444 + }, + { + "epoch": 0.14706766917293232, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.2887, + "step": 7445 + }, + { + "epoch": 0.1471278195488722, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.2347, + "step": 7446 + }, + { + "epoch": 0.14718796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3438, + "step": 7447 + }, + { + "epoch": 0.14724812030075188, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2121, + "step": 7448 + }, + { + "epoch": 0.14730827067669172, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 7449 + }, + { + "epoch": 0.14736842105263157, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.2598, + "step": 7450 + }, + { + "epoch": 0.14742857142857144, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 7451 + }, + { + "epoch": 0.14748872180451128, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 7452 + }, + { + "epoch": 0.14754887218045112, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2585, + "step": 7453 + }, + { + "epoch": 0.14760902255639097, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.265, + "step": 7454 + }, + { + "epoch": 0.14766917293233084, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 7455 + }, + { + "epoch": 0.14772932330827068, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2398, + "step": 7456 + }, + { + "epoch": 0.14778947368421053, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2045, + "step": 7457 + }, + { + "epoch": 0.14784962406015037, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1633, + "step": 7458 + }, + { + "epoch": 0.1479097744360902, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.234, + "step": 7459 + }, + { + "epoch": 0.14796992481203008, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 7460 + }, + { + "epoch": 0.14803007518796993, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2856, + "step": 7461 + }, + { + "epoch": 0.14809022556390977, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1972, + "step": 7462 + }, + { + "epoch": 0.14815037593984962, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 7463 + }, + { + "epoch": 0.1482105263157895, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 7464 + }, + { + "epoch": 0.14827067669172933, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2145, + "step": 7465 + }, + { + "epoch": 0.14833082706766917, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2846, + "step": 7466 + }, + { + "epoch": 0.14839097744360902, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2373, + "step": 7467 + }, + { + "epoch": 0.14845112781954886, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2312, + "step": 7468 + }, + { + "epoch": 0.14851127819548873, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.2498, + "step": 7469 + }, + { + "epoch": 0.14857142857142858, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2331, + "step": 7470 + }, + { + "epoch": 0.14863157894736842, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2201, + "step": 7471 + }, + { + "epoch": 0.14869172932330826, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 7472 + }, + { + "epoch": 0.1487518796992481, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.223, + "step": 7473 + }, + { + "epoch": 0.14881203007518798, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2972, + "step": 7474 + }, + { + "epoch": 0.14887218045112782, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 7475 + }, + { + "epoch": 0.14893233082706767, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2765, + "step": 7476 + }, + { + "epoch": 0.1489924812030075, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3235, + "step": 7477 + }, + { + "epoch": 0.14905263157894738, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 7478 + }, + { + "epoch": 0.14911278195488722, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 7479 + }, + { + "epoch": 0.14917293233082707, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 7480 + }, + { + "epoch": 0.1492330827067669, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 7481 + }, + { + "epoch": 0.14929323308270676, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2099, + "step": 7482 + }, + { + "epoch": 0.14935338345864663, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2472, + "step": 7483 + }, + { + "epoch": 0.14941353383458647, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2264, + "step": 7484 + }, + { + "epoch": 0.14947368421052631, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 7485 + }, + { + "epoch": 0.14953383458646616, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2785, + "step": 7486 + }, + { + "epoch": 0.149593984962406, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2287, + "step": 7487 + }, + { + "epoch": 0.14965413533834587, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2715, + "step": 7488 + }, + { + "epoch": 0.14971428571428572, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2957, + "step": 7489 + }, + { + "epoch": 0.14977443609022556, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2752, + "step": 7490 + }, + { + "epoch": 0.1498345864661654, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2326, + "step": 7491 + }, + { + "epoch": 0.14989473684210528, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1911, + "step": 7492 + }, + { + "epoch": 0.14995488721804512, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2433, + "step": 7493 + }, + { + "epoch": 0.15001503759398496, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 7494 + }, + { + "epoch": 0.1500751879699248, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 7495 + }, + { + "epoch": 0.15013533834586465, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.233, + "step": 7496 + }, + { + "epoch": 0.15019548872180452, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 7497 + }, + { + "epoch": 0.15025563909774436, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 7498 + }, + { + "epoch": 0.1503157894736842, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 7499 + }, + { + "epoch": 0.15037593984962405, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2328, + "step": 7500 + }, + { + "epoch": 0.15043609022556392, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2013, + "step": 7501 + }, + { + "epoch": 0.15049624060150377, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.242, + "step": 7502 + }, + { + "epoch": 0.1505563909774436, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2861, + "step": 7503 + }, + { + "epoch": 0.15061654135338345, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 7504 + }, + { + "epoch": 0.1506766917293233, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 7505 + }, + { + "epoch": 0.15073684210526317, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.301, + "step": 7506 + }, + { + "epoch": 0.150796992481203, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 7507 + }, + { + "epoch": 0.15085714285714286, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 7508 + }, + { + "epoch": 0.1509172932330827, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2867, + "step": 7509 + }, + { + "epoch": 0.15097744360902254, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 7510 + }, + { + "epoch": 0.15103759398496241, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 7511 + }, + { + "epoch": 0.15109774436090226, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2756, + "step": 7512 + }, + { + "epoch": 0.1511578947368421, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1928, + "step": 7513 + }, + { + "epoch": 0.15121804511278195, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 7514 + }, + { + "epoch": 0.15127819548872182, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3135, + "step": 7515 + }, + { + "epoch": 0.15133834586466166, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1808, + "step": 7516 + }, + { + "epoch": 0.1513984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1735, + "step": 7517 + }, + { + "epoch": 0.15145864661654135, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 7518 + }, + { + "epoch": 0.1515187969924812, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2115, + "step": 7519 + }, + { + "epoch": 0.15157894736842106, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 7520 + }, + { + "epoch": 0.1516390977443609, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 7521 + }, + { + "epoch": 0.15169924812030075, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2191, + "step": 7522 + }, + { + "epoch": 0.1517593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2379, + "step": 7523 + }, + { + "epoch": 0.15181954887218044, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2659, + "step": 7524 + }, + { + "epoch": 0.1518796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2155, + "step": 7525 + }, + { + "epoch": 0.15193984962406015, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2449, + "step": 7526 + }, + { + "epoch": 0.152, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1418, + "step": 7527 + }, + { + "epoch": 0.15206015037593984, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 7528 + }, + { + "epoch": 0.1521203007518797, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2761, + "step": 7529 + }, + { + "epoch": 0.15218045112781955, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 7530 + }, + { + "epoch": 0.1522406015037594, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.305, + "step": 7531 + }, + { + "epoch": 0.15230075187969924, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 7532 + }, + { + "epoch": 0.15236090225563909, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2457, + "step": 7533 + }, + { + "epoch": 0.15242105263157896, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2717, + "step": 7534 + }, + { + "epoch": 0.1524812030075188, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2373, + "step": 7535 + }, + { + "epoch": 0.15254135338345864, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2976, + "step": 7536 + }, + { + "epoch": 0.1526015037593985, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2805, + "step": 7537 + }, + { + "epoch": 0.15266165413533836, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 7538 + }, + { + "epoch": 0.1527218045112782, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2998, + "step": 7539 + }, + { + "epoch": 0.15278195488721805, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 7540 + }, + { + "epoch": 0.1528421052631579, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1946, + "step": 7541 + }, + { + "epoch": 0.15290225563909773, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.2038, + "step": 7542 + }, + { + "epoch": 0.1529624060150376, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2635, + "step": 7543 + }, + { + "epoch": 0.15302255639097745, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2995, + "step": 7544 + }, + { + "epoch": 0.1530827067669173, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2349, + "step": 7545 + }, + { + "epoch": 0.15314285714285714, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.241, + "step": 7546 + }, + { + "epoch": 0.15320300751879698, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1839, + "step": 7547 + }, + { + "epoch": 0.15326315789473685, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2822, + "step": 7548 + }, + { + "epoch": 0.1533233082706767, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2563, + "step": 7549 + }, + { + "epoch": 0.15338345864661654, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1919, + "step": 7550 + }, + { + "epoch": 0.15344360902255638, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2979, + "step": 7551 + }, + { + "epoch": 0.15350375939849625, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1908, + "step": 7552 + }, + { + "epoch": 0.1535639097744361, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.189, + "step": 7553 + }, + { + "epoch": 0.15362406015037594, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2482, + "step": 7554 + }, + { + "epoch": 0.15368421052631578, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2739, + "step": 7555 + }, + { + "epoch": 0.15374436090225563, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3609, + "step": 7556 + }, + { + "epoch": 0.1538045112781955, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 7557 + }, + { + "epoch": 0.15386466165413534, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3795, + "step": 7558 + }, + { + "epoch": 0.1539248120300752, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 7559 + }, + { + "epoch": 0.15398496240601503, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2115, + "step": 7560 + }, + { + "epoch": 0.15404511278195487, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2796, + "step": 7561 + }, + { + "epoch": 0.15410526315789475, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.245, + "step": 7562 + }, + { + "epoch": 0.1541654135338346, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 7563 + }, + { + "epoch": 0.15422556390977443, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 7564 + }, + { + "epoch": 0.15428571428571428, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2037, + "step": 7565 + }, + { + "epoch": 0.15434586466165415, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2088, + "step": 7566 + }, + { + "epoch": 0.154406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2862, + "step": 7567 + }, + { + "epoch": 0.15446616541353383, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2507, + "step": 7568 + }, + { + "epoch": 0.15452631578947368, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2556, + "step": 7569 + }, + { + "epoch": 0.15458646616541352, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.277, + "step": 7570 + }, + { + "epoch": 0.1546466165413534, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2336, + "step": 7571 + }, + { + "epoch": 0.15470676691729324, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.4703, + "step": 7572 + }, + { + "epoch": 0.15476691729323308, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2763, + "step": 7573 + }, + { + "epoch": 0.15482706766917292, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2321, + "step": 7574 + }, + { + "epoch": 0.1548872180451128, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2716, + "step": 7575 + }, + { + "epoch": 0.15494736842105264, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2696, + "step": 7576 + }, + { + "epoch": 0.15500751879699248, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.293, + "step": 7577 + }, + { + "epoch": 0.15506766917293233, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2179, + "step": 7578 + }, + { + "epoch": 0.15512781954887217, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3145, + "step": 7579 + }, + { + "epoch": 0.15518796992481204, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2, + "step": 7580 + }, + { + "epoch": 0.15524812030075189, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2096, + "step": 7581 + }, + { + "epoch": 0.15530827067669173, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2869, + "step": 7582 + }, + { + "epoch": 0.15536842105263157, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 7583 + }, + { + "epoch": 0.15542857142857142, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 7584 + }, + { + "epoch": 0.1554887218045113, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2242, + "step": 7585 + }, + { + "epoch": 0.15554887218045113, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 7586 + }, + { + "epoch": 0.15560902255639097, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.28, + "step": 7587 + }, + { + "epoch": 0.15566917293233082, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 7588 + }, + { + "epoch": 0.1557293233082707, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2315, + "step": 7589 + }, + { + "epoch": 0.15578947368421053, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1841, + "step": 7590 + }, + { + "epoch": 0.15584962406015038, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1886, + "step": 7591 + }, + { + "epoch": 0.15590977443609022, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2568, + "step": 7592 + }, + { + "epoch": 0.15596992481203006, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 7593 + }, + { + "epoch": 0.15603007518796994, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 7594 + }, + { + "epoch": 0.15609022556390978, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.26, + "step": 7595 + }, + { + "epoch": 0.15615037593984962, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.323, + "step": 7596 + }, + { + "epoch": 0.15621052631578947, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2609, + "step": 7597 + }, + { + "epoch": 0.1562706766917293, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2027, + "step": 7598 + }, + { + "epoch": 0.15633082706766918, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3397, + "step": 7599 + }, + { + "epoch": 0.15639097744360902, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 7600 + }, + { + "epoch": 0.15645112781954887, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.3141, + "step": 7601 + }, + { + "epoch": 0.1565112781954887, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 7602 + }, + { + "epoch": 0.15657142857142858, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2749, + "step": 7603 + }, + { + "epoch": 0.15663157894736843, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2132, + "step": 7604 + }, + { + "epoch": 0.15669172932330827, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2439, + "step": 7605 + }, + { + "epoch": 0.15675187969924811, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1957, + "step": 7606 + }, + { + "epoch": 0.15681203007518796, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.262, + "step": 7607 + }, + { + "epoch": 0.15687218045112783, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.3, + "step": 7608 + }, + { + "epoch": 0.15693233082706767, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3047, + "step": 7609 + }, + { + "epoch": 0.15699248120300752, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2577, + "step": 7610 + }, + { + "epoch": 0.15705263157894736, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2413, + "step": 7611 + }, + { + "epoch": 0.15711278195488723, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 7612 + }, + { + "epoch": 0.15717293233082708, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2715, + "step": 7613 + }, + { + "epoch": 0.15723308270676692, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.296, + "step": 7614 + }, + { + "epoch": 0.15729323308270676, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.192, + "step": 7615 + }, + { + "epoch": 0.1573533834586466, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2799, + "step": 7616 + }, + { + "epoch": 0.15741353383458648, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.3088, + "step": 7617 + }, + { + "epoch": 0.15747368421052632, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1852, + "step": 7618 + }, + { + "epoch": 0.15753383458646616, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 7619 + }, + { + "epoch": 0.157593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2556, + "step": 7620 + }, + { + "epoch": 0.15765413533834585, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 7621 + }, + { + "epoch": 0.15771428571428572, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3232, + "step": 7622 + }, + { + "epoch": 0.15777443609022557, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2756, + "step": 7623 + }, + { + "epoch": 0.1578345864661654, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2703, + "step": 7624 + }, + { + "epoch": 0.15789473684210525, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.226, + "step": 7625 + }, + { + "epoch": 0.15795488721804513, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1959, + "step": 7626 + }, + { + "epoch": 0.15801503759398497, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2755, + "step": 7627 + }, + { + "epoch": 0.1580751879699248, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2191, + "step": 7628 + }, + { + "epoch": 0.15813533834586466, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.253, + "step": 7629 + }, + { + "epoch": 0.1581954887218045, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.435, + "step": 7630 + }, + { + "epoch": 0.15825563909774437, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 7631 + }, + { + "epoch": 0.15831578947368422, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2772, + "step": 7632 + }, + { + "epoch": 0.15837593984962406, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2891, + "step": 7633 + }, + { + "epoch": 0.1584360902255639, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 7634 + }, + { + "epoch": 0.15849624060150375, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2153, + "step": 7635 + }, + { + "epoch": 0.15855639097744362, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 7636 + }, + { + "epoch": 0.15861654135338346, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 7637 + }, + { + "epoch": 0.1586766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 7638 + }, + { + "epoch": 0.15873684210526315, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2616, + "step": 7639 + }, + { + "epoch": 0.15879699248120302, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2374, + "step": 7640 + }, + { + "epoch": 0.15885714285714286, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2884, + "step": 7641 + }, + { + "epoch": 0.1589172932330827, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.236, + "step": 7642 + }, + { + "epoch": 0.15897744360902255, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2767, + "step": 7643 + }, + { + "epoch": 0.1590375939849624, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1915, + "step": 7644 + }, + { + "epoch": 0.15909774436090227, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 7645 + }, + { + "epoch": 0.1591578947368421, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2387, + "step": 7646 + }, + { + "epoch": 0.15921804511278195, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2355, + "step": 7647 + }, + { + "epoch": 0.1592781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 7648 + }, + { + "epoch": 0.15933834586466167, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.3066, + "step": 7649 + }, + { + "epoch": 0.1593984962406015, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2445, + "step": 7650 + }, + { + "epoch": 0.15945864661654136, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2401, + "step": 7651 + }, + { + "epoch": 0.1595187969924812, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2204, + "step": 7652 + }, + { + "epoch": 0.15957894736842104, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1847, + "step": 7653 + }, + { + "epoch": 0.1596390977443609, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.1891, + "step": 7654 + }, + { + "epoch": 0.15969924812030076, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 7655 + }, + { + "epoch": 0.1597593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 7656 + }, + { + "epoch": 0.15981954887218044, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2624, + "step": 7657 + }, + { + "epoch": 0.1598796992481203, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1954, + "step": 7658 + }, + { + "epoch": 0.15993984962406016, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2077, + "step": 7659 + }, + { + "epoch": 0.16, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2323, + "step": 7660 + }, + { + "epoch": 0.16006015037593985, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2531, + "step": 7661 + }, + { + "epoch": 0.1601203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 7662 + }, + { + "epoch": 0.16018045112781956, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2434, + "step": 7663 + }, + { + "epoch": 0.1602406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2488, + "step": 7664 + }, + { + "epoch": 0.16030075187969925, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2583, + "step": 7665 + }, + { + "epoch": 0.1603609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 7666 + }, + { + "epoch": 0.16042105263157894, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 7667 + }, + { + "epoch": 0.1604812030075188, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2025, + "step": 7668 + }, + { + "epoch": 0.16054135338345865, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 7669 + }, + { + "epoch": 0.1606015037593985, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1418, + "step": 7670 + }, + { + "epoch": 0.16066165413533834, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.281, + "step": 7671 + }, + { + "epoch": 0.16072180451127818, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2319, + "step": 7672 + }, + { + "epoch": 0.16078195488721805, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 7673 + }, + { + "epoch": 0.1608421052631579, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 7674 + }, + { + "epoch": 0.16090225563909774, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 7675 + }, + { + "epoch": 0.16096240601503758, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.213, + "step": 7676 + }, + { + "epoch": 0.16102255639097746, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 7677 + }, + { + "epoch": 0.1610827067669173, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1919, + "step": 7678 + }, + { + "epoch": 0.16114285714285714, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 7679 + }, + { + "epoch": 0.161203007518797, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.2147, + "step": 7680 + }, + { + "epoch": 0.16126315789473683, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2477, + "step": 7681 + }, + { + "epoch": 0.1613233082706767, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2624, + "step": 7682 + }, + { + "epoch": 0.16138345864661655, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 7683 + }, + { + "epoch": 0.1614436090225564, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 7684 + }, + { + "epoch": 0.16150375939849623, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.163, + "step": 7685 + }, + { + "epoch": 0.16156390977443608, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2548, + "step": 7686 + }, + { + "epoch": 0.16162406015037595, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2118, + "step": 7687 + }, + { + "epoch": 0.1616842105263158, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1785, + "step": 7688 + }, + { + "epoch": 0.16174436090225564, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 7689 + }, + { + "epoch": 0.16180451127819548, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 7690 + }, + { + "epoch": 0.16186466165413535, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2666, + "step": 7691 + }, + { + "epoch": 0.1619248120300752, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2542, + "step": 7692 + }, + { + "epoch": 0.16198496240601504, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2161, + "step": 7693 + }, + { + "epoch": 0.16204511278195488, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2818, + "step": 7694 + }, + { + "epoch": 0.16210526315789472, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 7695 + }, + { + "epoch": 0.1621654135338346, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 7696 + }, + { + "epoch": 0.16222556390977444, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 7697 + }, + { + "epoch": 0.16228571428571428, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2678, + "step": 7698 + }, + { + "epoch": 0.16234586466165413, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.3247, + "step": 7699 + }, + { + "epoch": 0.162406015037594, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2767, + "step": 7700 + }, + { + "epoch": 0.16246616541353384, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.2052, + "step": 7701 + }, + { + "epoch": 0.16252631578947369, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.23, + "step": 7702 + }, + { + "epoch": 0.16258646616541353, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 7703 + }, + { + "epoch": 0.16264661654135337, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2393, + "step": 7704 + }, + { + "epoch": 0.16270676691729324, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2966, + "step": 7705 + }, + { + "epoch": 0.1627669172932331, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2844, + "step": 7706 + }, + { + "epoch": 0.16282706766917293, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2061, + "step": 7707 + }, + { + "epoch": 0.16288721804511277, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2649, + "step": 7708 + }, + { + "epoch": 0.16294736842105262, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.245, + "step": 7709 + }, + { + "epoch": 0.1630075187969925, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 7710 + }, + { + "epoch": 0.16306766917293233, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2356, + "step": 7711 + }, + { + "epoch": 0.16312781954887218, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2166, + "step": 7712 + }, + { + "epoch": 0.16318796992481202, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2739, + "step": 7713 + }, + { + "epoch": 0.1632481203007519, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1545, + "step": 7714 + }, + { + "epoch": 0.16330827067669174, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2955, + "step": 7715 + }, + { + "epoch": 0.16336842105263158, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2557, + "step": 7716 + }, + { + "epoch": 0.16342857142857142, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1872, + "step": 7717 + }, + { + "epoch": 0.16348872180451127, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 7718 + }, + { + "epoch": 0.16354887218045114, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2498, + "step": 7719 + }, + { + "epoch": 0.16360902255639098, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2393, + "step": 7720 + }, + { + "epoch": 0.16366917293233083, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 7721 + }, + { + "epoch": 0.16372932330827067, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 7722 + }, + { + "epoch": 0.1637894736842105, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1704, + "step": 7723 + }, + { + "epoch": 0.16384962406015038, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 7724 + }, + { + "epoch": 0.16390977443609023, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.4191, + "step": 7725 + }, + { + "epoch": 0.16396992481203007, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2931, + "step": 7726 + }, + { + "epoch": 0.16403007518796991, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 7727 + }, + { + "epoch": 0.1640902255639098, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3035, + "step": 7728 + }, + { + "epoch": 0.16415037593984963, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 7729 + }, + { + "epoch": 0.16421052631578947, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2548, + "step": 7730 + }, + { + "epoch": 0.16427067669172932, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2333, + "step": 7731 + }, + { + "epoch": 0.16433082706766916, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3039, + "step": 7732 + }, + { + "epoch": 0.16439097744360903, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2388, + "step": 7733 + }, + { + "epoch": 0.16445112781954888, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2853, + "step": 7734 + }, + { + "epoch": 0.16451127819548872, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.3846, + "step": 7735 + }, + { + "epoch": 0.16457142857142856, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2037, + "step": 7736 + }, + { + "epoch": 0.16463157894736843, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.239, + "step": 7737 + }, + { + "epoch": 0.16469172932330828, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 7738 + }, + { + "epoch": 0.16475187969924812, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2271, + "step": 7739 + }, + { + "epoch": 0.16481203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 7740 + }, + { + "epoch": 0.1648721804511278, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 7741 + }, + { + "epoch": 0.16493233082706768, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2609, + "step": 7742 + }, + { + "epoch": 0.16499248120300752, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 7743 + }, + { + "epoch": 0.16505263157894737, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 7744 + }, + { + "epoch": 0.1651127819548872, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3097, + "step": 7745 + }, + { + "epoch": 0.16517293233082705, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 7746 + }, + { + "epoch": 0.16523308270676693, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2601, + "step": 7747 + }, + { + "epoch": 0.16529323308270677, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 7748 + }, + { + "epoch": 0.1653533834586466, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 7749 + }, + { + "epoch": 0.16541353383458646, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2456, + "step": 7750 + }, + { + "epoch": 0.16547368421052633, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2069, + "step": 7751 + }, + { + "epoch": 0.16553383458646617, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 7752 + }, + { + "epoch": 0.16559398496240602, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2042, + "step": 7753 + }, + { + "epoch": 0.16565413533834586, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 7754 + }, + { + "epoch": 0.1657142857142857, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3064, + "step": 7755 + }, + { + "epoch": 0.16577443609022557, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2182, + "step": 7756 + }, + { + "epoch": 0.16583458646616542, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 7757 + }, + { + "epoch": 0.16589473684210526, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.3412, + "step": 7758 + }, + { + "epoch": 0.1659548872180451, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1793, + "step": 7759 + }, + { + "epoch": 0.16601503759398495, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2044, + "step": 7760 + }, + { + "epoch": 0.16607518796992482, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2587, + "step": 7761 + }, + { + "epoch": 0.16613533834586466, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2391, + "step": 7762 + }, + { + "epoch": 0.1661954887218045, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2165, + "step": 7763 + }, + { + "epoch": 0.16625563909774435, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2, + "step": 7764 + }, + { + "epoch": 0.16631578947368422, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 7765 + }, + { + "epoch": 0.16637593984962407, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2135, + "step": 7766 + }, + { + "epoch": 0.1664360902255639, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.2022, + "step": 7767 + }, + { + "epoch": 0.16649624060150375, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2861, + "step": 7768 + }, + { + "epoch": 0.1665563909774436, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 7769 + }, + { + "epoch": 0.16661654135338347, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.331, + "step": 7770 + }, + { + "epoch": 0.1666766917293233, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2969, + "step": 7771 + }, + { + "epoch": 0.16673684210526316, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1949, + "step": 7772 + }, + { + "epoch": 0.166796992481203, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2046, + "step": 7773 + }, + { + "epoch": 0.16685714285714287, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 7774 + }, + { + "epoch": 0.16691729323308271, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2391, + "step": 7775 + }, + { + "epoch": 0.16697744360902256, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.238, + "step": 7776 + }, + { + "epoch": 0.1670375939849624, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 7777 + }, + { + "epoch": 0.16709774436090225, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 7778 + }, + { + "epoch": 0.16715789473684212, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 7779 + }, + { + "epoch": 0.16721804511278196, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 7780 + }, + { + "epoch": 0.1672781954887218, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2177, + "step": 7781 + }, + { + "epoch": 0.16733834586466165, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2609, + "step": 7782 + }, + { + "epoch": 0.1673984962406015, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2484, + "step": 7783 + }, + { + "epoch": 0.16745864661654136, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 7784 + }, + { + "epoch": 0.1675187969924812, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 7785 + }, + { + "epoch": 0.16757894736842105, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1682, + "step": 7786 + }, + { + "epoch": 0.1676390977443609, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.2629, + "step": 7787 + }, + { + "epoch": 0.16769924812030076, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2481, + "step": 7788 + }, + { + "epoch": 0.1677593984962406, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 7789 + }, + { + "epoch": 0.16781954887218045, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2433, + "step": 7790 + }, + { + "epoch": 0.1678796992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 7791 + }, + { + "epoch": 0.16793984962406014, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.136, + "step": 7792 + }, + { + "epoch": 0.168, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 7793 + }, + { + "epoch": 0.16806015037593985, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 7794 + }, + { + "epoch": 0.1681203007518797, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 7795 + }, + { + "epoch": 0.16818045112781954, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2577, + "step": 7796 + }, + { + "epoch": 0.16824060150375938, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 7797 + }, + { + "epoch": 0.16830075187969926, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1732, + "step": 7798 + }, + { + "epoch": 0.1683609022556391, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 7799 + }, + { + "epoch": 0.16842105263157894, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1635, + "step": 7800 + }, + { + "epoch": 0.1684812030075188, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2177, + "step": 7801 + }, + { + "epoch": 0.16854135338345866, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2532, + "step": 7802 + }, + { + "epoch": 0.1686015037593985, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2303, + "step": 7803 + }, + { + "epoch": 0.16866165413533835, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2096, + "step": 7804 + }, + { + "epoch": 0.1687218045112782, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2339, + "step": 7805 + }, + { + "epoch": 0.16878195488721803, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.27, + "step": 7806 + }, + { + "epoch": 0.1688421052631579, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2134, + "step": 7807 + }, + { + "epoch": 0.16890225563909775, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2144, + "step": 7808 + }, + { + "epoch": 0.1689624060150376, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3523, + "step": 7809 + }, + { + "epoch": 0.16902255639097744, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2608, + "step": 7810 + }, + { + "epoch": 0.1690827067669173, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1985, + "step": 7811 + }, + { + "epoch": 0.16914285714285715, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3189, + "step": 7812 + }, + { + "epoch": 0.169203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 7813 + }, + { + "epoch": 0.16926315789473684, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2289, + "step": 7814 + }, + { + "epoch": 0.16932330827067668, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 7815 + }, + { + "epoch": 0.16938345864661655, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.3286, + "step": 7816 + }, + { + "epoch": 0.1694436090225564, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1971, + "step": 7817 + }, + { + "epoch": 0.16950375939849624, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 7818 + }, + { + "epoch": 0.16956390977443608, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2637, + "step": 7819 + }, + { + "epoch": 0.16962406015037593, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3913, + "step": 7820 + }, + { + "epoch": 0.1696842105263158, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2884, + "step": 7821 + }, + { + "epoch": 0.16974436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2036, + "step": 7822 + }, + { + "epoch": 0.16980451127819549, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2074, + "step": 7823 + }, + { + "epoch": 0.16986466165413533, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2401, + "step": 7824 + }, + { + "epoch": 0.1699248120300752, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 7825 + }, + { + "epoch": 0.16998496240601504, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2321, + "step": 7826 + }, + { + "epoch": 0.1700451127819549, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 7827 + }, + { + "epoch": 0.17010526315789473, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 7828 + }, + { + "epoch": 0.17016541353383458, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 7829 + }, + { + "epoch": 0.17022556390977445, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3289, + "step": 7830 + }, + { + "epoch": 0.1702857142857143, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 7831 + }, + { + "epoch": 0.17034586466165413, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2865, + "step": 7832 + }, + { + "epoch": 0.17040601503759398, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.312, + "step": 7833 + }, + { + "epoch": 0.17046616541353382, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 7834 + }, + { + "epoch": 0.1705263157894737, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1916, + "step": 7835 + }, + { + "epoch": 0.17058646616541354, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1959, + "step": 7836 + }, + { + "epoch": 0.17064661654135338, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.307, + "step": 7837 + }, + { + "epoch": 0.17070676691729322, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 7838 + }, + { + "epoch": 0.1707669172932331, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2336, + "step": 7839 + }, + { + "epoch": 0.17082706766917294, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 7840 + }, + { + "epoch": 0.17088721804511278, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 7841 + }, + { + "epoch": 0.17094736842105263, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1936, + "step": 7842 + }, + { + "epoch": 0.17100751879699247, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 7843 + }, + { + "epoch": 0.17106766917293234, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2082, + "step": 7844 + }, + { + "epoch": 0.17112781954887218, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2011, + "step": 7845 + }, + { + "epoch": 0.17118796992481203, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 7846 + }, + { + "epoch": 0.17124812030075187, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2473, + "step": 7847 + }, + { + "epoch": 0.17130827067669174, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1888, + "step": 7848 + }, + { + "epoch": 0.1713684210526316, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2065, + "step": 7849 + }, + { + "epoch": 0.17142857142857143, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 7850 + }, + { + "epoch": 0.17148872180451127, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2453, + "step": 7851 + }, + { + "epoch": 0.17154887218045112, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2338, + "step": 7852 + }, + { + "epoch": 0.171609022556391, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1482, + "step": 7853 + }, + { + "epoch": 0.17166917293233083, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2184, + "step": 7854 + }, + { + "epoch": 0.17172932330827068, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 7855 + }, + { + "epoch": 0.17178947368421052, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2527, + "step": 7856 + }, + { + "epoch": 0.17184962406015036, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.238, + "step": 7857 + }, + { + "epoch": 0.17190977443609023, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 7858 + }, + { + "epoch": 0.17196992481203008, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2462, + "step": 7859 + }, + { + "epoch": 0.17203007518796992, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3745, + "step": 7860 + }, + { + "epoch": 0.17209022556390977, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2249, + "step": 7861 + }, + { + "epoch": 0.17215037593984964, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 7862 + }, + { + "epoch": 0.17221052631578948, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2286, + "step": 7863 + }, + { + "epoch": 0.17227067669172932, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.281, + "step": 7864 + }, + { + "epoch": 0.17233082706766917, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 7865 + }, + { + "epoch": 0.172390977443609, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2284, + "step": 7866 + }, + { + "epoch": 0.17245112781954888, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 7867 + }, + { + "epoch": 0.17251127819548873, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2797, + "step": 7868 + }, + { + "epoch": 0.17257142857142857, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2696, + "step": 7869 + }, + { + "epoch": 0.1726315789473684, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2453, + "step": 7870 + }, + { + "epoch": 0.17269172932330826, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 7871 + }, + { + "epoch": 0.17275187969924813, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 7872 + }, + { + "epoch": 0.17281203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2759, + "step": 7873 + }, + { + "epoch": 0.17287218045112782, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.3944, + "step": 7874 + }, + { + "epoch": 0.17293233082706766, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2621, + "step": 7875 + }, + { + "epoch": 0.17299248120300753, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3134, + "step": 7876 + }, + { + "epoch": 0.17305263157894737, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1746, + "step": 7877 + }, + { + "epoch": 0.17311278195488722, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.3345, + "step": 7878 + }, + { + "epoch": 0.17317293233082706, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2516, + "step": 7879 + }, + { + "epoch": 0.1732330827067669, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 7880 + }, + { + "epoch": 0.17329323308270678, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2628, + "step": 7881 + }, + { + "epoch": 0.17335338345864662, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2896, + "step": 7882 + }, + { + "epoch": 0.17341353383458646, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 7883 + }, + { + "epoch": 0.1734736842105263, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 7884 + }, + { + "epoch": 0.17353383458646618, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3229, + "step": 7885 + }, + { + "epoch": 0.17359398496240602, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.2624, + "step": 7886 + }, + { + "epoch": 0.17365413533834587, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2578, + "step": 7887 + }, + { + "epoch": 0.1737142857142857, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 7888 + }, + { + "epoch": 0.17377443609022555, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2603, + "step": 7889 + }, + { + "epoch": 0.17383458646616542, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 7890 + }, + { + "epoch": 0.17389473684210527, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 7891 + }, + { + "epoch": 0.1739548872180451, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 7892 + }, + { + "epoch": 0.17401503759398496, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2157, + "step": 7893 + }, + { + "epoch": 0.1740751879699248, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1954, + "step": 7894 + }, + { + "epoch": 0.17413533834586467, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2789, + "step": 7895 + }, + { + "epoch": 0.17419548872180451, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.202, + "step": 7896 + }, + { + "epoch": 0.17425563909774436, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3009, + "step": 7897 + }, + { + "epoch": 0.1743157894736842, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2441, + "step": 7898 + }, + { + "epoch": 0.17437593984962407, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2386, + "step": 7899 + }, + { + "epoch": 0.17443609022556392, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2599, + "step": 7900 + }, + { + "epoch": 0.17449624060150376, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 7901 + }, + { + "epoch": 0.1745563909774436, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2283, + "step": 7902 + }, + { + "epoch": 0.17461654135338345, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 7903 + }, + { + "epoch": 0.17467669172932332, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2322, + "step": 7904 + }, + { + "epoch": 0.17473684210526316, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 7905 + }, + { + "epoch": 0.174796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.203, + "step": 7906 + }, + { + "epoch": 0.17485714285714285, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1897, + "step": 7907 + }, + { + "epoch": 0.1749172932330827, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 7908 + }, + { + "epoch": 0.17497744360902256, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2529, + "step": 7909 + }, + { + "epoch": 0.1750375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1633, + "step": 7910 + }, + { + "epoch": 0.17509774436090225, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1886, + "step": 7911 + }, + { + "epoch": 0.1751578947368421, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.186, + "step": 7912 + }, + { + "epoch": 0.17521804511278197, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2824, + "step": 7913 + }, + { + "epoch": 0.1752781954887218, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 7914 + }, + { + "epoch": 0.17533834586466165, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2445, + "step": 7915 + }, + { + "epoch": 0.1753984962406015, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3121, + "step": 7916 + }, + { + "epoch": 0.17545864661654134, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1751, + "step": 7917 + }, + { + "epoch": 0.1755187969924812, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3194, + "step": 7918 + }, + { + "epoch": 0.17557894736842106, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 7919 + }, + { + "epoch": 0.1756390977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2505, + "step": 7920 + }, + { + "epoch": 0.17569924812030074, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2209, + "step": 7921 + }, + { + "epoch": 0.17575939849624062, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.233, + "step": 7922 + }, + { + "epoch": 0.17581954887218046, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2215, + "step": 7923 + }, + { + "epoch": 0.1758796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2261, + "step": 7924 + }, + { + "epoch": 0.17593984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2307, + "step": 7925 + }, + { + "epoch": 0.176, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 7926 + }, + { + "epoch": 0.17606015037593986, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2294, + "step": 7927 + }, + { + "epoch": 0.1761203007518797, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2392, + "step": 7928 + }, + { + "epoch": 0.17618045112781955, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 7929 + }, + { + "epoch": 0.1762406015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2793, + "step": 7930 + }, + { + "epoch": 0.17630075187969924, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 7931 + }, + { + "epoch": 0.1763609022556391, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2278, + "step": 7932 + }, + { + "epoch": 0.17642105263157895, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1879, + "step": 7933 + }, + { + "epoch": 0.1764812030075188, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 7934 + }, + { + "epoch": 0.17654135338345864, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2864, + "step": 7935 + }, + { + "epoch": 0.1766015037593985, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 7936 + }, + { + "epoch": 0.17666165413533835, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 7937 + }, + { + "epoch": 0.1767218045112782, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 7938 + }, + { + "epoch": 0.17678195488721804, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 7939 + }, + { + "epoch": 0.17684210526315788, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2402, + "step": 7940 + }, + { + "epoch": 0.17690225563909776, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1971, + "step": 7941 + }, + { + "epoch": 0.1769624060150376, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2185, + "step": 7942 + }, + { + "epoch": 0.17702255639097744, + "grad_norm": 0.44921875, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 7943 + }, + { + "epoch": 0.17708270676691729, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2121, + "step": 7944 + }, + { + "epoch": 0.17714285714285713, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2689, + "step": 7945 + }, + { + "epoch": 0.177203007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2413, + "step": 7946 + }, + { + "epoch": 0.17726315789473684, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 7947 + }, + { + "epoch": 0.1773233082706767, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2087, + "step": 7948 + }, + { + "epoch": 0.17738345864661653, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1847, + "step": 7949 + }, + { + "epoch": 0.1774436090225564, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 7950 + }, + { + "epoch": 0.17750375939849625, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2376, + "step": 7951 + }, + { + "epoch": 0.1775639097744361, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 7952 + }, + { + "epoch": 0.17762406015037593, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1766, + "step": 7953 + }, + { + "epoch": 0.17768421052631578, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1983, + "step": 7954 + }, + { + "epoch": 0.17774436090225565, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2255, + "step": 7955 + }, + { + "epoch": 0.1778045112781955, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 7956 + }, + { + "epoch": 0.17786466165413534, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 7957 + }, + { + "epoch": 0.17792481203007518, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 7958 + }, + { + "epoch": 0.17798496240601502, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1962, + "step": 7959 + }, + { + "epoch": 0.1780451127819549, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.282, + "step": 7960 + }, + { + "epoch": 0.17810526315789474, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 7961 + }, + { + "epoch": 0.17816541353383458, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 7962 + }, + { + "epoch": 0.17822556390977443, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3009, + "step": 7963 + }, + { + "epoch": 0.1782857142857143, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2287, + "step": 7964 + }, + { + "epoch": 0.17834586466165414, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.3134, + "step": 7965 + }, + { + "epoch": 0.17840601503759398, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2353, + "step": 7966 + }, + { + "epoch": 0.17846616541353383, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1766, + "step": 7967 + }, + { + "epoch": 0.17852631578947367, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2388, + "step": 7968 + }, + { + "epoch": 0.17858646616541354, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1955, + "step": 7969 + }, + { + "epoch": 0.1786466165413534, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 7970 + }, + { + "epoch": 0.17870676691729323, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.2277, + "step": 7971 + }, + { + "epoch": 0.17876691729323307, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1937, + "step": 7972 + }, + { + "epoch": 0.17882706766917295, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.2185, + "step": 7973 + }, + { + "epoch": 0.1788872180451128, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2474, + "step": 7974 + }, + { + "epoch": 0.17894736842105263, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.2865, + "step": 7975 + }, + { + "epoch": 0.17900751879699248, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2471, + "step": 7976 + }, + { + "epoch": 0.17906766917293232, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2297, + "step": 7977 + }, + { + "epoch": 0.1791278195488722, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.197, + "step": 7978 + }, + { + "epoch": 0.17918796992481204, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2932, + "step": 7979 + }, + { + "epoch": 0.17924812030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.223, + "step": 7980 + }, + { + "epoch": 0.17930827067669172, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2149, + "step": 7981 + }, + { + "epoch": 0.17936842105263157, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 7982 + }, + { + "epoch": 0.17942857142857144, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2321, + "step": 7983 + }, + { + "epoch": 0.17948872180451128, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2591, + "step": 7984 + }, + { + "epoch": 0.17954887218045112, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2241, + "step": 7985 + }, + { + "epoch": 0.17960902255639097, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3206, + "step": 7986 + }, + { + "epoch": 0.17966917293233084, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.256, + "step": 7987 + }, + { + "epoch": 0.17972932330827068, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 7988 + }, + { + "epoch": 0.17978947368421053, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1995, + "step": 7989 + }, + { + "epoch": 0.17984962406015037, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3124, + "step": 7990 + }, + { + "epoch": 0.17990977443609021, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2206, + "step": 7991 + }, + { + "epoch": 0.17996992481203009, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1882, + "step": 7992 + }, + { + "epoch": 0.18003007518796993, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2677, + "step": 7993 + }, + { + "epoch": 0.18009022556390977, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 7994 + }, + { + "epoch": 0.18015037593984962, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 7995 + }, + { + "epoch": 0.18021052631578946, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2825, + "step": 7996 + }, + { + "epoch": 0.18027067669172933, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3704, + "step": 7997 + }, + { + "epoch": 0.18033082706766917, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 7998 + }, + { + "epoch": 0.18039097744360902, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 7999 + }, + { + "epoch": 0.18045112781954886, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 8000 + }, + { + "epoch": 0.18051127819548873, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2088, + "step": 8001 + }, + { + "epoch": 0.18057142857142858, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2559, + "step": 8002 + }, + { + "epoch": 0.18063157894736842, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3176, + "step": 8003 + }, + { + "epoch": 0.18069172932330826, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 8004 + }, + { + "epoch": 0.1807518796992481, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 8005 + }, + { + "epoch": 0.18081203007518798, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2347, + "step": 8006 + }, + { + "epoch": 0.18087218045112782, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.3798, + "step": 8007 + }, + { + "epoch": 0.18093233082706767, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.216, + "step": 8008 + }, + { + "epoch": 0.1809924812030075, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2019, + "step": 8009 + }, + { + "epoch": 0.18105263157894738, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 8010 + }, + { + "epoch": 0.18111278195488723, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2303, + "step": 8011 + }, + { + "epoch": 0.18117293233082707, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 8012 + }, + { + "epoch": 0.1812330827067669, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2752, + "step": 8013 + }, + { + "epoch": 0.18129323308270676, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.2823, + "step": 8014 + }, + { + "epoch": 0.18135338345864663, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2401, + "step": 8015 + }, + { + "epoch": 0.18141353383458647, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 8016 + }, + { + "epoch": 0.18147368421052631, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 8017 + }, + { + "epoch": 0.18153383458646616, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.309, + "step": 8018 + }, + { + "epoch": 0.181593984962406, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3851, + "step": 8019 + }, + { + "epoch": 0.18165413533834587, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3772, + "step": 8020 + }, + { + "epoch": 0.18171428571428572, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 8021 + }, + { + "epoch": 0.18177443609022556, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2264, + "step": 8022 + }, + { + "epoch": 0.1818345864661654, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2925, + "step": 8023 + }, + { + "epoch": 0.18189473684210528, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 8024 + }, + { + "epoch": 0.18195488721804512, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2569, + "step": 8025 + }, + { + "epoch": 0.18201503759398496, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.4657, + "step": 8026 + }, + { + "epoch": 0.1820751879699248, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 8027 + }, + { + "epoch": 0.18213533834586465, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 8028 + }, + { + "epoch": 0.18219548872180452, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2608, + "step": 8029 + }, + { + "epoch": 0.18225563909774437, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 8030 + }, + { + "epoch": 0.1823157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1735, + "step": 8031 + }, + { + "epoch": 0.18237593984962405, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 8032 + }, + { + "epoch": 0.1824360902255639, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.342, + "step": 8033 + }, + { + "epoch": 0.18249624060150377, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 8034 + }, + { + "epoch": 0.1825563909774436, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2349, + "step": 8035 + }, + { + "epoch": 0.18261654135338345, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 8036 + }, + { + "epoch": 0.1826766917293233, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 8037 + }, + { + "epoch": 0.18273684210526317, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 8038 + }, + { + "epoch": 0.182796992481203, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2025, + "step": 8039 + }, + { + "epoch": 0.18285714285714286, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2112, + "step": 8040 + }, + { + "epoch": 0.1829172932330827, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2675, + "step": 8041 + }, + { + "epoch": 0.18297744360902254, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 8042 + }, + { + "epoch": 0.18303759398496242, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 8043 + }, + { + "epoch": 0.18309774436090226, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 8044 + }, + { + "epoch": 0.1831578947368421, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2005, + "step": 8045 + }, + { + "epoch": 0.18321804511278195, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2919, + "step": 8046 + }, + { + "epoch": 0.18327819548872182, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 8047 + }, + { + "epoch": 0.18333834586466166, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2495, + "step": 8048 + }, + { + "epoch": 0.1833984962406015, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2504, + "step": 8049 + }, + { + "epoch": 0.18345864661654135, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 8050 + }, + { + "epoch": 0.1835187969924812, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2978, + "step": 8051 + }, + { + "epoch": 0.18357894736842106, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2028, + "step": 8052 + }, + { + "epoch": 0.1836390977443609, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 8053 + }, + { + "epoch": 0.18369924812030075, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 8054 + }, + { + "epoch": 0.1837593984962406, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1582, + "step": 8055 + }, + { + "epoch": 0.18381954887218044, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2115, + "step": 8056 + }, + { + "epoch": 0.1838796992481203, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 8057 + }, + { + "epoch": 0.18393984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2313, + "step": 8058 + }, + { + "epoch": 0.184, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 8059 + }, + { + "epoch": 0.18406015037593984, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 8060 + }, + { + "epoch": 0.1841203007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2014, + "step": 8061 + }, + { + "epoch": 0.18418045112781956, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2271, + "step": 8062 + }, + { + "epoch": 0.1842406015037594, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2037, + "step": 8063 + }, + { + "epoch": 0.18430075187969924, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 8064 + }, + { + "epoch": 0.1843609022556391, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2601, + "step": 8065 + }, + { + "epoch": 0.18442105263157896, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2463, + "step": 8066 + }, + { + "epoch": 0.1844812030075188, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2692, + "step": 8067 + }, + { + "epoch": 0.18454135338345865, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 8068 + }, + { + "epoch": 0.1846015037593985, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 8069 + }, + { + "epoch": 0.18466165413533833, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.349, + "step": 8070 + }, + { + "epoch": 0.1847218045112782, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2129, + "step": 8071 + }, + { + "epoch": 0.18478195488721805, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 8072 + }, + { + "epoch": 0.1848421052631579, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2264, + "step": 8073 + }, + { + "epoch": 0.18490225563909773, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2907, + "step": 8074 + }, + { + "epoch": 0.1849624060150376, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 8075 + }, + { + "epoch": 0.18502255639097745, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.204, + "step": 8076 + }, + { + "epoch": 0.1850827067669173, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 8077 + }, + { + "epoch": 0.18514285714285714, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 8078 + }, + { + "epoch": 0.18520300751879698, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2196, + "step": 8079 + }, + { + "epoch": 0.18526315789473685, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2584, + "step": 8080 + }, + { + "epoch": 0.1853233082706767, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3272, + "step": 8081 + }, + { + "epoch": 0.18538345864661654, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 8082 + }, + { + "epoch": 0.18544360902255638, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2215, + "step": 8083 + }, + { + "epoch": 0.18550375939849625, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.3278, + "step": 8084 + }, + { + "epoch": 0.1855639097744361, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1658, + "step": 8085 + }, + { + "epoch": 0.18562406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2509, + "step": 8086 + }, + { + "epoch": 0.18568421052631578, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2389, + "step": 8087 + }, + { + "epoch": 0.18574436090225563, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2444, + "step": 8088 + }, + { + "epoch": 0.1858045112781955, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 8089 + }, + { + "epoch": 0.18586466165413534, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2838, + "step": 8090 + }, + { + "epoch": 0.1859248120300752, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 8091 + }, + { + "epoch": 0.18598496240601503, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2339, + "step": 8092 + }, + { + "epoch": 0.18604511278195487, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2814, + "step": 8093 + }, + { + "epoch": 0.18610526315789475, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.2309, + "step": 8094 + }, + { + "epoch": 0.1861654135338346, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 8095 + }, + { + "epoch": 0.18622556390977443, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 8096 + }, + { + "epoch": 0.18628571428571428, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2682, + "step": 8097 + }, + { + "epoch": 0.18634586466165415, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.3121, + "step": 8098 + }, + { + "epoch": 0.186406015037594, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2255, + "step": 8099 + }, + { + "epoch": 0.18646616541353384, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 8100 + }, + { + "epoch": 0.18652631578947368, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 8101 + }, + { + "epoch": 0.18658646616541352, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2497, + "step": 8102 + }, + { + "epoch": 0.1866466165413534, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 8103 + }, + { + "epoch": 0.18670676691729324, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.2964, + "step": 8104 + }, + { + "epoch": 0.18676691729323308, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1918, + "step": 8105 + }, + { + "epoch": 0.18682706766917292, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2499, + "step": 8106 + }, + { + "epoch": 0.18688721804511277, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2765, + "step": 8107 + }, + { + "epoch": 0.18694736842105264, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3079, + "step": 8108 + }, + { + "epoch": 0.18700751879699248, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2083, + "step": 8109 + }, + { + "epoch": 0.18706766917293233, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2402, + "step": 8110 + }, + { + "epoch": 0.18712781954887217, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2834, + "step": 8111 + }, + { + "epoch": 0.18718796992481204, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 8112 + }, + { + "epoch": 0.18724812030075189, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2597, + "step": 8113 + }, + { + "epoch": 0.18730827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1585, + "step": 8114 + }, + { + "epoch": 0.18736842105263157, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2486, + "step": 8115 + }, + { + "epoch": 0.18742857142857142, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1811, + "step": 8116 + }, + { + "epoch": 0.1874887218045113, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2568, + "step": 8117 + }, + { + "epoch": 0.18754887218045113, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 8118 + }, + { + "epoch": 0.18760902255639098, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 8119 + }, + { + "epoch": 0.18766917293233082, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2258, + "step": 8120 + }, + { + "epoch": 0.1877293233082707, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 8121 + }, + { + "epoch": 0.18778947368421053, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2564, + "step": 8122 + }, + { + "epoch": 0.18784962406015038, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1623, + "step": 8123 + }, + { + "epoch": 0.18790977443609022, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2581, + "step": 8124 + }, + { + "epoch": 0.18796992481203006, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2213, + "step": 8125 + }, + { + "epoch": 0.18803007518796994, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3088, + "step": 8126 + }, + { + "epoch": 0.18809022556390978, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 8127 + }, + { + "epoch": 0.18815037593984962, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2928, + "step": 8128 + }, + { + "epoch": 0.18821052631578947, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 8129 + }, + { + "epoch": 0.1882706766917293, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 8130 + }, + { + "epoch": 0.18833082706766918, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 8131 + }, + { + "epoch": 0.18839097744360903, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2161, + "step": 8132 + }, + { + "epoch": 0.18845112781954887, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2356, + "step": 8133 + }, + { + "epoch": 0.1885112781954887, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.3085, + "step": 8134 + }, + { + "epoch": 0.18857142857142858, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.3948, + "step": 8135 + }, + { + "epoch": 0.18863157894736843, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2375, + "step": 8136 + }, + { + "epoch": 0.18869172932330827, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2953, + "step": 8137 + }, + { + "epoch": 0.18875187969924812, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2378, + "step": 8138 + }, + { + "epoch": 0.18881203007518796, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 8139 + }, + { + "epoch": 0.18887218045112783, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2781, + "step": 8140 + }, + { + "epoch": 0.18893233082706767, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2667, + "step": 8141 + }, + { + "epoch": 0.18899248120300752, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2037, + "step": 8142 + }, + { + "epoch": 0.18905263157894736, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2014, + "step": 8143 + }, + { + "epoch": 0.1891127819548872, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 8144 + }, + { + "epoch": 0.18917293233082708, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2048, + "step": 8145 + }, + { + "epoch": 0.18923308270676692, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2976, + "step": 8146 + }, + { + "epoch": 0.18929323308270676, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2426, + "step": 8147 + }, + { + "epoch": 0.1893533834586466, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 8148 + }, + { + "epoch": 0.18941353383458648, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2767, + "step": 8149 + }, + { + "epoch": 0.18947368421052632, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.223, + "step": 8150 + }, + { + "epoch": 0.18953383458646617, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 8151 + }, + { + "epoch": 0.189593984962406, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2972, + "step": 8152 + }, + { + "epoch": 0.18965413533834585, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2135, + "step": 8153 + }, + { + "epoch": 0.18971428571428572, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 8154 + }, + { + "epoch": 0.18977443609022557, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3316, + "step": 8155 + }, + { + "epoch": 0.1898345864661654, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2185, + "step": 8156 + }, + { + "epoch": 0.18989473684210526, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 8157 + }, + { + "epoch": 0.18995488721804513, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.24, + "step": 8158 + }, + { + "epoch": 0.19001503759398497, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2055, + "step": 8159 + }, + { + "epoch": 0.1900751879699248, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 8160 + }, + { + "epoch": 0.19013533834586466, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2669, + "step": 8161 + }, + { + "epoch": 0.1901954887218045, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 8162 + }, + { + "epoch": 0.19025563909774437, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1903, + "step": 8163 + }, + { + "epoch": 0.19031578947368422, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 8164 + }, + { + "epoch": 0.19037593984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.227, + "step": 8165 + }, + { + "epoch": 0.1904360902255639, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 8166 + }, + { + "epoch": 0.19049624060150375, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 8167 + }, + { + "epoch": 0.19055639097744362, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2128, + "step": 8168 + }, + { + "epoch": 0.19061654135338346, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2232, + "step": 8169 + }, + { + "epoch": 0.1906766917293233, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2784, + "step": 8170 + }, + { + "epoch": 0.19073684210526315, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.5229, + "step": 8171 + }, + { + "epoch": 0.19079699248120302, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2141, + "step": 8172 + }, + { + "epoch": 0.19085714285714286, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 8173 + }, + { + "epoch": 0.1909172932330827, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2286, + "step": 8174 + }, + { + "epoch": 0.19097744360902255, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1817, + "step": 8175 + }, + { + "epoch": 0.1910375939849624, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.232, + "step": 8176 + }, + { + "epoch": 0.19109774436090227, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 8177 + }, + { + "epoch": 0.1911578947368421, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2895, + "step": 8178 + }, + { + "epoch": 0.19121804511278195, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2897, + "step": 8179 + }, + { + "epoch": 0.1912781954887218, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2387, + "step": 8180 + }, + { + "epoch": 0.19133834586466164, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2581, + "step": 8181 + }, + { + "epoch": 0.1913984962406015, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3212, + "step": 8182 + }, + { + "epoch": 0.19145864661654136, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.332, + "step": 8183 + }, + { + "epoch": 0.1915187969924812, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.5479, + "step": 8184 + }, + { + "epoch": 0.19157894736842104, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.199, + "step": 8185 + }, + { + "epoch": 0.19163909774436091, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 8186 + }, + { + "epoch": 0.19169924812030076, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 8187 + }, + { + "epoch": 0.1917593984962406, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2928, + "step": 8188 + }, + { + "epoch": 0.19181954887218045, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2576, + "step": 8189 + }, + { + "epoch": 0.1918796992481203, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 8190 + }, + { + "epoch": 0.19193984962406016, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 8191 + }, + { + "epoch": 0.192, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2698, + "step": 8192 + }, + { + "epoch": 0.19206015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2304, + "step": 8193 + }, + { + "epoch": 0.1921203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 8194 + }, + { + "epoch": 0.19218045112781956, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2032, + "step": 8195 + }, + { + "epoch": 0.1922406015037594, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2169, + "step": 8196 + }, + { + "epoch": 0.19230075187969925, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 8197 + }, + { + "epoch": 0.1923609022556391, + "grad_norm": 1.6484375, + "learning_rate": 3e-05, + "loss": 1.3426, + "step": 8198 + }, + { + "epoch": 0.19242105263157894, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1643, + "step": 8199 + }, + { + "epoch": 0.1924812030075188, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2712, + "step": 8200 + }, + { + "epoch": 0.19254135338345865, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2253, + "step": 8201 + }, + { + "epoch": 0.1926015037593985, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.1907, + "step": 8202 + }, + { + "epoch": 0.19266165413533834, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 8203 + }, + { + "epoch": 0.19272180451127818, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 8204 + }, + { + "epoch": 0.19278195488721805, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2966, + "step": 8205 + }, + { + "epoch": 0.1928421052631579, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2163, + "step": 8206 + }, + { + "epoch": 0.19290225563909774, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 8207 + }, + { + "epoch": 0.19296240601503759, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2485, + "step": 8208 + }, + { + "epoch": 0.19302255639097746, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2457, + "step": 8209 + }, + { + "epoch": 0.1930827067669173, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2346, + "step": 8210 + }, + { + "epoch": 0.19314285714285714, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.4375, + "step": 8211 + }, + { + "epoch": 0.193203007518797, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 8212 + }, + { + "epoch": 0.19326315789473683, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2853, + "step": 8213 + }, + { + "epoch": 0.1933233082706767, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 8214 + }, + { + "epoch": 0.19338345864661655, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 8215 + }, + { + "epoch": 0.1934436090225564, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 8216 + }, + { + "epoch": 0.19350375939849623, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 8217 + }, + { + "epoch": 0.19356390977443608, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2732, + "step": 8218 + }, + { + "epoch": 0.19362406015037595, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3582, + "step": 8219 + }, + { + "epoch": 0.1936842105263158, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2714, + "step": 8220 + }, + { + "epoch": 0.19374436090225564, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2393, + "step": 8221 + }, + { + "epoch": 0.19380451127819548, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 8222 + }, + { + "epoch": 0.19386466165413535, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2328, + "step": 8223 + }, + { + "epoch": 0.1939248120300752, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 8224 + }, + { + "epoch": 0.19398496240601504, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 8225 + }, + { + "epoch": 0.19404511278195488, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 8226 + }, + { + "epoch": 0.19410526315789473, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 8227 + }, + { + "epoch": 0.1941654135338346, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.2492, + "step": 8228 + }, + { + "epoch": 0.19422556390977444, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 8229 + }, + { + "epoch": 0.19428571428571428, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2756, + "step": 8230 + }, + { + "epoch": 0.19434586466165413, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 8231 + }, + { + "epoch": 0.194406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2058, + "step": 8232 + }, + { + "epoch": 0.19446616541353384, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 8233 + }, + { + "epoch": 0.19452631578947369, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2507, + "step": 8234 + }, + { + "epoch": 0.19458646616541353, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 8235 + }, + { + "epoch": 0.19464661654135337, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 8236 + }, + { + "epoch": 0.19470676691729324, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.23, + "step": 8237 + }, + { + "epoch": 0.1947669172932331, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2181, + "step": 8238 + }, + { + "epoch": 0.19482706766917293, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 8239 + }, + { + "epoch": 0.19488721804511278, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 8240 + }, + { + "epoch": 0.19494736842105262, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2057, + "step": 8241 + }, + { + "epoch": 0.1950075187969925, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 8242 + }, + { + "epoch": 0.19506766917293233, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 8243 + }, + { + "epoch": 0.19512781954887218, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 8244 + }, + { + "epoch": 0.19518796992481202, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2032, + "step": 8245 + }, + { + "epoch": 0.1952481203007519, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 8246 + }, + { + "epoch": 0.19530827067669174, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3071, + "step": 8247 + }, + { + "epoch": 0.19536842105263158, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 8248 + }, + { + "epoch": 0.19542857142857142, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 8249 + }, + { + "epoch": 0.19548872180451127, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2474, + "step": 8250 + }, + { + "epoch": 0.19554887218045114, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2058, + "step": 8251 + }, + { + "epoch": 0.19560902255639098, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2978, + "step": 8252 + }, + { + "epoch": 0.19566917293233083, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 8253 + }, + { + "epoch": 0.19572932330827067, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 8254 + }, + { + "epoch": 0.1957894736842105, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.3314, + "step": 8255 + }, + { + "epoch": 0.19584962406015038, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2602, + "step": 8256 + }, + { + "epoch": 0.19590977443609023, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1888, + "step": 8257 + }, + { + "epoch": 0.19596992481203007, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2635, + "step": 8258 + }, + { + "epoch": 0.19603007518796992, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 8259 + }, + { + "epoch": 0.1960902255639098, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.213, + "step": 8260 + }, + { + "epoch": 0.19615037593984963, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2951, + "step": 8261 + }, + { + "epoch": 0.19621052631578947, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1915, + "step": 8262 + }, + { + "epoch": 0.19627067669172932, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2546, + "step": 8263 + }, + { + "epoch": 0.19633082706766916, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1781, + "step": 8264 + }, + { + "epoch": 0.19639097744360903, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2166, + "step": 8265 + }, + { + "epoch": 0.19645112781954888, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2052, + "step": 8266 + }, + { + "epoch": 0.19651127819548872, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1746, + "step": 8267 + }, + { + "epoch": 0.19657142857142856, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 8268 + }, + { + "epoch": 0.1966315789473684, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2439, + "step": 8269 + }, + { + "epoch": 0.19669172932330828, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.187, + "step": 8270 + }, + { + "epoch": 0.19675187969924812, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 8271 + }, + { + "epoch": 0.19681203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2162, + "step": 8272 + }, + { + "epoch": 0.1968721804511278, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 8273 + }, + { + "epoch": 0.19693233082706768, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 8274 + }, + { + "epoch": 0.19699248120300752, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2284, + "step": 8275 + }, + { + "epoch": 0.19705263157894737, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.236, + "step": 8276 + }, + { + "epoch": 0.1971127819548872, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3792, + "step": 8277 + }, + { + "epoch": 0.19717293233082706, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1567, + "step": 8278 + }, + { + "epoch": 0.19723308270676693, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1894, + "step": 8279 + }, + { + "epoch": 0.19729323308270677, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 8280 + }, + { + "epoch": 0.19735338345864661, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1652, + "step": 8281 + }, + { + "epoch": 0.19741353383458646, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 8282 + }, + { + "epoch": 0.19747368421052633, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 8283 + }, + { + "epoch": 0.19753383458646617, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 8284 + }, + { + "epoch": 0.19759398496240602, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 8285 + }, + { + "epoch": 0.19765413533834586, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 8286 + }, + { + "epoch": 0.1977142857142857, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 8287 + }, + { + "epoch": 0.19777443609022557, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2331, + "step": 8288 + }, + { + "epoch": 0.19783458646616542, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 8289 + }, + { + "epoch": 0.19789473684210526, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.245, + "step": 8290 + }, + { + "epoch": 0.1979548872180451, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2911, + "step": 8291 + }, + { + "epoch": 0.19801503759398495, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 8292 + }, + { + "epoch": 0.19807518796992482, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2488, + "step": 8293 + }, + { + "epoch": 0.19813533834586466, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 8294 + }, + { + "epoch": 0.1981954887218045, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2401, + "step": 8295 + }, + { + "epoch": 0.19825563909774435, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 8296 + }, + { + "epoch": 0.19831578947368422, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 8297 + }, + { + "epoch": 0.19837593984962407, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2135, + "step": 8298 + }, + { + "epoch": 0.1984360902255639, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 8299 + }, + { + "epoch": 0.19849624060150375, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2394, + "step": 8300 + }, + { + "epoch": 0.1985563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3268, + "step": 8301 + }, + { + "epoch": 0.19861654135338347, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2004, + "step": 8302 + }, + { + "epoch": 0.1986766917293233, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2644, + "step": 8303 + }, + { + "epoch": 0.19873684210526316, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 8304 + }, + { + "epoch": 0.198796992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 8305 + }, + { + "epoch": 0.19885714285714284, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3115, + "step": 8306 + }, + { + "epoch": 0.19891729323308271, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 8307 + }, + { + "epoch": 0.19897744360902256, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2615, + "step": 8308 + }, + { + "epoch": 0.1990375939849624, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3331, + "step": 8309 + }, + { + "epoch": 0.19909774436090225, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2192, + "step": 8310 + }, + { + "epoch": 0.19915789473684212, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.27, + "step": 8311 + }, + { + "epoch": 0.19921804511278196, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2569, + "step": 8312 + }, + { + "epoch": 0.1992781954887218, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1823, + "step": 8313 + }, + { + "epoch": 0.19933834586466165, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 8314 + }, + { + "epoch": 0.1993984962406015, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 8315 + }, + { + "epoch": 0.19945864661654136, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 8316 + }, + { + "epoch": 0.1995187969924812, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 8317 + }, + { + "epoch": 0.19957894736842105, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 8318 + }, + { + "epoch": 0.1996390977443609, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2243, + "step": 8319 + }, + { + "epoch": 0.19969924812030077, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2414, + "step": 8320 + }, + { + "epoch": 0.1997593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 8321 + }, + { + "epoch": 0.19981954887218045, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2382, + "step": 8322 + }, + { + "epoch": 0.1998796992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2608, + "step": 8323 + }, + { + "epoch": 0.19993984962406014, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 8324 + }, + { + "epoch": 0.2, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2352, + "step": 8325 + }, + { + "epoch": 0.20006015037593985, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.388, + "step": 8326 + }, + { + "epoch": 0.2001203007518797, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2476, + "step": 8327 + }, + { + "epoch": 0.20018045112781954, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3484, + "step": 8328 + }, + { + "epoch": 0.20024060150375939, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 8329 + }, + { + "epoch": 0.20030075187969926, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 8330 + }, + { + "epoch": 0.2003609022556391, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1843, + "step": 8331 + }, + { + "epoch": 0.20042105263157894, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2846, + "step": 8332 + }, + { + "epoch": 0.2004812030075188, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 8333 + }, + { + "epoch": 0.20054135338345866, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2671, + "step": 8334 + }, + { + "epoch": 0.2006015037593985, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2558, + "step": 8335 + }, + { + "epoch": 0.20066165413533835, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 8336 + }, + { + "epoch": 0.2007218045112782, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 8337 + }, + { + "epoch": 0.20078195488721803, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 8338 + }, + { + "epoch": 0.2008421052631579, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2356, + "step": 8339 + }, + { + "epoch": 0.20090225563909775, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 8340 + }, + { + "epoch": 0.2009624060150376, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 8341 + }, + { + "epoch": 0.20102255639097744, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 8342 + }, + { + "epoch": 0.20108270676691728, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2224, + "step": 8343 + }, + { + "epoch": 0.20114285714285715, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2318, + "step": 8344 + }, + { + "epoch": 0.201203007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1861, + "step": 8345 + }, + { + "epoch": 0.20126315789473684, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2581, + "step": 8346 + }, + { + "epoch": 0.20132330827067668, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 8347 + }, + { + "epoch": 0.20138345864661655, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 8348 + }, + { + "epoch": 0.2014436090225564, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2405, + "step": 8349 + }, + { + "epoch": 0.20150375939849624, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2546, + "step": 8350 + }, + { + "epoch": 0.20156390977443608, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 8351 + }, + { + "epoch": 0.20162406015037593, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.2879, + "step": 8352 + }, + { + "epoch": 0.2016842105263158, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2807, + "step": 8353 + }, + { + "epoch": 0.20174436090225564, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 8354 + }, + { + "epoch": 0.2018045112781955, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3203, + "step": 8355 + }, + { + "epoch": 0.20186466165413533, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 8356 + }, + { + "epoch": 0.2019248120300752, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2539, + "step": 8357 + }, + { + "epoch": 0.20198496240601505, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2513, + "step": 8358 + }, + { + "epoch": 0.2020451127819549, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2976, + "step": 8359 + }, + { + "epoch": 0.20210526315789473, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2638, + "step": 8360 + }, + { + "epoch": 0.20216541353383458, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3019, + "step": 8361 + }, + { + "epoch": 0.20222556390977445, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2193, + "step": 8362 + }, + { + "epoch": 0.2022857142857143, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2351, + "step": 8363 + }, + { + "epoch": 0.20234586466165413, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 8364 + }, + { + "epoch": 0.20240601503759398, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1986, + "step": 8365 + }, + { + "epoch": 0.20246616541353382, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2092, + "step": 8366 + }, + { + "epoch": 0.2025263157894737, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.294, + "step": 8367 + }, + { + "epoch": 0.20258646616541354, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2696, + "step": 8368 + }, + { + "epoch": 0.20264661654135338, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3551, + "step": 8369 + }, + { + "epoch": 0.20270676691729322, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 8370 + }, + { + "epoch": 0.2027669172932331, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.187, + "step": 8371 + }, + { + "epoch": 0.20282706766917294, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3407, + "step": 8372 + }, + { + "epoch": 0.20288721804511278, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 8373 + }, + { + "epoch": 0.20294736842105263, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 8374 + }, + { + "epoch": 0.20300751879699247, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 8375 + }, + { + "epoch": 0.20306766917293234, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.4222, + "step": 8376 + }, + { + "epoch": 0.20312781954887218, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1761, + "step": 8377 + }, + { + "epoch": 0.20318796992481203, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2179, + "step": 8378 + }, + { + "epoch": 0.20324812030075187, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2669, + "step": 8379 + }, + { + "epoch": 0.20330827067669172, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.359, + "step": 8380 + }, + { + "epoch": 0.2033684210526316, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3092, + "step": 8381 + }, + { + "epoch": 0.20342857142857143, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.2558, + "step": 8382 + }, + { + "epoch": 0.20348872180451127, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.4198, + "step": 8383 + }, + { + "epoch": 0.20354887218045112, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 8384 + }, + { + "epoch": 0.203609022556391, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2381, + "step": 8385 + }, + { + "epoch": 0.20366917293233083, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 8386 + }, + { + "epoch": 0.20372932330827068, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2732, + "step": 8387 + }, + { + "epoch": 0.20378947368421052, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.238, + "step": 8388 + }, + { + "epoch": 0.20384962406015036, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2933, + "step": 8389 + }, + { + "epoch": 0.20390977443609024, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 8390 + }, + { + "epoch": 0.20396992481203008, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 8391 + }, + { + "epoch": 0.20403007518796992, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3594, + "step": 8392 + }, + { + "epoch": 0.20409022556390977, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 8393 + }, + { + "epoch": 0.20415037593984964, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3093, + "step": 8394 + }, + { + "epoch": 0.20421052631578948, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2478, + "step": 8395 + }, + { + "epoch": 0.20427067669172932, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2935, + "step": 8396 + }, + { + "epoch": 0.20433082706766917, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1604, + "step": 8397 + }, + { + "epoch": 0.204390977443609, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1961, + "step": 8398 + }, + { + "epoch": 0.20445112781954888, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2898, + "step": 8399 + }, + { + "epoch": 0.20451127819548873, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2097, + "step": 8400 + }, + { + "epoch": 0.20457142857142857, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 8401 + }, + { + "epoch": 0.20463157894736841, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2692, + "step": 8402 + }, + { + "epoch": 0.20469172932330826, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2412, + "step": 8403 + }, + { + "epoch": 0.20475187969924813, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2633, + "step": 8404 + }, + { + "epoch": 0.20481203007518797, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 8405 + }, + { + "epoch": 0.20487218045112782, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 8406 + }, + { + "epoch": 0.20493233082706766, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3334, + "step": 8407 + }, + { + "epoch": 0.20499248120300753, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1881, + "step": 8408 + }, + { + "epoch": 0.20505263157894738, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3956, + "step": 8409 + }, + { + "epoch": 0.20511278195488722, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2255, + "step": 8410 + }, + { + "epoch": 0.20517293233082706, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 8411 + }, + { + "epoch": 0.2052330827067669, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 8412 + }, + { + "epoch": 0.20529323308270678, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1671, + "step": 8413 + }, + { + "epoch": 0.20535338345864662, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.214, + "step": 8414 + }, + { + "epoch": 0.20541353383458646, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 8415 + }, + { + "epoch": 0.2054736842105263, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2814, + "step": 8416 + }, + { + "epoch": 0.20553383458646615, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.245, + "step": 8417 + }, + { + "epoch": 0.20559398496240602, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1646, + "step": 8418 + }, + { + "epoch": 0.20565413533834587, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1728, + "step": 8419 + }, + { + "epoch": 0.2057142857142857, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 8420 + }, + { + "epoch": 0.20577443609022555, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2883, + "step": 8421 + }, + { + "epoch": 0.20583458646616543, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2004, + "step": 8422 + }, + { + "epoch": 0.20589473684210527, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 8423 + }, + { + "epoch": 0.2059548872180451, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 8424 + }, + { + "epoch": 0.20601503759398496, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2321, + "step": 8425 + }, + { + "epoch": 0.2060751879699248, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2632, + "step": 8426 + }, + { + "epoch": 0.20613533834586467, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 8427 + }, + { + "epoch": 0.20619548872180452, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 8428 + }, + { + "epoch": 0.20625563909774436, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2513, + "step": 8429 + }, + { + "epoch": 0.2063157894736842, + "grad_norm": 1.6640625, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 8430 + }, + { + "epoch": 0.20637593984962407, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1637, + "step": 8431 + }, + { + "epoch": 0.20643609022556392, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 8432 + }, + { + "epoch": 0.20649624060150376, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2754, + "step": 8433 + }, + { + "epoch": 0.2065563909774436, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 8434 + }, + { + "epoch": 0.20661654135338345, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1643, + "step": 8435 + }, + { + "epoch": 0.20667669172932332, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2178, + "step": 8436 + }, + { + "epoch": 0.20673684210526316, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2095, + "step": 8437 + }, + { + "epoch": 0.206796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2329, + "step": 8438 + }, + { + "epoch": 0.20685714285714285, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 8439 + }, + { + "epoch": 0.2069172932330827, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2025, + "step": 8440 + }, + { + "epoch": 0.20697744360902257, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 8441 + }, + { + "epoch": 0.2070375939849624, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1732, + "step": 8442 + }, + { + "epoch": 0.20709774436090225, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3227, + "step": 8443 + }, + { + "epoch": 0.2071578947368421, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2377, + "step": 8444 + }, + { + "epoch": 0.20721804511278197, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.2229, + "step": 8445 + }, + { + "epoch": 0.2072781954887218, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2721, + "step": 8446 + }, + { + "epoch": 0.20733834586466166, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.4295, + "step": 8447 + }, + { + "epoch": 0.2073984962406015, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2553, + "step": 8448 + }, + { + "epoch": 0.20745864661654134, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2691, + "step": 8449 + }, + { + "epoch": 0.2075187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2916, + "step": 8450 + }, + { + "epoch": 0.20757894736842106, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 8451 + }, + { + "epoch": 0.2076390977443609, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2737, + "step": 8452 + }, + { + "epoch": 0.20769924812030074, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2197, + "step": 8453 + }, + { + "epoch": 0.2077593984962406, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 8454 + }, + { + "epoch": 0.20781954887218046, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 8455 + }, + { + "epoch": 0.2078796992481203, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.2628, + "step": 8456 + }, + { + "epoch": 0.20793984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.252, + "step": 8457 + }, + { + "epoch": 0.208, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2744, + "step": 8458 + }, + { + "epoch": 0.20806015037593986, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.262, + "step": 8459 + }, + { + "epoch": 0.2081203007518797, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.2074, + "step": 8460 + }, + { + "epoch": 0.20818045112781955, + "grad_norm": 0.435546875, + "learning_rate": 3e-05, + "loss": 1.2336, + "step": 8461 + }, + { + "epoch": 0.2082406015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 8462 + }, + { + "epoch": 0.20830075187969924, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2323, + "step": 8463 + }, + { + "epoch": 0.2083609022556391, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 8464 + }, + { + "epoch": 0.20842105263157895, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 8465 + }, + { + "epoch": 0.2084812030075188, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2434, + "step": 8466 + }, + { + "epoch": 0.20854135338345864, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 8467 + }, + { + "epoch": 0.2086015037593985, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 8468 + }, + { + "epoch": 0.20866165413533835, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2362, + "step": 8469 + }, + { + "epoch": 0.2087218045112782, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1938, + "step": 8470 + }, + { + "epoch": 0.20878195488721804, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2088, + "step": 8471 + }, + { + "epoch": 0.20884210526315788, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 8472 + }, + { + "epoch": 0.20890225563909776, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 8473 + }, + { + "epoch": 0.2089624060150376, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 8474 + }, + { + "epoch": 0.20902255639097744, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 8475 + }, + { + "epoch": 0.2090827067669173, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2788, + "step": 8476 + }, + { + "epoch": 0.20914285714285713, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2776, + "step": 8477 + }, + { + "epoch": 0.209203007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 8478 + }, + { + "epoch": 0.20926315789473685, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.7869, + "step": 8479 + }, + { + "epoch": 0.2093233082706767, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.227, + "step": 8480 + }, + { + "epoch": 0.20938345864661653, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 8481 + }, + { + "epoch": 0.2094436090225564, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 8482 + }, + { + "epoch": 0.20950375939849625, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2033, + "step": 8483 + }, + { + "epoch": 0.2095639097744361, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2294, + "step": 8484 + }, + { + "epoch": 0.20962406015037593, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2598, + "step": 8485 + }, + { + "epoch": 0.20968421052631578, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 8486 + }, + { + "epoch": 0.20974436090225565, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 8487 + }, + { + "epoch": 0.2098045112781955, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2603, + "step": 8488 + }, + { + "epoch": 0.20986466165413534, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 8489 + }, + { + "epoch": 0.20992481203007518, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 8490 + }, + { + "epoch": 0.20998496240601502, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 8491 + }, + { + "epoch": 0.2100451127819549, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 8492 + }, + { + "epoch": 0.21010526315789474, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 8493 + }, + { + "epoch": 0.21016541353383458, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 8494 + }, + { + "epoch": 0.21022556390977443, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2969, + "step": 8495 + }, + { + "epoch": 0.2102857142857143, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 8496 + }, + { + "epoch": 0.21034586466165414, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 8497 + }, + { + "epoch": 0.21040601503759399, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 8498 + }, + { + "epoch": 0.21046616541353383, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 8499 + }, + { + "epoch": 0.21052631578947367, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1429, + "step": 8500 + }, + { + "epoch": 0.21058646616541354, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.4184, + "step": 8501 + }, + { + "epoch": 0.2106466165413534, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 8502 + }, + { + "epoch": 0.21070676691729323, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 8503 + }, + { + "epoch": 0.21076691729323307, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2226, + "step": 8504 + }, + { + "epoch": 0.21082706766917295, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.3268, + "step": 8505 + }, + { + "epoch": 0.2108872180451128, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 8506 + }, + { + "epoch": 0.21094736842105263, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 8507 + }, + { + "epoch": 0.21100751879699248, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2627, + "step": 8508 + }, + { + "epoch": 0.21106766917293232, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 8509 + }, + { + "epoch": 0.2111278195488722, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.264, + "step": 8510 + }, + { + "epoch": 0.21118796992481204, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 8511 + }, + { + "epoch": 0.21124812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2985, + "step": 8512 + }, + { + "epoch": 0.21130827067669172, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1965, + "step": 8513 + }, + { + "epoch": 0.21136842105263157, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2563, + "step": 8514 + }, + { + "epoch": 0.21142857142857144, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 8515 + }, + { + "epoch": 0.21148872180451128, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2013, + "step": 8516 + }, + { + "epoch": 0.21154887218045113, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 8517 + }, + { + "epoch": 0.21160902255639097, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 8518 + }, + { + "epoch": 0.21166917293233084, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.266, + "step": 8519 + }, + { + "epoch": 0.21172932330827068, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2329, + "step": 8520 + }, + { + "epoch": 0.21178947368421053, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1817, + "step": 8521 + }, + { + "epoch": 0.21184962406015037, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 8522 + }, + { + "epoch": 0.21190977443609021, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1567, + "step": 8523 + }, + { + "epoch": 0.21196992481203009, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 8524 + }, + { + "epoch": 0.21203007518796993, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2401, + "step": 8525 + }, + { + "epoch": 0.21209022556390977, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3421, + "step": 8526 + }, + { + "epoch": 0.21215037593984962, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2221, + "step": 8527 + }, + { + "epoch": 0.21221052631578946, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 8528 + }, + { + "epoch": 0.21227067669172933, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2994, + "step": 8529 + }, + { + "epoch": 0.21233082706766918, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2861, + "step": 8530 + }, + { + "epoch": 0.21239097744360902, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.3794, + "step": 8531 + }, + { + "epoch": 0.21245112781954886, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2882, + "step": 8532 + }, + { + "epoch": 0.21251127819548873, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 8533 + }, + { + "epoch": 0.21257142857142858, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 8534 + }, + { + "epoch": 0.21263157894736842, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.4241, + "step": 8535 + }, + { + "epoch": 0.21269172932330827, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 8536 + }, + { + "epoch": 0.2127518796992481, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2339, + "step": 8537 + }, + { + "epoch": 0.21281203007518798, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2283, + "step": 8538 + }, + { + "epoch": 0.21287218045112782, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2186, + "step": 8539 + }, + { + "epoch": 0.21293233082706767, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 8540 + }, + { + "epoch": 0.2129924812030075, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 8541 + }, + { + "epoch": 0.21305263157894735, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 8542 + }, + { + "epoch": 0.21311278195488723, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 8543 + }, + { + "epoch": 0.21317293233082707, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1786, + "step": 8544 + }, + { + "epoch": 0.2132330827067669, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1651, + "step": 8545 + }, + { + "epoch": 0.21329323308270676, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2349, + "step": 8546 + }, + { + "epoch": 0.21335338345864663, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 8547 + }, + { + "epoch": 0.21341353383458647, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2581, + "step": 8548 + }, + { + "epoch": 0.21347368421052632, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3236, + "step": 8549 + }, + { + "epoch": 0.21353383458646616, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3166, + "step": 8550 + }, + { + "epoch": 0.213593984962406, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1919, + "step": 8551 + }, + { + "epoch": 0.21365413533834587, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 8552 + }, + { + "epoch": 0.21371428571428572, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3582, + "step": 8553 + }, + { + "epoch": 0.21377443609022556, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1903, + "step": 8554 + }, + { + "epoch": 0.2138345864661654, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2856, + "step": 8555 + }, + { + "epoch": 0.21389473684210528, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2373, + "step": 8556 + }, + { + "epoch": 0.21395488721804512, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2389, + "step": 8557 + }, + { + "epoch": 0.21401503759398496, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1723, + "step": 8558 + }, + { + "epoch": 0.2140751879699248, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2846, + "step": 8559 + }, + { + "epoch": 0.21413533834586465, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 8560 + }, + { + "epoch": 0.21419548872180452, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2018, + "step": 8561 + }, + { + "epoch": 0.21425563909774437, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2085, + "step": 8562 + }, + { + "epoch": 0.2143157894736842, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2095, + "step": 8563 + }, + { + "epoch": 0.21437593984962405, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 8564 + }, + { + "epoch": 0.2144360902255639, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.3816, + "step": 8565 + }, + { + "epoch": 0.21449624060150377, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 8566 + }, + { + "epoch": 0.2145563909774436, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.3007, + "step": 8567 + }, + { + "epoch": 0.21461654135338346, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 8568 + }, + { + "epoch": 0.2146766917293233, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2726, + "step": 8569 + }, + { + "epoch": 0.21473684210526317, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.267, + "step": 8570 + }, + { + "epoch": 0.21479699248120301, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 8571 + }, + { + "epoch": 0.21485714285714286, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.218, + "step": 8572 + }, + { + "epoch": 0.2149172932330827, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2204, + "step": 8573 + }, + { + "epoch": 0.21497744360902254, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 8574 + }, + { + "epoch": 0.21503759398496242, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2748, + "step": 8575 + }, + { + "epoch": 0.21509774436090226, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2878, + "step": 8576 + }, + { + "epoch": 0.2151578947368421, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2298, + "step": 8577 + }, + { + "epoch": 0.21521804511278195, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2279, + "step": 8578 + }, + { + "epoch": 0.2152781954887218, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.207, + "step": 8579 + }, + { + "epoch": 0.21533834586466166, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1526, + "step": 8580 + }, + { + "epoch": 0.2153984962406015, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2183, + "step": 8581 + }, + { + "epoch": 0.21545864661654135, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 8582 + }, + { + "epoch": 0.2155187969924812, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2235, + "step": 8583 + }, + { + "epoch": 0.21557894736842106, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2366, + "step": 8584 + }, + { + "epoch": 0.2156390977443609, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.237, + "step": 8585 + }, + { + "epoch": 0.21569924812030075, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 8586 + }, + { + "epoch": 0.2157593984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.314, + "step": 8587 + }, + { + "epoch": 0.21581954887218044, + "grad_norm": 1.5, + "learning_rate": 3e-05, + "loss": 1.3524, + "step": 8588 + }, + { + "epoch": 0.2158796992481203, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2706, + "step": 8589 + }, + { + "epoch": 0.21593984962406015, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1738, + "step": 8590 + }, + { + "epoch": 0.216, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 8591 + }, + { + "epoch": 0.21606015037593984, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 8592 + }, + { + "epoch": 0.2161203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2726, + "step": 8593 + }, + { + "epoch": 0.21618045112781956, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2433, + "step": 8594 + }, + { + "epoch": 0.2162406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1833, + "step": 8595 + }, + { + "epoch": 0.21630075187969924, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 8596 + }, + { + "epoch": 0.2163609022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 8597 + }, + { + "epoch": 0.21642105263157896, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.3238, + "step": 8598 + }, + { + "epoch": 0.2164812030075188, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2261, + "step": 8599 + }, + { + "epoch": 0.21654135338345865, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 8600 + }, + { + "epoch": 0.2166015037593985, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2162, + "step": 8601 + }, + { + "epoch": 0.21666165413533833, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 8602 + }, + { + "epoch": 0.2167218045112782, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.35, + "step": 8603 + }, + { + "epoch": 0.21678195488721805, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2459, + "step": 8604 + }, + { + "epoch": 0.2168421052631579, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 8605 + }, + { + "epoch": 0.21690225563909774, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2211, + "step": 8606 + }, + { + "epoch": 0.2169624060150376, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2119, + "step": 8607 + }, + { + "epoch": 0.21702255639097745, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 8608 + }, + { + "epoch": 0.2170827067669173, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2149, + "step": 8609 + }, + { + "epoch": 0.21714285714285714, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2401, + "step": 8610 + }, + { + "epoch": 0.21720300751879698, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 8611 + }, + { + "epoch": 0.21726315789473685, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.203, + "step": 8612 + }, + { + "epoch": 0.2173233082706767, + "grad_norm": 1.4140625, + "learning_rate": 3e-05, + "loss": 1.2283, + "step": 8613 + }, + { + "epoch": 0.21738345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1822, + "step": 8614 + }, + { + "epoch": 0.21744360902255638, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1966, + "step": 8615 + }, + { + "epoch": 0.21750375939849623, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2494, + "step": 8616 + }, + { + "epoch": 0.2175639097744361, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 8617 + }, + { + "epoch": 0.21762406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2171, + "step": 8618 + }, + { + "epoch": 0.21768421052631579, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2362, + "step": 8619 + }, + { + "epoch": 0.21774436090225563, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.226, + "step": 8620 + }, + { + "epoch": 0.2178045112781955, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 8621 + }, + { + "epoch": 0.21786466165413534, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 8622 + }, + { + "epoch": 0.2179248120300752, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 8623 + }, + { + "epoch": 0.21798496240601503, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 8624 + }, + { + "epoch": 0.21804511278195488, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 8625 + }, + { + "epoch": 0.21810526315789475, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2163, + "step": 8626 + }, + { + "epoch": 0.2181654135338346, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2682, + "step": 8627 + }, + { + "epoch": 0.21822556390977443, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.225, + "step": 8628 + }, + { + "epoch": 0.21828571428571428, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 8629 + }, + { + "epoch": 0.21834586466165415, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2854, + "step": 8630 + }, + { + "epoch": 0.218406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 8631 + }, + { + "epoch": 0.21846616541353384, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.1773, + "step": 8632 + }, + { + "epoch": 0.21852631578947368, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.2811, + "step": 8633 + }, + { + "epoch": 0.21858646616541352, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2972, + "step": 8634 + }, + { + "epoch": 0.2186466165413534, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 8635 + }, + { + "epoch": 0.21870676691729324, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2648, + "step": 8636 + }, + { + "epoch": 0.21876691729323308, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2389, + "step": 8637 + }, + { + "epoch": 0.21882706766917293, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2726, + "step": 8638 + }, + { + "epoch": 0.21888721804511277, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 8639 + }, + { + "epoch": 0.21894736842105264, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 8640 + }, + { + "epoch": 0.21900751879699248, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1982, + "step": 8641 + }, + { + "epoch": 0.21906766917293233, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 8642 + }, + { + "epoch": 0.21912781954887217, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 8643 + }, + { + "epoch": 0.21918796992481204, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2027, + "step": 8644 + }, + { + "epoch": 0.2192481203007519, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2595, + "step": 8645 + }, + { + "epoch": 0.21930827067669173, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2531, + "step": 8646 + }, + { + "epoch": 0.21936842105263157, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2253, + "step": 8647 + }, + { + "epoch": 0.21942857142857142, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 8648 + }, + { + "epoch": 0.2194887218045113, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 8649 + }, + { + "epoch": 0.21954887218045113, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 8650 + }, + { + "epoch": 0.21960902255639098, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2995, + "step": 8651 + }, + { + "epoch": 0.21966917293233082, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2138, + "step": 8652 + }, + { + "epoch": 0.21972932330827066, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2488, + "step": 8653 + }, + { + "epoch": 0.21978947368421053, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3535, + "step": 8654 + }, + { + "epoch": 0.21984962406015038, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 8655 + }, + { + "epoch": 0.21990977443609022, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2461, + "step": 8656 + }, + { + "epoch": 0.21996992481203007, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 8657 + }, + { + "epoch": 0.22003007518796994, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 8658 + }, + { + "epoch": 0.22009022556390978, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2034, + "step": 8659 + }, + { + "epoch": 0.22015037593984962, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2087, + "step": 8660 + }, + { + "epoch": 0.22021052631578947, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2176, + "step": 8661 + }, + { + "epoch": 0.2202706766917293, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.2637, + "step": 8662 + }, + { + "epoch": 0.22033082706766918, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 8663 + }, + { + "epoch": 0.22039097744360903, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2005, + "step": 8664 + }, + { + "epoch": 0.22045112781954887, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.236, + "step": 8665 + }, + { + "epoch": 0.2205112781954887, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3411, + "step": 8666 + }, + { + "epoch": 0.22057142857142858, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2403, + "step": 8667 + }, + { + "epoch": 0.22063157894736843, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 8668 + }, + { + "epoch": 0.22069172932330827, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 8669 + }, + { + "epoch": 0.22075187969924812, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.229, + "step": 8670 + }, + { + "epoch": 0.22081203007518796, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2062, + "step": 8671 + }, + { + "epoch": 0.22087218045112783, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2937, + "step": 8672 + }, + { + "epoch": 0.22093233082706767, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2713, + "step": 8673 + }, + { + "epoch": 0.22099248120300752, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 8674 + }, + { + "epoch": 0.22105263157894736, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2436, + "step": 8675 + }, + { + "epoch": 0.2211127819548872, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2138, + "step": 8676 + }, + { + "epoch": 0.22117293233082708, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 8677 + }, + { + "epoch": 0.22123308270676692, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 8678 + }, + { + "epoch": 0.22129323308270676, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 8679 + }, + { + "epoch": 0.2213533834586466, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2343, + "step": 8680 + }, + { + "epoch": 0.22141353383458648, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 8681 + }, + { + "epoch": 0.22147368421052632, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 8682 + }, + { + "epoch": 0.22153383458646617, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 8683 + }, + { + "epoch": 0.221593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2075, + "step": 8684 + }, + { + "epoch": 0.22165413533834585, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 8685 + }, + { + "epoch": 0.22171428571428572, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1529, + "step": 8686 + }, + { + "epoch": 0.22177443609022557, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 8687 + }, + { + "epoch": 0.2218345864661654, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2281, + "step": 8688 + }, + { + "epoch": 0.22189473684210526, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 8689 + }, + { + "epoch": 0.2219548872180451, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 8690 + }, + { + "epoch": 0.22201503759398497, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.216, + "step": 8691 + }, + { + "epoch": 0.22207518796992481, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 8692 + }, + { + "epoch": 0.22213533834586466, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1496, + "step": 8693 + }, + { + "epoch": 0.2221954887218045, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 8694 + }, + { + "epoch": 0.22225563909774437, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 8695 + }, + { + "epoch": 0.22231578947368422, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2654, + "step": 8696 + }, + { + "epoch": 0.22237593984962406, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2191, + "step": 8697 + }, + { + "epoch": 0.2224360902255639, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.3077, + "step": 8698 + }, + { + "epoch": 0.22249624060150375, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2313, + "step": 8699 + }, + { + "epoch": 0.22255639097744362, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1738, + "step": 8700 + }, + { + "epoch": 0.22261654135338346, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2255, + "step": 8701 + }, + { + "epoch": 0.2226766917293233, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2725, + "step": 8702 + }, + { + "epoch": 0.22273684210526315, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1646, + "step": 8703 + }, + { + "epoch": 0.22279699248120302, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2674, + "step": 8704 + }, + { + "epoch": 0.22285714285714286, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 8705 + }, + { + "epoch": 0.2229172932330827, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 8706 + }, + { + "epoch": 0.22297744360902255, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2066, + "step": 8707 + }, + { + "epoch": 0.2230375939849624, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 8708 + }, + { + "epoch": 0.22309774436090227, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2338, + "step": 8709 + }, + { + "epoch": 0.2231578947368421, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 8710 + }, + { + "epoch": 0.22321804511278195, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2866, + "step": 8711 + }, + { + "epoch": 0.2232781954887218, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2761, + "step": 8712 + }, + { + "epoch": 0.22333834586466164, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2606, + "step": 8713 + }, + { + "epoch": 0.2233984962406015, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2441, + "step": 8714 + }, + { + "epoch": 0.22345864661654136, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 8715 + }, + { + "epoch": 0.2235187969924812, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2035, + "step": 8716 + }, + { + "epoch": 0.22357894736842104, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2236, + "step": 8717 + }, + { + "epoch": 0.22363909774436092, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.2857, + "step": 8718 + }, + { + "epoch": 0.22369924812030076, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.274, + "step": 8719 + }, + { + "epoch": 0.2237593984962406, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2709, + "step": 8720 + }, + { + "epoch": 0.22381954887218045, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.276, + "step": 8721 + }, + { + "epoch": 0.2238796992481203, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 8722 + }, + { + "epoch": 0.22393984962406016, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2963, + "step": 8723 + }, + { + "epoch": 0.224, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 8724 + }, + { + "epoch": 0.22406015037593985, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 8725 + }, + { + "epoch": 0.2241203007518797, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1845, + "step": 8726 + }, + { + "epoch": 0.22418045112781954, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2341, + "step": 8727 + }, + { + "epoch": 0.2242406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 8728 + }, + { + "epoch": 0.22430075187969925, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 8729 + }, + { + "epoch": 0.2243609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2363, + "step": 8730 + }, + { + "epoch": 0.22442105263157894, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 8731 + }, + { + "epoch": 0.2244812030075188, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 8732 + }, + { + "epoch": 0.22454135338345865, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 8733 + }, + { + "epoch": 0.2246015037593985, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 8734 + }, + { + "epoch": 0.22466165413533834, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2537, + "step": 8735 + }, + { + "epoch": 0.22472180451127818, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 8736 + }, + { + "epoch": 0.22478195488721806, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 8737 + }, + { + "epoch": 0.2248421052631579, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.3151, + "step": 8738 + }, + { + "epoch": 0.22490225563909774, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 8739 + }, + { + "epoch": 0.22496240601503759, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2493, + "step": 8740 + }, + { + "epoch": 0.22502255639097746, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2206, + "step": 8741 + }, + { + "epoch": 0.2250827067669173, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 8742 + }, + { + "epoch": 0.22514285714285714, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1585, + "step": 8743 + }, + { + "epoch": 0.225203007518797, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2446, + "step": 8744 + }, + { + "epoch": 0.22526315789473683, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2336, + "step": 8745 + }, + { + "epoch": 0.2253233082706767, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2475, + "step": 8746 + }, + { + "epoch": 0.22538345864661655, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2594, + "step": 8747 + }, + { + "epoch": 0.2254436090225564, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 8748 + }, + { + "epoch": 0.22550375939849623, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 8749 + }, + { + "epoch": 0.22556390977443608, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2007, + "step": 8750 + }, + { + "epoch": 0.22562406015037595, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 8751 + }, + { + "epoch": 0.2256842105263158, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.189, + "step": 8752 + }, + { + "epoch": 0.22574436090225564, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 8753 + }, + { + "epoch": 0.22580451127819548, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1763, + "step": 8754 + }, + { + "epoch": 0.22586466165413535, + "grad_norm": 0.443359375, + "learning_rate": 3e-05, + "loss": 1.2133, + "step": 8755 + }, + { + "epoch": 0.2259248120300752, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2048, + "step": 8756 + }, + { + "epoch": 0.22598496240601504, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 8757 + }, + { + "epoch": 0.22604511278195488, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2116, + "step": 8758 + }, + { + "epoch": 0.22610526315789473, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 8759 + }, + { + "epoch": 0.2261654135338346, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.3889, + "step": 8760 + }, + { + "epoch": 0.22622556390977444, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2149, + "step": 8761 + }, + { + "epoch": 0.22628571428571428, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.308, + "step": 8762 + }, + { + "epoch": 0.22634586466165413, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2617, + "step": 8763 + }, + { + "epoch": 0.22640601503759397, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 8764 + }, + { + "epoch": 0.22646616541353384, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2846, + "step": 8765 + }, + { + "epoch": 0.2265263157894737, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 8766 + }, + { + "epoch": 0.22658646616541353, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 8767 + }, + { + "epoch": 0.22664661654135337, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2676, + "step": 8768 + }, + { + "epoch": 0.22670676691729325, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.199, + "step": 8769 + }, + { + "epoch": 0.2267669172932331, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2363, + "step": 8770 + }, + { + "epoch": 0.22682706766917293, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1421, + "step": 8771 + }, + { + "epoch": 0.22688721804511278, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1888, + "step": 8772 + }, + { + "epoch": 0.22694736842105262, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 8773 + }, + { + "epoch": 0.2270075187969925, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2235, + "step": 8774 + }, + { + "epoch": 0.22706766917293233, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.1839, + "step": 8775 + }, + { + "epoch": 0.22712781954887218, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 8776 + }, + { + "epoch": 0.22718796992481202, + "grad_norm": 0.4453125, + "learning_rate": 3e-05, + "loss": 1.2236, + "step": 8777 + }, + { + "epoch": 0.2272481203007519, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 8778 + }, + { + "epoch": 0.22730827067669174, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 8779 + }, + { + "epoch": 0.22736842105263158, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2363, + "step": 8780 + }, + { + "epoch": 0.22742857142857142, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 8781 + }, + { + "epoch": 0.22748872180451127, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 8782 + }, + { + "epoch": 0.22754887218045114, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1721, + "step": 8783 + }, + { + "epoch": 0.22760902255639098, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2256, + "step": 8784 + }, + { + "epoch": 0.22766917293233083, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2606, + "step": 8785 + }, + { + "epoch": 0.22772932330827067, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2131, + "step": 8786 + }, + { + "epoch": 0.22778947368421051, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 8787 + }, + { + "epoch": 0.22784962406015039, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2263, + "step": 8788 + }, + { + "epoch": 0.22790977443609023, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2131, + "step": 8789 + }, + { + "epoch": 0.22796992481203007, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 8790 + }, + { + "epoch": 0.22803007518796992, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 8791 + }, + { + "epoch": 0.2280902255639098, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.3052, + "step": 8792 + }, + { + "epoch": 0.22815037593984963, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2644, + "step": 8793 + }, + { + "epoch": 0.22821052631578947, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 8794 + }, + { + "epoch": 0.22827067669172932, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 8795 + }, + { + "epoch": 0.22833082706766916, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 8796 + }, + { + "epoch": 0.22839097744360903, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.21, + "step": 8797 + }, + { + "epoch": 0.22845112781954888, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 8798 + }, + { + "epoch": 0.22851127819548872, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2228, + "step": 8799 + }, + { + "epoch": 0.22857142857142856, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 8800 + }, + { + "epoch": 0.2286315789473684, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2658, + "step": 8801 + }, + { + "epoch": 0.22869172932330828, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2387, + "step": 8802 + }, + { + "epoch": 0.22875187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 8803 + }, + { + "epoch": 0.22881203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2283, + "step": 8804 + }, + { + "epoch": 0.2288721804511278, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2057, + "step": 8805 + }, + { + "epoch": 0.22893233082706768, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 8806 + }, + { + "epoch": 0.22899248120300753, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3098, + "step": 8807 + }, + { + "epoch": 0.22905263157894737, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2, + "step": 8808 + }, + { + "epoch": 0.2291127819548872, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 8809 + }, + { + "epoch": 0.22917293233082706, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 8810 + }, + { + "epoch": 0.22923308270676693, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.214, + "step": 8811 + }, + { + "epoch": 0.22929323308270677, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2259, + "step": 8812 + }, + { + "epoch": 0.22935338345864661, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2187, + "step": 8813 + }, + { + "epoch": 0.22941353383458646, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.148, + "step": 8814 + }, + { + "epoch": 0.2294736842105263, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2615, + "step": 8815 + }, + { + "epoch": 0.22953383458646617, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2296, + "step": 8816 + }, + { + "epoch": 0.22959398496240602, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.5297, + "step": 8817 + }, + { + "epoch": 0.22965413533834586, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2766, + "step": 8818 + }, + { + "epoch": 0.2297142857142857, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2927, + "step": 8819 + }, + { + "epoch": 0.22977443609022558, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2334, + "step": 8820 + }, + { + "epoch": 0.22983458646616542, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 8821 + }, + { + "epoch": 0.22989473684210526, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 8822 + }, + { + "epoch": 0.2299548872180451, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 8823 + }, + { + "epoch": 0.23001503759398495, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2142, + "step": 8824 + }, + { + "epoch": 0.23007518796992482, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 8825 + }, + { + "epoch": 0.23013533834586467, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2614, + "step": 8826 + }, + { + "epoch": 0.2301954887218045, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 8827 + }, + { + "epoch": 0.23025563909774435, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 8828 + }, + { + "epoch": 0.23031578947368422, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 8829 + }, + { + "epoch": 0.23037593984962407, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1859, + "step": 8830 + }, + { + "epoch": 0.2304360902255639, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 8831 + }, + { + "epoch": 0.23049624060150375, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1467, + "step": 8832 + }, + { + "epoch": 0.2305563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2651, + "step": 8833 + }, + { + "epoch": 0.23061654135338347, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2316, + "step": 8834 + }, + { + "epoch": 0.2306766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 8835 + }, + { + "epoch": 0.23073684210526316, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 8836 + }, + { + "epoch": 0.230796992481203, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1859, + "step": 8837 + }, + { + "epoch": 0.23085714285714284, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.3035, + "step": 8838 + }, + { + "epoch": 0.23091729323308272, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 8839 + }, + { + "epoch": 0.23097744360902256, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1719, + "step": 8840 + }, + { + "epoch": 0.2310375939849624, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.243, + "step": 8841 + }, + { + "epoch": 0.23109774436090225, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 8842 + }, + { + "epoch": 0.23115789473684212, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.192, + "step": 8843 + }, + { + "epoch": 0.23121804511278196, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 8844 + }, + { + "epoch": 0.2312781954887218, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.2953, + "step": 8845 + }, + { + "epoch": 0.23133834586466165, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1702, + "step": 8846 + }, + { + "epoch": 0.2313984962406015, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2315, + "step": 8847 + }, + { + "epoch": 0.23145864661654136, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.228, + "step": 8848 + }, + { + "epoch": 0.2315187969924812, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 8849 + }, + { + "epoch": 0.23157894736842105, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 8850 + }, + { + "epoch": 0.2316390977443609, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 8851 + }, + { + "epoch": 0.23169924812030074, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 8852 + }, + { + "epoch": 0.2317593984962406, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2528, + "step": 8853 + }, + { + "epoch": 0.23181954887218045, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 8854 + }, + { + "epoch": 0.2318796992481203, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2242, + "step": 8855 + }, + { + "epoch": 0.23193984962406014, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 8856 + }, + { + "epoch": 0.232, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3019, + "step": 8857 + }, + { + "epoch": 0.23206015037593986, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2514, + "step": 8858 + }, + { + "epoch": 0.2321203007518797, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2417, + "step": 8859 + }, + { + "epoch": 0.23218045112781954, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2352, + "step": 8860 + }, + { + "epoch": 0.2322406015037594, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 8861 + }, + { + "epoch": 0.23230075187969926, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 8862 + }, + { + "epoch": 0.2323609022556391, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2707, + "step": 8863 + }, + { + "epoch": 0.23242105263157894, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 8864 + }, + { + "epoch": 0.2324812030075188, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3455, + "step": 8865 + }, + { + "epoch": 0.23254135338345866, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2251, + "step": 8866 + }, + { + "epoch": 0.2326015037593985, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.3753, + "step": 8867 + }, + { + "epoch": 0.23266165413533835, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2011, + "step": 8868 + }, + { + "epoch": 0.2327218045112782, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 8869 + }, + { + "epoch": 0.23278195488721803, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.23, + "step": 8870 + }, + { + "epoch": 0.2328421052631579, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 8871 + }, + { + "epoch": 0.23290225563909775, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 8872 + }, + { + "epoch": 0.2329624060150376, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2553, + "step": 8873 + }, + { + "epoch": 0.23302255639097744, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2124, + "step": 8874 + }, + { + "epoch": 0.23308270676691728, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2335, + "step": 8875 + }, + { + "epoch": 0.23314285714285715, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 8876 + }, + { + "epoch": 0.233203007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2846, + "step": 8877 + }, + { + "epoch": 0.23326315789473684, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3364, + "step": 8878 + }, + { + "epoch": 0.23332330827067668, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.3156, + "step": 8879 + }, + { + "epoch": 0.23338345864661655, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2744, + "step": 8880 + }, + { + "epoch": 0.2334436090225564, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3938, + "step": 8881 + }, + { + "epoch": 0.23350375939849624, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2395, + "step": 8882 + }, + { + "epoch": 0.23356390977443608, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 8883 + }, + { + "epoch": 0.23362406015037593, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2197, + "step": 8884 + }, + { + "epoch": 0.2336842105263158, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3027, + "step": 8885 + }, + { + "epoch": 0.23374436090225564, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2732, + "step": 8886 + }, + { + "epoch": 0.2338045112781955, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.289, + "step": 8887 + }, + { + "epoch": 0.23386466165413533, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 8888 + }, + { + "epoch": 0.23392481203007517, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 8889 + }, + { + "epoch": 0.23398496240601505, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3874, + "step": 8890 + }, + { + "epoch": 0.2340451127819549, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2691, + "step": 8891 + }, + { + "epoch": 0.23410526315789473, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2489, + "step": 8892 + }, + { + "epoch": 0.23416541353383458, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 8893 + }, + { + "epoch": 0.23422556390977445, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2754, + "step": 8894 + }, + { + "epoch": 0.2342857142857143, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2109, + "step": 8895 + }, + { + "epoch": 0.23434586466165414, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 8896 + }, + { + "epoch": 0.23440601503759398, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2245, + "step": 8897 + }, + { + "epoch": 0.23446616541353382, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 8898 + }, + { + "epoch": 0.2345263157894737, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2438, + "step": 8899 + }, + { + "epoch": 0.23458646616541354, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 8900 + }, + { + "epoch": 0.23464661654135338, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2157, + "step": 8901 + }, + { + "epoch": 0.23470676691729322, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.207, + "step": 8902 + }, + { + "epoch": 0.2347669172932331, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 8903 + }, + { + "epoch": 0.23482706766917294, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 8904 + }, + { + "epoch": 0.23488721804511278, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2449, + "step": 8905 + }, + { + "epoch": 0.23494736842105263, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2042, + "step": 8906 + }, + { + "epoch": 0.23500751879699247, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1836, + "step": 8907 + }, + { + "epoch": 0.23506766917293234, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3, + "step": 8908 + }, + { + "epoch": 0.23512781954887219, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 8909 + }, + { + "epoch": 0.23518796992481203, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2278, + "step": 8910 + }, + { + "epoch": 0.23524812030075187, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 8911 + }, + { + "epoch": 0.23530827067669172, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2536, + "step": 8912 + }, + { + "epoch": 0.2353684210526316, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2436, + "step": 8913 + }, + { + "epoch": 0.23542857142857143, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2136, + "step": 8914 + }, + { + "epoch": 0.23548872180451128, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2808, + "step": 8915 + }, + { + "epoch": 0.23554887218045112, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 8916 + }, + { + "epoch": 0.235609022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 8917 + }, + { + "epoch": 0.23566917293233083, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.2241, + "step": 8918 + }, + { + "epoch": 0.23572932330827068, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 8919 + }, + { + "epoch": 0.23578947368421052, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2083, + "step": 8920 + }, + { + "epoch": 0.23584962406015036, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.2362, + "step": 8921 + }, + { + "epoch": 0.23590977443609024, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1936, + "step": 8922 + }, + { + "epoch": 0.23596992481203008, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.21, + "step": 8923 + }, + { + "epoch": 0.23603007518796992, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 8924 + }, + { + "epoch": 0.23609022556390977, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.26, + "step": 8925 + }, + { + "epoch": 0.2361503759398496, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 8926 + }, + { + "epoch": 0.23621052631578948, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2733, + "step": 8927 + }, + { + "epoch": 0.23627067669172933, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 8928 + }, + { + "epoch": 0.23633082706766917, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3039, + "step": 8929 + }, + { + "epoch": 0.236390977443609, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 8930 + }, + { + "epoch": 0.23645112781954888, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 8931 + }, + { + "epoch": 0.23651127819548873, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2056, + "step": 8932 + }, + { + "epoch": 0.23657142857142857, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2806, + "step": 8933 + }, + { + "epoch": 0.23663157894736842, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 8934 + }, + { + "epoch": 0.23669172932330826, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2135, + "step": 8935 + }, + { + "epoch": 0.23675187969924813, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 8936 + }, + { + "epoch": 0.23681203007518797, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2299, + "step": 8937 + }, + { + "epoch": 0.23687218045112782, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2462, + "step": 8938 + }, + { + "epoch": 0.23693233082706766, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2279, + "step": 8939 + }, + { + "epoch": 0.23699248120300753, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.265, + "step": 8940 + }, + { + "epoch": 0.23705263157894738, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 8941 + }, + { + "epoch": 0.23711278195488722, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 8942 + }, + { + "epoch": 0.23717293233082706, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 8943 + }, + { + "epoch": 0.2372330827067669, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2152, + "step": 8944 + }, + { + "epoch": 0.23729323308270678, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 8945 + }, + { + "epoch": 0.23735338345864662, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.197, + "step": 8946 + }, + { + "epoch": 0.23741353383458647, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 8947 + }, + { + "epoch": 0.2374736842105263, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2521, + "step": 8948 + }, + { + "epoch": 0.23753383458646615, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2896, + "step": 8949 + }, + { + "epoch": 0.23759398496240602, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2619, + "step": 8950 + }, + { + "epoch": 0.23765413533834587, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 8951 + }, + { + "epoch": 0.2377142857142857, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 8952 + }, + { + "epoch": 0.23777443609022555, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 8953 + }, + { + "epoch": 0.23783458646616543, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.3061, + "step": 8954 + }, + { + "epoch": 0.23789473684210527, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 8955 + }, + { + "epoch": 0.2379548872180451, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2641, + "step": 8956 + }, + { + "epoch": 0.23801503759398496, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 8957 + }, + { + "epoch": 0.2380751879699248, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.3095, + "step": 8958 + }, + { + "epoch": 0.23813533834586467, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 8959 + }, + { + "epoch": 0.23819548872180452, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 8960 + }, + { + "epoch": 0.23825563909774436, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.239, + "step": 8961 + }, + { + "epoch": 0.2383157894736842, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 8962 + }, + { + "epoch": 0.23837593984962405, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 8963 + }, + { + "epoch": 0.23843609022556392, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 8964 + }, + { + "epoch": 0.23849624060150376, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2348, + "step": 8965 + }, + { + "epoch": 0.2385563909774436, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 8966 + }, + { + "epoch": 0.23861654135338345, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 8967 + }, + { + "epoch": 0.23867669172932332, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 8968 + }, + { + "epoch": 0.23873684210526316, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3618, + "step": 8969 + }, + { + "epoch": 0.238796992481203, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2748, + "step": 8970 + }, + { + "epoch": 0.23885714285714285, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 8971 + }, + { + "epoch": 0.2389172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.326, + "step": 8972 + }, + { + "epoch": 0.23897744360902257, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.3187, + "step": 8973 + }, + { + "epoch": 0.2390375939849624, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 8974 + }, + { + "epoch": 0.23909774436090225, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2006, + "step": 8975 + }, + { + "epoch": 0.2391578947368421, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2463, + "step": 8976 + }, + { + "epoch": 0.23921804511278197, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2316, + "step": 8977 + }, + { + "epoch": 0.2392781954887218, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2966, + "step": 8978 + }, + { + "epoch": 0.23933834586466166, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 8979 + }, + { + "epoch": 0.2393984962406015, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2444, + "step": 8980 + }, + { + "epoch": 0.23945864661654134, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2685, + "step": 8981 + }, + { + "epoch": 0.23951879699248121, + "grad_norm": 2.03125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 8982 + }, + { + "epoch": 0.23957894736842106, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1986, + "step": 8983 + }, + { + "epoch": 0.2396390977443609, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 8984 + }, + { + "epoch": 0.23969924812030075, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2141, + "step": 8985 + }, + { + "epoch": 0.2397593984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 8986 + }, + { + "epoch": 0.23981954887218046, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 8987 + }, + { + "epoch": 0.2398796992481203, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.156, + "step": 8988 + }, + { + "epoch": 0.23993984962406015, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2557, + "step": 8989 + }, + { + "epoch": 0.24, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2217, + "step": 8990 + }, + { + "epoch": 0.24006015037593986, + "grad_norm": 4.40625, + "learning_rate": 3e-05, + "loss": 1.4093, + "step": 8991 + }, + { + "epoch": 0.2401203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 8992 + }, + { + "epoch": 0.24018045112781955, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 8993 + }, + { + "epoch": 0.2402406015037594, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 8994 + }, + { + "epoch": 0.24030075187969924, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 8995 + }, + { + "epoch": 0.2403609022556391, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2313, + "step": 8996 + }, + { + "epoch": 0.24042105263157895, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3767, + "step": 8997 + }, + { + "epoch": 0.2404812030075188, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 8998 + }, + { + "epoch": 0.24054135338345864, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 8999 + }, + { + "epoch": 0.24060150375939848, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 9000 + }, + { + "epoch": 0.24066165413533835, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 9001 + }, + { + "epoch": 0.2407218045112782, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2024, + "step": 9002 + }, + { + "epoch": 0.24078195488721804, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 9003 + }, + { + "epoch": 0.24084210526315789, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2055, + "step": 9004 + }, + { + "epoch": 0.24090225563909776, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1708, + "step": 9005 + }, + { + "epoch": 0.2409624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 9006 + }, + { + "epoch": 0.24102255639097744, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.238, + "step": 9007 + }, + { + "epoch": 0.2410827067669173, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 9008 + }, + { + "epoch": 0.24114285714285713, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.261, + "step": 9009 + }, + { + "epoch": 0.241203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2112, + "step": 9010 + }, + { + "epoch": 0.24126315789473685, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 9011 + }, + { + "epoch": 0.2413233082706767, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2993, + "step": 9012 + }, + { + "epoch": 0.24138345864661653, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.402, + "step": 9013 + }, + { + "epoch": 0.2414436090225564, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 9014 + }, + { + "epoch": 0.24150375939849625, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2113, + "step": 9015 + }, + { + "epoch": 0.2415639097744361, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2141, + "step": 9016 + }, + { + "epoch": 0.24162406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2056, + "step": 9017 + }, + { + "epoch": 0.24168421052631578, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.3124, + "step": 9018 + }, + { + "epoch": 0.24174436090225565, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 9019 + }, + { + "epoch": 0.2418045112781955, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1638, + "step": 9020 + }, + { + "epoch": 0.24186466165413534, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 9021 + }, + { + "epoch": 0.24192481203007518, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.2585, + "step": 9022 + }, + { + "epoch": 0.24198496240601503, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2946, + "step": 9023 + }, + { + "epoch": 0.2420451127819549, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3665, + "step": 9024 + }, + { + "epoch": 0.24210526315789474, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2539, + "step": 9025 + }, + { + "epoch": 0.24216541353383458, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 9026 + }, + { + "epoch": 0.24222556390977443, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2512, + "step": 9027 + }, + { + "epoch": 0.2422857142857143, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2534, + "step": 9028 + }, + { + "epoch": 0.24234586466165414, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2122, + "step": 9029 + }, + { + "epoch": 0.24240601503759399, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 9030 + }, + { + "epoch": 0.24246616541353383, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2101, + "step": 9031 + }, + { + "epoch": 0.24252631578947367, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2556, + "step": 9032 + }, + { + "epoch": 0.24258646616541354, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 9033 + }, + { + "epoch": 0.2426466165413534, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 9034 + }, + { + "epoch": 0.24270676691729323, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2309, + "step": 9035 + }, + { + "epoch": 0.24276691729323308, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2333, + "step": 9036 + }, + { + "epoch": 0.24282706766917292, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2632, + "step": 9037 + }, + { + "epoch": 0.2428872180451128, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 9038 + }, + { + "epoch": 0.24294736842105263, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 9039 + }, + { + "epoch": 0.24300751879699248, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 9040 + }, + { + "epoch": 0.24306766917293232, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.238, + "step": 9041 + }, + { + "epoch": 0.2431278195488722, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2521, + "step": 9042 + }, + { + "epoch": 0.24318796992481204, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.284, + "step": 9043 + }, + { + "epoch": 0.24324812030075188, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1994, + "step": 9044 + }, + { + "epoch": 0.24330827067669172, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 9045 + }, + { + "epoch": 0.24336842105263157, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1841, + "step": 9046 + }, + { + "epoch": 0.24342857142857144, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 9047 + }, + { + "epoch": 0.24348872180451128, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2739, + "step": 9048 + }, + { + "epoch": 0.24354887218045113, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 9049 + }, + { + "epoch": 0.24360902255639097, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 9050 + }, + { + "epoch": 0.24366917293233084, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3028, + "step": 9051 + }, + { + "epoch": 0.24372932330827068, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2343, + "step": 9052 + }, + { + "epoch": 0.24378947368421053, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 9053 + }, + { + "epoch": 0.24384962406015037, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2592, + "step": 9054 + }, + { + "epoch": 0.24390977443609022, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2393, + "step": 9055 + }, + { + "epoch": 0.2439699248120301, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 9056 + }, + { + "epoch": 0.24403007518796993, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 9057 + }, + { + "epoch": 0.24409022556390977, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 9058 + }, + { + "epoch": 0.24415037593984962, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.4447, + "step": 9059 + }, + { + "epoch": 0.24421052631578946, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 9060 + }, + { + "epoch": 0.24427067669172933, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2077, + "step": 9061 + }, + { + "epoch": 0.24433082706766918, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 9062 + }, + { + "epoch": 0.24439097744360902, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2602, + "step": 9063 + }, + { + "epoch": 0.24445112781954886, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 9064 + }, + { + "epoch": 0.24451127819548873, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3053, + "step": 9065 + }, + { + "epoch": 0.24457142857142858, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1641, + "step": 9066 + }, + { + "epoch": 0.24463157894736842, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2077, + "step": 9067 + }, + { + "epoch": 0.24469172932330827, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2608, + "step": 9068 + }, + { + "epoch": 0.2447518796992481, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 9069 + }, + { + "epoch": 0.24481203007518798, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.2913, + "step": 9070 + }, + { + "epoch": 0.24487218045112782, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 9071 + }, + { + "epoch": 0.24493233082706767, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2763, + "step": 9072 + }, + { + "epoch": 0.2449924812030075, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1791, + "step": 9073 + }, + { + "epoch": 0.24505263157894736, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 9074 + }, + { + "epoch": 0.24511278195488723, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 9075 + }, + { + "epoch": 0.24517293233082707, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 9076 + }, + { + "epoch": 0.24523308270676691, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.204, + "step": 9077 + }, + { + "epoch": 0.24529323308270676, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3158, + "step": 9078 + }, + { + "epoch": 0.24535338345864663, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2206, + "step": 9079 + }, + { + "epoch": 0.24541353383458647, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 9080 + }, + { + "epoch": 0.24547368421052632, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1465, + "step": 9081 + }, + { + "epoch": 0.24553383458646616, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 9082 + }, + { + "epoch": 0.245593984962406, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.304, + "step": 9083 + }, + { + "epoch": 0.24565413533834587, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 9084 + }, + { + "epoch": 0.24571428571428572, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 9085 + }, + { + "epoch": 0.24577443609022556, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.25, + "step": 9086 + }, + { + "epoch": 0.2458345864661654, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.28, + "step": 9087 + }, + { + "epoch": 0.24589473684210525, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2169, + "step": 9088 + }, + { + "epoch": 0.24595488721804512, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2642, + "step": 9089 + }, + { + "epoch": 0.24601503759398496, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.3749, + "step": 9090 + }, + { + "epoch": 0.2460751879699248, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2626, + "step": 9091 + }, + { + "epoch": 0.24613533834586465, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2233, + "step": 9092 + }, + { + "epoch": 0.24619548872180452, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 9093 + }, + { + "epoch": 0.24625563909774437, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2263, + "step": 9094 + }, + { + "epoch": 0.2463157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 9095 + }, + { + "epoch": 0.24637593984962405, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 9096 + }, + { + "epoch": 0.2464360902255639, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.245, + "step": 9097 + }, + { + "epoch": 0.24649624060150377, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.256, + "step": 9098 + }, + { + "epoch": 0.2465563909774436, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 9099 + }, + { + "epoch": 0.24661654135338346, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3016, + "step": 9100 + }, + { + "epoch": 0.2466766917293233, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2225, + "step": 9101 + }, + { + "epoch": 0.24673684210526317, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 9102 + }, + { + "epoch": 0.24679699248120301, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 9103 + }, + { + "epoch": 0.24685714285714286, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 9104 + }, + { + "epoch": 0.2469172932330827, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 9105 + }, + { + "epoch": 0.24697744360902255, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3096, + "step": 9106 + }, + { + "epoch": 0.24703759398496242, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2212, + "step": 9107 + }, + { + "epoch": 0.24709774436090226, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2526, + "step": 9108 + }, + { + "epoch": 0.2471578947368421, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 9109 + }, + { + "epoch": 0.24721804511278195, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 9110 + }, + { + "epoch": 0.2472781954887218, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 9111 + }, + { + "epoch": 0.24733834586466166, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 9112 + }, + { + "epoch": 0.2473984962406015, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.3427, + "step": 9113 + }, + { + "epoch": 0.24745864661654135, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 9114 + }, + { + "epoch": 0.2475187969924812, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 9115 + }, + { + "epoch": 0.24757894736842107, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 9116 + }, + { + "epoch": 0.2476390977443609, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 9117 + }, + { + "epoch": 0.24769924812030075, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 9118 + }, + { + "epoch": 0.2477593984962406, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 9119 + }, + { + "epoch": 0.24781954887218044, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 9120 + }, + { + "epoch": 0.2478796992481203, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 9121 + }, + { + "epoch": 0.24793984962406015, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2193, + "step": 9122 + }, + { + "epoch": 0.248, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 9123 + }, + { + "epoch": 0.24806015037593984, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 9124 + }, + { + "epoch": 0.24812030075187969, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.199, + "step": 9125 + }, + { + "epoch": 0.24818045112781956, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2513, + "step": 9126 + }, + { + "epoch": 0.2482406015037594, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2069, + "step": 9127 + }, + { + "epoch": 0.24830075187969924, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 9128 + }, + { + "epoch": 0.2483609022556391, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 9129 + }, + { + "epoch": 0.24842105263157896, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.192, + "step": 9130 + }, + { + "epoch": 0.2484812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 9131 + }, + { + "epoch": 0.24854135338345865, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2091, + "step": 9132 + }, + { + "epoch": 0.2486015037593985, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2867, + "step": 9133 + }, + { + "epoch": 0.24866165413533833, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 9134 + }, + { + "epoch": 0.2487218045112782, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2407, + "step": 9135 + }, + { + "epoch": 0.24878195488721805, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 9136 + }, + { + "epoch": 0.2488421052631579, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2296, + "step": 9137 + }, + { + "epoch": 0.24890225563909774, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 9138 + }, + { + "epoch": 0.2489624060150376, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 9139 + }, + { + "epoch": 0.24902255639097745, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3635, + "step": 9140 + }, + { + "epoch": 0.2490827067669173, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 9141 + }, + { + "epoch": 0.24914285714285714, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.3463, + "step": 9142 + }, + { + "epoch": 0.24920300751879698, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.21, + "step": 9143 + }, + { + "epoch": 0.24926315789473685, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1912, + "step": 9144 + }, + { + "epoch": 0.2493233082706767, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 9145 + }, + { + "epoch": 0.24938345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2085, + "step": 9146 + }, + { + "epoch": 0.24944360902255638, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1696, + "step": 9147 + }, + { + "epoch": 0.24950375939849623, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1915, + "step": 9148 + }, + { + "epoch": 0.2495639097744361, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2288, + "step": 9149 + }, + { + "epoch": 0.24962406015037594, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 9150 + }, + { + "epoch": 0.2496842105263158, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2335, + "step": 9151 + }, + { + "epoch": 0.24974436090225563, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3101, + "step": 9152 + }, + { + "epoch": 0.2498045112781955, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2044, + "step": 9153 + }, + { + "epoch": 0.24986466165413534, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2522, + "step": 9154 + }, + { + "epoch": 0.2499248120300752, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2162, + "step": 9155 + }, + { + "epoch": 0.24998496240601503, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 9156 + }, + { + "epoch": 0.2500451127819549, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 9157 + }, + { + "epoch": 0.2501052631578947, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 9158 + }, + { + "epoch": 0.2501654135338346, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.262, + "step": 9159 + }, + { + "epoch": 0.25022556390977446, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 9160 + }, + { + "epoch": 0.2502857142857143, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 9161 + }, + { + "epoch": 0.25034586466165415, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1808, + "step": 9162 + }, + { + "epoch": 0.25040601503759397, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 9163 + }, + { + "epoch": 0.25046616541353384, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 9164 + }, + { + "epoch": 0.2505263157894737, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2336, + "step": 9165 + }, + { + "epoch": 0.2505864661654135, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2833, + "step": 9166 + }, + { + "epoch": 0.2506466165413534, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2206, + "step": 9167 + }, + { + "epoch": 0.2507067669172932, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1334, + "step": 9168 + }, + { + "epoch": 0.2507669172932331, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 9169 + }, + { + "epoch": 0.25082706766917295, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3038, + "step": 9170 + }, + { + "epoch": 0.25088721804511277, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2536, + "step": 9171 + }, + { + "epoch": 0.25094736842105264, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 9172 + }, + { + "epoch": 0.25100751879699246, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 9173 + }, + { + "epoch": 0.25106766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2122, + "step": 9174 + }, + { + "epoch": 0.2511278195488722, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 9175 + }, + { + "epoch": 0.251187969924812, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 9176 + }, + { + "epoch": 0.2512481203007519, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3158, + "step": 9177 + }, + { + "epoch": 0.2513082706766917, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 9178 + }, + { + "epoch": 0.2513684210526316, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2545, + "step": 9179 + }, + { + "epoch": 0.25142857142857145, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2661, + "step": 9180 + }, + { + "epoch": 0.25148872180451126, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2575, + "step": 9181 + }, + { + "epoch": 0.25154887218045113, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 9182 + }, + { + "epoch": 0.251609022556391, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2526, + "step": 9183 + }, + { + "epoch": 0.2516691729323308, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 9184 + }, + { + "epoch": 0.2517293233082707, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2382, + "step": 9185 + }, + { + "epoch": 0.2517894736842105, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1959, + "step": 9186 + }, + { + "epoch": 0.2518496240601504, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 9187 + }, + { + "epoch": 0.25190977443609025, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 9188 + }, + { + "epoch": 0.25196992481203007, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2434, + "step": 9189 + }, + { + "epoch": 0.25203007518796994, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 9190 + }, + { + "epoch": 0.25209022556390975, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 9191 + }, + { + "epoch": 0.2521503759398496, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2166, + "step": 9192 + }, + { + "epoch": 0.2522105263157895, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2588, + "step": 9193 + }, + { + "epoch": 0.2522706766917293, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 9194 + }, + { + "epoch": 0.2523308270676692, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 9195 + }, + { + "epoch": 0.252390977443609, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2405, + "step": 9196 + }, + { + "epoch": 0.25245112781954887, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2194, + "step": 9197 + }, + { + "epoch": 0.25251127819548874, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 9198 + }, + { + "epoch": 0.25257142857142856, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 9199 + }, + { + "epoch": 0.25263157894736843, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2722, + "step": 9200 + }, + { + "epoch": 0.25269172932330825, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 9201 + }, + { + "epoch": 0.2527518796992481, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 9202 + }, + { + "epoch": 0.252812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 9203 + }, + { + "epoch": 0.2528721804511278, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 9204 + }, + { + "epoch": 0.2529323308270677, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 9205 + }, + { + "epoch": 0.25299248120300755, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1943, + "step": 9206 + }, + { + "epoch": 0.25305263157894736, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 9207 + }, + { + "epoch": 0.25311278195488723, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 9208 + }, + { + "epoch": 0.25317293233082705, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2603, + "step": 9209 + }, + { + "epoch": 0.2532330827067669, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 9210 + }, + { + "epoch": 0.2532932330827068, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2523, + "step": 9211 + }, + { + "epoch": 0.2533533834586466, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2862, + "step": 9212 + }, + { + "epoch": 0.2534135338345865, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.202, + "step": 9213 + }, + { + "epoch": 0.2534736842105263, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.3309, + "step": 9214 + }, + { + "epoch": 0.25353383458646617, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 9215 + }, + { + "epoch": 0.25359398496240604, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 9216 + }, + { + "epoch": 0.25365413533834585, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2961, + "step": 9217 + }, + { + "epoch": 0.2537142857142857, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2816, + "step": 9218 + }, + { + "epoch": 0.25377443609022554, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 9219 + }, + { + "epoch": 0.2538345864661654, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 9220 + }, + { + "epoch": 0.2538947368421053, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 9221 + }, + { + "epoch": 0.2539548872180451, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 9222 + }, + { + "epoch": 0.25401503759398497, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.186, + "step": 9223 + }, + { + "epoch": 0.2540751879699248, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2805, + "step": 9224 + }, + { + "epoch": 0.25413533834586466, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2008, + "step": 9225 + }, + { + "epoch": 0.25419548872180453, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2514, + "step": 9226 + }, + { + "epoch": 0.25425563909774435, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2227, + "step": 9227 + }, + { + "epoch": 0.2543157894736842, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.304, + "step": 9228 + }, + { + "epoch": 0.25437593984962403, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1758, + "step": 9229 + }, + { + "epoch": 0.2544360902255639, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 9230 + }, + { + "epoch": 0.2544962406015038, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.2032, + "step": 9231 + }, + { + "epoch": 0.2545563909774436, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2808, + "step": 9232 + }, + { + "epoch": 0.25461654135338346, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2635, + "step": 9233 + }, + { + "epoch": 0.25467669172932333, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 9234 + }, + { + "epoch": 0.25473684210526315, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2152, + "step": 9235 + }, + { + "epoch": 0.254796992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 9236 + }, + { + "epoch": 0.25485714285714284, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2674, + "step": 9237 + }, + { + "epoch": 0.2549172932330827, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2694, + "step": 9238 + }, + { + "epoch": 0.2549774436090226, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 9239 + }, + { + "epoch": 0.2550375939849624, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2194, + "step": 9240 + }, + { + "epoch": 0.25509774436090227, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 9241 + }, + { + "epoch": 0.2551578947368421, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 9242 + }, + { + "epoch": 0.25521804511278195, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 9243 + }, + { + "epoch": 0.2552781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2008, + "step": 9244 + }, + { + "epoch": 0.25533834586466164, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1597, + "step": 9245 + }, + { + "epoch": 0.2553984962406015, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 9246 + }, + { + "epoch": 0.25545864661654133, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2573, + "step": 9247 + }, + { + "epoch": 0.2555187969924812, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 9248 + }, + { + "epoch": 0.25557894736842107, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2721, + "step": 9249 + }, + { + "epoch": 0.2556390977443609, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 9250 + }, + { + "epoch": 0.25569924812030076, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1786, + "step": 9251 + }, + { + "epoch": 0.2557593984962406, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2296, + "step": 9252 + }, + { + "epoch": 0.25581954887218045, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1332, + "step": 9253 + }, + { + "epoch": 0.2558796992481203, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2308, + "step": 9254 + }, + { + "epoch": 0.25593984962406013, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2652, + "step": 9255 + }, + { + "epoch": 0.256, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 9256 + }, + { + "epoch": 0.2560601503759399, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.3351, + "step": 9257 + }, + { + "epoch": 0.2561203007518797, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 9258 + }, + { + "epoch": 0.25618045112781956, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2286, + "step": 9259 + }, + { + "epoch": 0.2562406015037594, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2722, + "step": 9260 + }, + { + "epoch": 0.25630075187969925, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 9261 + }, + { + "epoch": 0.2563609022556391, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2596, + "step": 9262 + }, + { + "epoch": 0.25642105263157894, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3109, + "step": 9263 + }, + { + "epoch": 0.2564812030075188, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2021, + "step": 9264 + }, + { + "epoch": 0.2565413533834586, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2492, + "step": 9265 + }, + { + "epoch": 0.2566015037593985, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 9266 + }, + { + "epoch": 0.25666165413533837, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 9267 + }, + { + "epoch": 0.2567218045112782, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 9268 + }, + { + "epoch": 0.25678195488721806, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3454, + "step": 9269 + }, + { + "epoch": 0.25684210526315787, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 9270 + }, + { + "epoch": 0.25690225563909774, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 9271 + }, + { + "epoch": 0.2569624060150376, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2334, + "step": 9272 + }, + { + "epoch": 0.25702255639097743, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 9273 + }, + { + "epoch": 0.2570827067669173, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 9274 + }, + { + "epoch": 0.2571428571428571, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2652, + "step": 9275 + }, + { + "epoch": 0.257203007518797, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 9276 + }, + { + "epoch": 0.25726315789473686, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 9277 + }, + { + "epoch": 0.2573233082706767, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 9278 + }, + { + "epoch": 0.25738345864661655, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 9279 + }, + { + "epoch": 0.25744360902255636, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2492, + "step": 9280 + }, + { + "epoch": 0.25750375939849623, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.246, + "step": 9281 + }, + { + "epoch": 0.2575639097744361, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 9282 + }, + { + "epoch": 0.2576240601503759, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2437, + "step": 9283 + }, + { + "epoch": 0.2576842105263158, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1955, + "step": 9284 + }, + { + "epoch": 0.25774436090225566, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 9285 + }, + { + "epoch": 0.2578045112781955, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2613, + "step": 9286 + }, + { + "epoch": 0.25786466165413535, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2485, + "step": 9287 + }, + { + "epoch": 0.25792481203007517, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 9288 + }, + { + "epoch": 0.25798496240601504, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2376, + "step": 9289 + }, + { + "epoch": 0.2580451127819549, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2664, + "step": 9290 + }, + { + "epoch": 0.2581052631578947, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2339, + "step": 9291 + }, + { + "epoch": 0.2581654135338346, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 9292 + }, + { + "epoch": 0.2582255639097744, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 9293 + }, + { + "epoch": 0.2582857142857143, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1872, + "step": 9294 + }, + { + "epoch": 0.25834586466165416, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2451, + "step": 9295 + }, + { + "epoch": 0.258406015037594, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2475, + "step": 9296 + }, + { + "epoch": 0.25846616541353384, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 9297 + }, + { + "epoch": 0.25852631578947366, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1677, + "step": 9298 + }, + { + "epoch": 0.25858646616541353, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.215, + "step": 9299 + }, + { + "epoch": 0.2586466165413534, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 9300 + }, + { + "epoch": 0.2587067669172932, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1834, + "step": 9301 + }, + { + "epoch": 0.2587669172932331, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 9302 + }, + { + "epoch": 0.2588270676691729, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 9303 + }, + { + "epoch": 0.2588872180451128, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2528, + "step": 9304 + }, + { + "epoch": 0.25894736842105265, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 9305 + }, + { + "epoch": 0.25900751879699246, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 9306 + }, + { + "epoch": 0.25906766917293234, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2166, + "step": 9307 + }, + { + "epoch": 0.2591278195488722, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2017, + "step": 9308 + }, + { + "epoch": 0.259187969924812, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2872, + "step": 9309 + }, + { + "epoch": 0.2592481203007519, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2787, + "step": 9310 + }, + { + "epoch": 0.2593082706766917, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2734, + "step": 9311 + }, + { + "epoch": 0.2593684210526316, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 9312 + }, + { + "epoch": 0.25942857142857145, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1751, + "step": 9313 + }, + { + "epoch": 0.25948872180451127, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 9314 + }, + { + "epoch": 0.25954887218045114, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2412, + "step": 9315 + }, + { + "epoch": 0.25960902255639096, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 9316 + }, + { + "epoch": 0.2596691729323308, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2575, + "step": 9317 + }, + { + "epoch": 0.2597293233082707, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2769, + "step": 9318 + }, + { + "epoch": 0.2597894736842105, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 9319 + }, + { + "epoch": 0.2598496240601504, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 9320 + }, + { + "epoch": 0.2599097744360902, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2194, + "step": 9321 + }, + { + "epoch": 0.2599699248120301, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 9322 + }, + { + "epoch": 0.26003007518796994, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 9323 + }, + { + "epoch": 0.26009022556390976, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 9324 + }, + { + "epoch": 0.26015037593984963, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 9325 + }, + { + "epoch": 0.26021052631578945, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 9326 + }, + { + "epoch": 0.2602706766917293, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 9327 + }, + { + "epoch": 0.2603308270676692, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 9328 + }, + { + "epoch": 0.260390977443609, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2057, + "step": 9329 + }, + { + "epoch": 0.2604511278195489, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 9330 + }, + { + "epoch": 0.26051127819548875, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1903, + "step": 9331 + }, + { + "epoch": 0.26057142857142856, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 9332 + }, + { + "epoch": 0.26063157894736844, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2907, + "step": 9333 + }, + { + "epoch": 0.26069172932330825, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.3362, + "step": 9334 + }, + { + "epoch": 0.2607518796992481, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1882, + "step": 9335 + }, + { + "epoch": 0.260812030075188, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 9336 + }, + { + "epoch": 0.2608721804511278, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2154, + "step": 9337 + }, + { + "epoch": 0.2609323308270677, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 9338 + }, + { + "epoch": 0.2609924812030075, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 9339 + }, + { + "epoch": 0.26105263157894737, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2197, + "step": 9340 + }, + { + "epoch": 0.26111278195488724, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2871, + "step": 9341 + }, + { + "epoch": 0.26117293233082706, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1868, + "step": 9342 + }, + { + "epoch": 0.26123308270676693, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1758, + "step": 9343 + }, + { + "epoch": 0.26129323308270674, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2466, + "step": 9344 + }, + { + "epoch": 0.2613533834586466, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 9345 + }, + { + "epoch": 0.2614135338345865, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2093, + "step": 9346 + }, + { + "epoch": 0.2614736842105263, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1852, + "step": 9347 + }, + { + "epoch": 0.2615338345864662, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2386, + "step": 9348 + }, + { + "epoch": 0.261593984962406, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 9349 + }, + { + "epoch": 0.26165413533834586, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2442, + "step": 9350 + }, + { + "epoch": 0.26171428571428573, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2047, + "step": 9351 + }, + { + "epoch": 0.26177443609022555, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 9352 + }, + { + "epoch": 0.2618345864661654, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2312, + "step": 9353 + }, + { + "epoch": 0.26189473684210524, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3361, + "step": 9354 + }, + { + "epoch": 0.2619548872180451, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 9355 + }, + { + "epoch": 0.262015037593985, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2986, + "step": 9356 + }, + { + "epoch": 0.2620751879699248, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2644, + "step": 9357 + }, + { + "epoch": 0.26213533834586467, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1667, + "step": 9358 + }, + { + "epoch": 0.26219548872180454, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2635, + "step": 9359 + }, + { + "epoch": 0.26225563909774435, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2483, + "step": 9360 + }, + { + "epoch": 0.2623157894736842, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2496, + "step": 9361 + }, + { + "epoch": 0.26237593984962404, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2258, + "step": 9362 + }, + { + "epoch": 0.2624360902255639, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 9363 + }, + { + "epoch": 0.2624962406015038, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2397, + "step": 9364 + }, + { + "epoch": 0.2625563909774436, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2027, + "step": 9365 + }, + { + "epoch": 0.26261654135338347, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 9366 + }, + { + "epoch": 0.2626766917293233, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2541, + "step": 9367 + }, + { + "epoch": 0.26273684210526316, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2446, + "step": 9368 + }, + { + "epoch": 0.26279699248120303, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.27, + "step": 9369 + }, + { + "epoch": 0.26285714285714284, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 9370 + }, + { + "epoch": 0.2629172932330827, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 9371 + }, + { + "epoch": 0.26297744360902253, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.136, + "step": 9372 + }, + { + "epoch": 0.2630375939849624, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.227, + "step": 9373 + }, + { + "epoch": 0.2630977443609023, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2637, + "step": 9374 + }, + { + "epoch": 0.2631578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 9375 + }, + { + "epoch": 0.26321804511278196, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 9376 + }, + { + "epoch": 0.2632781954887218, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 9377 + }, + { + "epoch": 0.26333834586466165, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2554, + "step": 9378 + }, + { + "epoch": 0.2633984962406015, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1905, + "step": 9379 + }, + { + "epoch": 0.26345864661654134, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 9380 + }, + { + "epoch": 0.2635187969924812, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2778, + "step": 9381 + }, + { + "epoch": 0.2635789473684211, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3007, + "step": 9382 + }, + { + "epoch": 0.2636390977443609, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2333, + "step": 9383 + }, + { + "epoch": 0.26369924812030077, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 9384 + }, + { + "epoch": 0.2637593984962406, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 9385 + }, + { + "epoch": 0.26381954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 9386 + }, + { + "epoch": 0.2638796992481203, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 9387 + }, + { + "epoch": 0.26393984962406014, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2699, + "step": 9388 + }, + { + "epoch": 0.264, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1689, + "step": 9389 + }, + { + "epoch": 0.26406015037593983, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1667, + "step": 9390 + }, + { + "epoch": 0.2641203007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 9391 + }, + { + "epoch": 0.26418045112781957, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2305, + "step": 9392 + }, + { + "epoch": 0.2642406015037594, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 9393 + }, + { + "epoch": 0.26430075187969926, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 9394 + }, + { + "epoch": 0.2643609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 9395 + }, + { + "epoch": 0.26442105263157895, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.4064, + "step": 9396 + }, + { + "epoch": 0.2644812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 9397 + }, + { + "epoch": 0.26454135338345863, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 9398 + }, + { + "epoch": 0.2646015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 9399 + }, + { + "epoch": 0.2646616541353383, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 9400 + }, + { + "epoch": 0.2647218045112782, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1937, + "step": 9401 + }, + { + "epoch": 0.26478195488721806, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2164, + "step": 9402 + }, + { + "epoch": 0.2648421052631579, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2382, + "step": 9403 + }, + { + "epoch": 0.26490225563909775, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 9404 + }, + { + "epoch": 0.2649624060150376, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2223, + "step": 9405 + }, + { + "epoch": 0.26502255639097744, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2267, + "step": 9406 + }, + { + "epoch": 0.2650827067669173, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 9407 + }, + { + "epoch": 0.2651428571428571, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2463, + "step": 9408 + }, + { + "epoch": 0.265203007518797, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 9409 + }, + { + "epoch": 0.26526315789473687, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 9410 + }, + { + "epoch": 0.2653233082706767, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 9411 + }, + { + "epoch": 0.26538345864661655, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2575, + "step": 9412 + }, + { + "epoch": 0.26544360902255637, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 9413 + }, + { + "epoch": 0.26550375939849624, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 9414 + }, + { + "epoch": 0.2655639097744361, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 9415 + }, + { + "epoch": 0.26562406015037593, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 9416 + }, + { + "epoch": 0.2656842105263158, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 9417 + }, + { + "epoch": 0.2657443609022556, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 9418 + }, + { + "epoch": 0.2658045112781955, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.146, + "step": 9419 + }, + { + "epoch": 0.26586466165413536, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 9420 + }, + { + "epoch": 0.2659248120300752, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2011, + "step": 9421 + }, + { + "epoch": 0.26598496240601505, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2427, + "step": 9422 + }, + { + "epoch": 0.26604511278195486, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 9423 + }, + { + "epoch": 0.26610526315789473, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2786, + "step": 9424 + }, + { + "epoch": 0.2661654135338346, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2092, + "step": 9425 + }, + { + "epoch": 0.2662255639097744, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 9426 + }, + { + "epoch": 0.2662857142857143, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2677, + "step": 9427 + }, + { + "epoch": 0.2663458646616541, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 9428 + }, + { + "epoch": 0.266406015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1957, + "step": 9429 + }, + { + "epoch": 0.26646616541353385, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 9430 + }, + { + "epoch": 0.26652631578947367, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 9431 + }, + { + "epoch": 0.26658646616541354, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2014, + "step": 9432 + }, + { + "epoch": 0.2666466165413534, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2783, + "step": 9433 + }, + { + "epoch": 0.2667067669172932, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1852, + "step": 9434 + }, + { + "epoch": 0.2667669172932331, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1886, + "step": 9435 + }, + { + "epoch": 0.2668270676691729, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 9436 + }, + { + "epoch": 0.2668872180451128, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 9437 + }, + { + "epoch": 0.26694736842105266, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2317, + "step": 9438 + }, + { + "epoch": 0.26700751879699247, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2551, + "step": 9439 + }, + { + "epoch": 0.26706766917293234, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 9440 + }, + { + "epoch": 0.26712781954887216, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 9441 + }, + { + "epoch": 0.26718796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 9442 + }, + { + "epoch": 0.2672481203007519, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.384, + "step": 9443 + }, + { + "epoch": 0.2673082706766917, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 9444 + }, + { + "epoch": 0.2673684210526316, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2404, + "step": 9445 + }, + { + "epoch": 0.2674285714285714, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 9446 + }, + { + "epoch": 0.2674887218045113, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.3667, + "step": 9447 + }, + { + "epoch": 0.26754887218045115, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2035, + "step": 9448 + }, + { + "epoch": 0.26760902255639096, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 9449 + }, + { + "epoch": 0.26766917293233083, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2411, + "step": 9450 + }, + { + "epoch": 0.26772932330827065, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2557, + "step": 9451 + }, + { + "epoch": 0.2677894736842105, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.3167, + "step": 9452 + }, + { + "epoch": 0.2678496240601504, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 9453 + }, + { + "epoch": 0.2679097744360902, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 9454 + }, + { + "epoch": 0.2679699248120301, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 9455 + }, + { + "epoch": 0.26803007518796995, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 9456 + }, + { + "epoch": 0.26809022556390977, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 9457 + }, + { + "epoch": 0.26815037593984964, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 9458 + }, + { + "epoch": 0.26821052631578945, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.23, + "step": 9459 + }, + { + "epoch": 0.2682706766917293, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2666, + "step": 9460 + }, + { + "epoch": 0.2683308270676692, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.3891, + "step": 9461 + }, + { + "epoch": 0.268390977443609, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.1564, + "step": 9462 + }, + { + "epoch": 0.2684511278195489, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2308, + "step": 9463 + }, + { + "epoch": 0.2685112781954887, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.3152, + "step": 9464 + }, + { + "epoch": 0.26857142857142857, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 9465 + }, + { + "epoch": 0.26863157894736844, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 9466 + }, + { + "epoch": 0.26869172932330826, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1959, + "step": 9467 + }, + { + "epoch": 0.26875187969924813, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1814, + "step": 9468 + }, + { + "epoch": 0.26881203007518795, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 9469 + }, + { + "epoch": 0.2688721804511278, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1776, + "step": 9470 + }, + { + "epoch": 0.2689323308270677, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 9471 + }, + { + "epoch": 0.2689924812030075, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2738, + "step": 9472 + }, + { + "epoch": 0.2690526315789474, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2248, + "step": 9473 + }, + { + "epoch": 0.2691127819548872, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 9474 + }, + { + "epoch": 0.26917293233082706, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3237, + "step": 9475 + }, + { + "epoch": 0.26923308270676694, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1998, + "step": 9476 + }, + { + "epoch": 0.26929323308270675, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.249, + "step": 9477 + }, + { + "epoch": 0.2693533834586466, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 9478 + }, + { + "epoch": 0.2694135338345865, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2237, + "step": 9479 + }, + { + "epoch": 0.2694736842105263, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2987, + "step": 9480 + }, + { + "epoch": 0.2695338345864662, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 9481 + }, + { + "epoch": 0.269593984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2232, + "step": 9482 + }, + { + "epoch": 0.26965413533834587, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2156, + "step": 9483 + }, + { + "epoch": 0.26971428571428574, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1582, + "step": 9484 + }, + { + "epoch": 0.26977443609022556, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 9485 + }, + { + "epoch": 0.2698345864661654, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2499, + "step": 9486 + }, + { + "epoch": 0.26989473684210524, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2591, + "step": 9487 + }, + { + "epoch": 0.2699548872180451, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2153, + "step": 9488 + }, + { + "epoch": 0.270015037593985, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 9489 + }, + { + "epoch": 0.2700751879699248, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 9490 + }, + { + "epoch": 0.2701353383458647, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2479, + "step": 9491 + }, + { + "epoch": 0.2701954887218045, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2021, + "step": 9492 + }, + { + "epoch": 0.27025563909774436, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2192, + "step": 9493 + }, + { + "epoch": 0.27031578947368423, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2263, + "step": 9494 + }, + { + "epoch": 0.27037593984962405, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 9495 + }, + { + "epoch": 0.2704360902255639, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 9496 + }, + { + "epoch": 0.27049624060150373, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 9497 + }, + { + "epoch": 0.2705563909774436, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2888, + "step": 9498 + }, + { + "epoch": 0.2706165413533835, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 9499 + }, + { + "epoch": 0.2706766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3244, + "step": 9500 + }, + { + "epoch": 0.27073684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2712, + "step": 9501 + }, + { + "epoch": 0.270796992481203, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 9502 + }, + { + "epoch": 0.27085714285714285, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 9503 + }, + { + "epoch": 0.2709172932330827, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1943, + "step": 9504 + }, + { + "epoch": 0.27097744360902254, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2008, + "step": 9505 + }, + { + "epoch": 0.2710375939849624, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 9506 + }, + { + "epoch": 0.2710977443609023, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2576, + "step": 9507 + }, + { + "epoch": 0.2711578947368421, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 9508 + }, + { + "epoch": 0.27121804511278197, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3795, + "step": 9509 + }, + { + "epoch": 0.2712781954887218, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2508, + "step": 9510 + }, + { + "epoch": 0.27133834586466166, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2808, + "step": 9511 + }, + { + "epoch": 0.27139849624060153, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2334, + "step": 9512 + }, + { + "epoch": 0.27145864661654134, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 9513 + }, + { + "epoch": 0.2715187969924812, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2152, + "step": 9514 + }, + { + "epoch": 0.27157894736842103, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 9515 + }, + { + "epoch": 0.2716390977443609, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 9516 + }, + { + "epoch": 0.2716992481203008, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.228, + "step": 9517 + }, + { + "epoch": 0.2717593984962406, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 9518 + }, + { + "epoch": 0.27181954887218046, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2744, + "step": 9519 + }, + { + "epoch": 0.2718796992481203, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 9520 + }, + { + "epoch": 0.27193984962406015, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3055, + "step": 9521 + }, + { + "epoch": 0.272, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2569, + "step": 9522 + }, + { + "epoch": 0.27206015037593984, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2152, + "step": 9523 + }, + { + "epoch": 0.2721203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 9524 + }, + { + "epoch": 0.2721804511278195, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 9525 + }, + { + "epoch": 0.2722406015037594, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2199, + "step": 9526 + }, + { + "epoch": 0.27230075187969927, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2306, + "step": 9527 + }, + { + "epoch": 0.2723609022556391, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.219, + "step": 9528 + }, + { + "epoch": 0.27242105263157895, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 9529 + }, + { + "epoch": 0.2724812030075188, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2351, + "step": 9530 + }, + { + "epoch": 0.27254135338345864, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 9531 + }, + { + "epoch": 0.2726015037593985, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 9532 + }, + { + "epoch": 0.2726616541353383, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2694, + "step": 9533 + }, + { + "epoch": 0.2727218045112782, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2199, + "step": 9534 + }, + { + "epoch": 0.27278195488721807, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2618, + "step": 9535 + }, + { + "epoch": 0.2728421052631579, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2263, + "step": 9536 + }, + { + "epoch": 0.27290225563909776, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.3407, + "step": 9537 + }, + { + "epoch": 0.2729624060150376, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 9538 + }, + { + "epoch": 0.27302255639097744, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1689, + "step": 9539 + }, + { + "epoch": 0.2730827067669173, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 9540 + }, + { + "epoch": 0.27314285714285713, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 9541 + }, + { + "epoch": 0.273203007518797, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3411, + "step": 9542 + }, + { + "epoch": 0.2732631578947368, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1872, + "step": 9543 + }, + { + "epoch": 0.2733233082706767, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 9544 + }, + { + "epoch": 0.27338345864661656, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 9545 + }, + { + "epoch": 0.2734436090225564, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2756, + "step": 9546 + }, + { + "epoch": 0.27350375939849625, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2232, + "step": 9547 + }, + { + "epoch": 0.27356390977443606, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2808, + "step": 9548 + }, + { + "epoch": 0.27362406015037594, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.195, + "step": 9549 + }, + { + "epoch": 0.2736842105263158, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 9550 + }, + { + "epoch": 0.2737443609022556, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 9551 + }, + { + "epoch": 0.2738045112781955, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1721, + "step": 9552 + }, + { + "epoch": 0.2738646616541353, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2185, + "step": 9553 + }, + { + "epoch": 0.2739248120300752, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 9554 + }, + { + "epoch": 0.27398496240601505, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 9555 + }, + { + "epoch": 0.27404511278195487, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1735, + "step": 9556 + }, + { + "epoch": 0.27410526315789474, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2045, + "step": 9557 + }, + { + "epoch": 0.2741654135338346, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2599, + "step": 9558 + }, + { + "epoch": 0.27422556390977443, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 9559 + }, + { + "epoch": 0.2742857142857143, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2589, + "step": 9560 + }, + { + "epoch": 0.2743458646616541, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2361, + "step": 9561 + }, + { + "epoch": 0.274406015037594, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2721, + "step": 9562 + }, + { + "epoch": 0.27446616541353386, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 9563 + }, + { + "epoch": 0.2745263157894737, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 9564 + }, + { + "epoch": 0.27458646616541355, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 9565 + }, + { + "epoch": 0.27464661654135336, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 9566 + }, + { + "epoch": 0.27470676691729323, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2657, + "step": 9567 + }, + { + "epoch": 0.2747669172932331, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 9568 + }, + { + "epoch": 0.2748270676691729, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 9569 + }, + { + "epoch": 0.2748872180451128, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3684, + "step": 9570 + }, + { + "epoch": 0.2749473684210526, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2306, + "step": 9571 + }, + { + "epoch": 0.2750075187969925, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2035, + "step": 9572 + }, + { + "epoch": 0.27506766917293235, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2657, + "step": 9573 + }, + { + "epoch": 0.27512781954887217, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 9574 + }, + { + "epoch": 0.27518796992481204, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 9575 + }, + { + "epoch": 0.27524812030075185, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 9576 + }, + { + "epoch": 0.2753082706766917, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 9577 + }, + { + "epoch": 0.2753684210526316, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2383, + "step": 9578 + }, + { + "epoch": 0.2754285714285714, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2246, + "step": 9579 + }, + { + "epoch": 0.2754887218045113, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2566, + "step": 9580 + }, + { + "epoch": 0.27554887218045115, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 9581 + }, + { + "epoch": 0.27560902255639097, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 9582 + }, + { + "epoch": 0.27566917293233084, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1982, + "step": 9583 + }, + { + "epoch": 0.27572932330827066, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 9584 + }, + { + "epoch": 0.27578947368421053, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 9585 + }, + { + "epoch": 0.2758496240601504, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.21, + "step": 9586 + }, + { + "epoch": 0.2759097744360902, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2233, + "step": 9587 + }, + { + "epoch": 0.2759699248120301, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 9588 + }, + { + "epoch": 0.2760300751879699, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2462, + "step": 9589 + }, + { + "epoch": 0.2760902255639098, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.209, + "step": 9590 + }, + { + "epoch": 0.27615037593984965, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 9591 + }, + { + "epoch": 0.27621052631578946, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2281, + "step": 9592 + }, + { + "epoch": 0.27627067669172933, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1847, + "step": 9593 + }, + { + "epoch": 0.27633082706766915, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 9594 + }, + { + "epoch": 0.276390977443609, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2322, + "step": 9595 + }, + { + "epoch": 0.2764511278195489, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 9596 + }, + { + "epoch": 0.2765112781954887, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 9597 + }, + { + "epoch": 0.2765714285714286, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2308, + "step": 9598 + }, + { + "epoch": 0.2766315789473684, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3336, + "step": 9599 + }, + { + "epoch": 0.27669172932330827, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 9600 + }, + { + "epoch": 0.27675187969924814, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 9601 + }, + { + "epoch": 0.27681203007518795, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 9602 + }, + { + "epoch": 0.2768721804511278, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2181, + "step": 9603 + }, + { + "epoch": 0.2769323308270677, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 9604 + }, + { + "epoch": 0.2769924812030075, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 9605 + }, + { + "epoch": 0.2770526315789474, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1728, + "step": 9606 + }, + { + "epoch": 0.2771127819548872, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2161, + "step": 9607 + }, + { + "epoch": 0.27717293233082707, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 9608 + }, + { + "epoch": 0.27723308270676694, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2553, + "step": 9609 + }, + { + "epoch": 0.27729323308270676, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.3685, + "step": 9610 + }, + { + "epoch": 0.27735338345864663, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 9611 + }, + { + "epoch": 0.27741353383458645, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 9612 + }, + { + "epoch": 0.2774736842105263, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 9613 + }, + { + "epoch": 0.2775338345864662, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 9614 + }, + { + "epoch": 0.277593984962406, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.2967, + "step": 9615 + }, + { + "epoch": 0.2776541353383459, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2057, + "step": 9616 + }, + { + "epoch": 0.2777142857142857, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2342, + "step": 9617 + }, + { + "epoch": 0.27777443609022556, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2154, + "step": 9618 + }, + { + "epoch": 0.27783458646616543, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 9619 + }, + { + "epoch": 0.27789473684210525, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2214, + "step": 9620 + }, + { + "epoch": 0.2779548872180451, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1421, + "step": 9621 + }, + { + "epoch": 0.27801503759398494, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3018, + "step": 9622 + }, + { + "epoch": 0.2780751879699248, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 9623 + }, + { + "epoch": 0.2781353383458647, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2862, + "step": 9624 + }, + { + "epoch": 0.2781954887218045, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 9625 + }, + { + "epoch": 0.27825563909774437, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.188, + "step": 9626 + }, + { + "epoch": 0.2783157894736842, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2115, + "step": 9627 + }, + { + "epoch": 0.27837593984962405, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1996, + "step": 9628 + }, + { + "epoch": 0.2784360902255639, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2155, + "step": 9629 + }, + { + "epoch": 0.27849624060150374, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.2203, + "step": 9630 + }, + { + "epoch": 0.2785563909774436, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2605, + "step": 9631 + }, + { + "epoch": 0.2786165413533835, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3122, + "step": 9632 + }, + { + "epoch": 0.2786766917293233, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2794, + "step": 9633 + }, + { + "epoch": 0.27873684210526317, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2188, + "step": 9634 + }, + { + "epoch": 0.278796992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2042, + "step": 9635 + }, + { + "epoch": 0.27885714285714286, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1916, + "step": 9636 + }, + { + "epoch": 0.27891729323308273, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 9637 + }, + { + "epoch": 0.27897744360902255, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 9638 + }, + { + "epoch": 0.2790375939849624, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 9639 + }, + { + "epoch": 0.27909774436090223, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2563, + "step": 9640 + }, + { + "epoch": 0.2791578947368421, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 9641 + }, + { + "epoch": 0.279218045112782, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2877, + "step": 9642 + }, + { + "epoch": 0.2792781954887218, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 9643 + }, + { + "epoch": 0.27933834586466166, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2561, + "step": 9644 + }, + { + "epoch": 0.2793984962406015, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 9645 + }, + { + "epoch": 0.27945864661654135, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1986, + "step": 9646 + }, + { + "epoch": 0.2795187969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2445, + "step": 9647 + }, + { + "epoch": 0.27957894736842104, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 9648 + }, + { + "epoch": 0.2796390977443609, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 9649 + }, + { + "epoch": 0.2796992481203007, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 9650 + }, + { + "epoch": 0.2797593984962406, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 9651 + }, + { + "epoch": 0.27981954887218047, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1804, + "step": 9652 + }, + { + "epoch": 0.2798796992481203, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 9653 + }, + { + "epoch": 0.27993984962406016, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.3195, + "step": 9654 + }, + { + "epoch": 0.28, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1731, + "step": 9655 + }, + { + "epoch": 0.28006015037593984, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 9656 + }, + { + "epoch": 0.2801203007518797, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 9657 + }, + { + "epoch": 0.28018045112781953, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.199, + "step": 9658 + }, + { + "epoch": 0.2802406015037594, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.219, + "step": 9659 + }, + { + "epoch": 0.2803007518796993, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3002, + "step": 9660 + }, + { + "epoch": 0.2803609022556391, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 9661 + }, + { + "epoch": 0.28042105263157896, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2486, + "step": 9662 + }, + { + "epoch": 0.2804812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1758, + "step": 9663 + }, + { + "epoch": 0.28054135338345865, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 9664 + }, + { + "epoch": 0.2806015037593985, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.2256, + "step": 9665 + }, + { + "epoch": 0.28066165413533833, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2723, + "step": 9666 + }, + { + "epoch": 0.2807218045112782, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 9667 + }, + { + "epoch": 0.280781954887218, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2411, + "step": 9668 + }, + { + "epoch": 0.2808421052631579, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 9669 + }, + { + "epoch": 0.28090225563909776, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3099, + "step": 9670 + }, + { + "epoch": 0.2809624060150376, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 9671 + }, + { + "epoch": 0.28102255639097745, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1801, + "step": 9672 + }, + { + "epoch": 0.28108270676691727, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2781, + "step": 9673 + }, + { + "epoch": 0.28114285714285714, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 9674 + }, + { + "epoch": 0.281203007518797, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 9675 + }, + { + "epoch": 0.2812631578947368, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.324, + "step": 9676 + }, + { + "epoch": 0.2813233082706767, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3559, + "step": 9677 + }, + { + "epoch": 0.28138345864661657, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2733, + "step": 9678 + }, + { + "epoch": 0.2814436090225564, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2028, + "step": 9679 + }, + { + "epoch": 0.28150375939849626, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1834, + "step": 9680 + }, + { + "epoch": 0.28156390977443607, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 9681 + }, + { + "epoch": 0.28162406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2306, + "step": 9682 + }, + { + "epoch": 0.2816842105263158, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2535, + "step": 9683 + }, + { + "epoch": 0.28174436090225563, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1994, + "step": 9684 + }, + { + "epoch": 0.2818045112781955, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1732, + "step": 9685 + }, + { + "epoch": 0.2818646616541353, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 9686 + }, + { + "epoch": 0.2819248120300752, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2131, + "step": 9687 + }, + { + "epoch": 0.28198496240601506, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2067, + "step": 9688 + }, + { + "epoch": 0.2820451127819549, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 9689 + }, + { + "epoch": 0.28210526315789475, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 9690 + }, + { + "epoch": 0.28216541353383456, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 9691 + }, + { + "epoch": 0.28222556390977444, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1872, + "step": 9692 + }, + { + "epoch": 0.2822857142857143, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 9693 + }, + { + "epoch": 0.2823458646616541, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2451, + "step": 9694 + }, + { + "epoch": 0.282406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2458, + "step": 9695 + }, + { + "epoch": 0.2824661654135338, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 9696 + }, + { + "epoch": 0.2825263157894737, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 9697 + }, + { + "epoch": 0.28258646616541355, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 9698 + }, + { + "epoch": 0.28264661654135337, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 9699 + }, + { + "epoch": 0.28270676691729324, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 9700 + }, + { + "epoch": 0.28276691729323306, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2715, + "step": 9701 + }, + { + "epoch": 0.2828270676691729, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 9702 + }, + { + "epoch": 0.2828872180451128, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2199, + "step": 9703 + }, + { + "epoch": 0.2829473684210526, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1791, + "step": 9704 + }, + { + "epoch": 0.2830075187969925, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2472, + "step": 9705 + }, + { + "epoch": 0.28306766917293236, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 9706 + }, + { + "epoch": 0.2831278195488722, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 9707 + }, + { + "epoch": 0.28318796992481204, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2244, + "step": 9708 + }, + { + "epoch": 0.28324812030075186, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2466, + "step": 9709 + }, + { + "epoch": 0.28330827067669173, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 9710 + }, + { + "epoch": 0.2833684210526316, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3999, + "step": 9711 + }, + { + "epoch": 0.2834285714285714, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2141, + "step": 9712 + }, + { + "epoch": 0.2834887218045113, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3893, + "step": 9713 + }, + { + "epoch": 0.2835488721804511, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2184, + "step": 9714 + }, + { + "epoch": 0.283609022556391, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 9715 + }, + { + "epoch": 0.28366917293233085, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2413, + "step": 9716 + }, + { + "epoch": 0.28372932330827066, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 9717 + }, + { + "epoch": 0.28378947368421054, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 9718 + }, + { + "epoch": 0.28384962406015035, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 9719 + }, + { + "epoch": 0.2839097744360902, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2593, + "step": 9720 + }, + { + "epoch": 0.2839699248120301, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2107, + "step": 9721 + }, + { + "epoch": 0.2840300751879699, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 9722 + }, + { + "epoch": 0.2840902255639098, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.163, + "step": 9723 + }, + { + "epoch": 0.2841503759398496, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 9724 + }, + { + "epoch": 0.28421052631578947, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2607, + "step": 9725 + }, + { + "epoch": 0.28427067669172934, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 9726 + }, + { + "epoch": 0.28433082706766916, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 9727 + }, + { + "epoch": 0.284390977443609, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1754, + "step": 9728 + }, + { + "epoch": 0.2844511278195489, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.277, + "step": 9729 + }, + { + "epoch": 0.2845112781954887, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2495, + "step": 9730 + }, + { + "epoch": 0.2845714285714286, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.2092, + "step": 9731 + }, + { + "epoch": 0.2846315789473684, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.3046, + "step": 9732 + }, + { + "epoch": 0.2846917293233083, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1884, + "step": 9733 + }, + { + "epoch": 0.28475187969924814, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2531, + "step": 9734 + }, + { + "epoch": 0.28481203007518796, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 9735 + }, + { + "epoch": 0.28487218045112783, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2591, + "step": 9736 + }, + { + "epoch": 0.28493233082706765, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 9737 + }, + { + "epoch": 0.2849924812030075, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 9738 + }, + { + "epoch": 0.2850526315789474, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 9739 + }, + { + "epoch": 0.2851127819548872, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 9740 + }, + { + "epoch": 0.2851729323308271, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2587, + "step": 9741 + }, + { + "epoch": 0.2852330827067669, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.2784, + "step": 9742 + }, + { + "epoch": 0.28529323308270677, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 9743 + }, + { + "epoch": 0.28535338345864664, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2204, + "step": 9744 + }, + { + "epoch": 0.28541353383458645, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2307, + "step": 9745 + }, + { + "epoch": 0.2854736842105263, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 9746 + }, + { + "epoch": 0.28553383458646614, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 9747 + }, + { + "epoch": 0.285593984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2698, + "step": 9748 + }, + { + "epoch": 0.2856541353383459, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 9749 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 9750 + }, + { + "epoch": 0.28577443609022557, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2066, + "step": 9751 + }, + { + "epoch": 0.28583458646616544, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 9752 + }, + { + "epoch": 0.28589473684210526, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2355, + "step": 9753 + }, + { + "epoch": 0.28595488721804513, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 9754 + }, + { + "epoch": 0.28601503759398494, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 9755 + }, + { + "epoch": 0.2860751879699248, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.2945, + "step": 9756 + }, + { + "epoch": 0.2861353383458647, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 9757 + }, + { + "epoch": 0.2861954887218045, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 9758 + }, + { + "epoch": 0.2862556390977444, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 9759 + }, + { + "epoch": 0.2863157894736842, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 9760 + }, + { + "epoch": 0.28637593984962406, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 9761 + }, + { + "epoch": 0.28643609022556393, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 9762 + }, + { + "epoch": 0.28649624060150375, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 9763 + }, + { + "epoch": 0.2865563909774436, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 9764 + }, + { + "epoch": 0.28661654135338344, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1876, + "step": 9765 + }, + { + "epoch": 0.2866766917293233, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 9766 + }, + { + "epoch": 0.2867368421052632, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 9767 + }, + { + "epoch": 0.286796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 9768 + }, + { + "epoch": 0.28685714285714287, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2615, + "step": 9769 + }, + { + "epoch": 0.2869172932330827, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1781, + "step": 9770 + }, + { + "epoch": 0.28697744360902255, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 9771 + }, + { + "epoch": 0.2870375939849624, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.2796, + "step": 9772 + }, + { + "epoch": 0.28709774436090224, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 9773 + }, + { + "epoch": 0.2871578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.195, + "step": 9774 + }, + { + "epoch": 0.28721804511278193, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 9775 + }, + { + "epoch": 0.2872781954887218, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 9776 + }, + { + "epoch": 0.28733834586466167, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2196, + "step": 9777 + }, + { + "epoch": 0.2873984962406015, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2355, + "step": 9778 + }, + { + "epoch": 0.28745864661654136, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1526, + "step": 9779 + }, + { + "epoch": 0.28751879699248123, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2535, + "step": 9780 + }, + { + "epoch": 0.28757894736842105, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2, + "step": 9781 + }, + { + "epoch": 0.2876390977443609, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2676, + "step": 9782 + }, + { + "epoch": 0.28769924812030073, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 9783 + }, + { + "epoch": 0.2877593984962406, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 9784 + }, + { + "epoch": 0.2878195488721805, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2186, + "step": 9785 + }, + { + "epoch": 0.2878796992481203, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 9786 + }, + { + "epoch": 0.28793984962406016, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 9787 + }, + { + "epoch": 0.288, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 9788 + }, + { + "epoch": 0.28806015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 9789 + }, + { + "epoch": 0.2881203007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 9790 + }, + { + "epoch": 0.28818045112781954, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2383, + "step": 9791 + }, + { + "epoch": 0.2882406015037594, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2225, + "step": 9792 + }, + { + "epoch": 0.2883007518796992, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 9793 + }, + { + "epoch": 0.2883609022556391, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 9794 + }, + { + "epoch": 0.28842105263157897, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 9795 + }, + { + "epoch": 0.2884812030075188, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.214, + "step": 9796 + }, + { + "epoch": 0.28854135338345865, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1834, + "step": 9797 + }, + { + "epoch": 0.28860150375939847, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2032, + "step": 9798 + }, + { + "epoch": 0.28866165413533834, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3537, + "step": 9799 + }, + { + "epoch": 0.2887218045112782, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1859, + "step": 9800 + }, + { + "epoch": 0.28878195488721803, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 9801 + }, + { + "epoch": 0.2888421052631579, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 9802 + }, + { + "epoch": 0.28890225563909777, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2257, + "step": 9803 + }, + { + "epoch": 0.2889624060150376, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 9804 + }, + { + "epoch": 0.28902255639097746, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 9805 + }, + { + "epoch": 0.2890827067669173, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 9806 + }, + { + "epoch": 0.28914285714285715, + "grad_norm": 1.4375, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 9807 + }, + { + "epoch": 0.289203007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 9808 + }, + { + "epoch": 0.28926315789473683, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 9809 + }, + { + "epoch": 0.2893233082706767, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.353, + "step": 9810 + }, + { + "epoch": 0.2893834586466165, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 9811 + }, + { + "epoch": 0.2894436090225564, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 9812 + }, + { + "epoch": 0.28950375939849626, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 9813 + }, + { + "epoch": 0.2895639097744361, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2507, + "step": 9814 + }, + { + "epoch": 0.28962406015037595, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 9815 + }, + { + "epoch": 0.28968421052631577, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 9816 + }, + { + "epoch": 0.28974436090225564, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 9817 + }, + { + "epoch": 0.2898045112781955, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2617, + "step": 9818 + }, + { + "epoch": 0.2898646616541353, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 9819 + }, + { + "epoch": 0.2899248120300752, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3007, + "step": 9820 + }, + { + "epoch": 0.289984962406015, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2571, + "step": 9821 + }, + { + "epoch": 0.2900451127819549, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1995, + "step": 9822 + }, + { + "epoch": 0.29010526315789475, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 9823 + }, + { + "epoch": 0.29016541353383457, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2383, + "step": 9824 + }, + { + "epoch": 0.29022556390977444, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 9825 + }, + { + "epoch": 0.29028571428571426, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1761, + "step": 9826 + }, + { + "epoch": 0.29034586466165413, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.274, + "step": 9827 + }, + { + "epoch": 0.290406015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 9828 + }, + { + "epoch": 0.2904661654135338, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2575, + "step": 9829 + }, + { + "epoch": 0.2905263157894737, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2214, + "step": 9830 + }, + { + "epoch": 0.29058646616541356, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3539, + "step": 9831 + }, + { + "epoch": 0.2906466165413534, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.236, + "step": 9832 + }, + { + "epoch": 0.29070676691729325, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 9833 + }, + { + "epoch": 0.29076691729323306, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 9834 + }, + { + "epoch": 0.29082706766917293, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1597, + "step": 9835 + }, + { + "epoch": 0.2908872180451128, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 9836 + }, + { + "epoch": 0.2909473684210526, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1746, + "step": 9837 + }, + { + "epoch": 0.2910075187969925, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 9838 + }, + { + "epoch": 0.2910676691729323, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2203, + "step": 9839 + }, + { + "epoch": 0.2911278195488722, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2588, + "step": 9840 + }, + { + "epoch": 0.29118796992481205, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 9841 + }, + { + "epoch": 0.29124812030075187, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 9842 + }, + { + "epoch": 0.29130827067669174, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 9843 + }, + { + "epoch": 0.29136842105263155, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2897, + "step": 9844 + }, + { + "epoch": 0.2914285714285714, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 9845 + }, + { + "epoch": 0.2914887218045113, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 9846 + }, + { + "epoch": 0.2915488721804511, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2433, + "step": 9847 + }, + { + "epoch": 0.291609022556391, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 9848 + }, + { + "epoch": 0.2916691729323308, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2361, + "step": 9849 + }, + { + "epoch": 0.29172932330827067, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1822, + "step": 9850 + }, + { + "epoch": 0.29178947368421054, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 9851 + }, + { + "epoch": 0.29184962406015036, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3016, + "step": 9852 + }, + { + "epoch": 0.29190977443609023, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 9853 + }, + { + "epoch": 0.2919699248120301, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2952, + "step": 9854 + }, + { + "epoch": 0.2920300751879699, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 9855 + }, + { + "epoch": 0.2920902255639098, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 9856 + }, + { + "epoch": 0.2921503759398496, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 9857 + }, + { + "epoch": 0.2922105263157895, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 9858 + }, + { + "epoch": 0.29227067669172935, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1983, + "step": 9859 + }, + { + "epoch": 0.29233082706766916, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2736, + "step": 9860 + }, + { + "epoch": 0.29239097744360903, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2744, + "step": 9861 + }, + { + "epoch": 0.29245112781954885, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.2412, + "step": 9862 + }, + { + "epoch": 0.2925112781954887, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2241, + "step": 9863 + }, + { + "epoch": 0.2925714285714286, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2399, + "step": 9864 + }, + { + "epoch": 0.2926315789473684, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2094, + "step": 9865 + }, + { + "epoch": 0.2926917293233083, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2149, + "step": 9866 + }, + { + "epoch": 0.2927518796992481, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 9867 + }, + { + "epoch": 0.29281203007518797, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2615, + "step": 9868 + }, + { + "epoch": 0.29287218045112784, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 9869 + }, + { + "epoch": 0.29293233082706766, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.267, + "step": 9870 + }, + { + "epoch": 0.2929924812030075, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 9871 + }, + { + "epoch": 0.29305263157894734, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1907, + "step": 9872 + }, + { + "epoch": 0.2931127819548872, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 9873 + }, + { + "epoch": 0.2931729323308271, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.3551, + "step": 9874 + }, + { + "epoch": 0.2932330827067669, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1564, + "step": 9875 + }, + { + "epoch": 0.2932932330827068, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.267, + "step": 9876 + }, + { + "epoch": 0.29335338345864664, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2768, + "step": 9877 + }, + { + "epoch": 0.29341353383458646, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.2262, + "step": 9878 + }, + { + "epoch": 0.29347368421052633, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1914, + "step": 9879 + }, + { + "epoch": 0.29353383458646615, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2027, + "step": 9880 + }, + { + "epoch": 0.293593984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2041, + "step": 9881 + }, + { + "epoch": 0.2936541353383459, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2742, + "step": 9882 + }, + { + "epoch": 0.2937142857142857, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 9883 + }, + { + "epoch": 0.2937744360902256, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 9884 + }, + { + "epoch": 0.2938345864661654, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 9885 + }, + { + "epoch": 0.29389473684210526, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2494, + "step": 9886 + }, + { + "epoch": 0.29395488721804514, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 9887 + }, + { + "epoch": 0.29401503759398495, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 9888 + }, + { + "epoch": 0.2940751879699248, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2104, + "step": 9889 + }, + { + "epoch": 0.29413533834586464, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2062, + "step": 9890 + }, + { + "epoch": 0.2941954887218045, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 9891 + }, + { + "epoch": 0.2942556390977444, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2922, + "step": 9892 + }, + { + "epoch": 0.2943157894736842, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 9893 + }, + { + "epoch": 0.29437593984962407, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1904, + "step": 9894 + }, + { + "epoch": 0.2944360902255639, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2112, + "step": 9895 + }, + { + "epoch": 0.29449624060150376, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2601, + "step": 9896 + }, + { + "epoch": 0.2945563909774436, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.22, + "step": 9897 + }, + { + "epoch": 0.29461654135338344, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.4569, + "step": 9898 + }, + { + "epoch": 0.2946766917293233, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1793, + "step": 9899 + }, + { + "epoch": 0.29473684210526313, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 9900 + }, + { + "epoch": 0.294796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 9901 + }, + { + "epoch": 0.2948571428571429, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2502, + "step": 9902 + }, + { + "epoch": 0.2949172932330827, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1633, + "step": 9903 + }, + { + "epoch": 0.29497744360902256, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 9904 + }, + { + "epoch": 0.29503759398496243, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 9905 + }, + { + "epoch": 0.29509774436090225, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2044, + "step": 9906 + }, + { + "epoch": 0.2951578947368421, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2833, + "step": 9907 + }, + { + "epoch": 0.29521804511278193, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2584, + "step": 9908 + }, + { + "epoch": 0.2952781954887218, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 9909 + }, + { + "epoch": 0.2953383458646617, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 9910 + }, + { + "epoch": 0.2953984962406015, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2143, + "step": 9911 + }, + { + "epoch": 0.29545864661654136, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 9912 + }, + { + "epoch": 0.2955187969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 9913 + }, + { + "epoch": 0.29557894736842105, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2494, + "step": 9914 + }, + { + "epoch": 0.2956390977443609, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1671, + "step": 9915 + }, + { + "epoch": 0.29569924812030074, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2046, + "step": 9916 + }, + { + "epoch": 0.2957593984962406, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2357, + "step": 9917 + }, + { + "epoch": 0.2958195488721804, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 9918 + }, + { + "epoch": 0.2958796992481203, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2492, + "step": 9919 + }, + { + "epoch": 0.29593984962406017, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2439, + "step": 9920 + }, + { + "epoch": 0.296, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2794, + "step": 9921 + }, + { + "epoch": 0.29606015037593986, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 9922 + }, + { + "epoch": 0.2961203007518797, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 9923 + }, + { + "epoch": 0.29618045112781954, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3274, + "step": 9924 + }, + { + "epoch": 0.2962406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2357, + "step": 9925 + }, + { + "epoch": 0.29630075187969923, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2038, + "step": 9926 + }, + { + "epoch": 0.2963609022556391, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 9927 + }, + { + "epoch": 0.296421052631579, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 9928 + }, + { + "epoch": 0.2964812030075188, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2521, + "step": 9929 + }, + { + "epoch": 0.29654135338345866, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.216, + "step": 9930 + }, + { + "epoch": 0.2966015037593985, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 9931 + }, + { + "epoch": 0.29666165413533835, + "grad_norm": 1.5703125, + "learning_rate": 3e-05, + "loss": 1.2527, + "step": 9932 + }, + { + "epoch": 0.2967218045112782, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2398, + "step": 9933 + }, + { + "epoch": 0.29678195488721804, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1326, + "step": 9934 + }, + { + "epoch": 0.2968421052631579, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1696, + "step": 9935 + }, + { + "epoch": 0.2969022556390977, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 9936 + }, + { + "epoch": 0.2969624060150376, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 9937 + }, + { + "epoch": 0.29702255639097747, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2926, + "step": 9938 + }, + { + "epoch": 0.2970827067669173, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2267, + "step": 9939 + }, + { + "epoch": 0.29714285714285715, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 9940 + }, + { + "epoch": 0.29720300751879697, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 9941 + }, + { + "epoch": 0.29726315789473684, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 9942 + }, + { + "epoch": 0.2973233082706767, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 9943 + }, + { + "epoch": 0.2973834586466165, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1937, + "step": 9944 + }, + { + "epoch": 0.2974436090225564, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2759, + "step": 9945 + }, + { + "epoch": 0.2975037593984962, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 9946 + }, + { + "epoch": 0.2975639097744361, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2349, + "step": 9947 + }, + { + "epoch": 0.29762406015037596, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 9948 + }, + { + "epoch": 0.2976842105263158, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 9949 + }, + { + "epoch": 0.29774436090225564, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2157, + "step": 9950 + }, + { + "epoch": 0.2978045112781955, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 9951 + }, + { + "epoch": 0.29786466165413533, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1436, + "step": 9952 + }, + { + "epoch": 0.2979248120300752, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 9953 + }, + { + "epoch": 0.297984962406015, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2203, + "step": 9954 + }, + { + "epoch": 0.2980451127819549, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 9955 + }, + { + "epoch": 0.29810526315789476, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1726, + "step": 9956 + }, + { + "epoch": 0.2981654135338346, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 9957 + }, + { + "epoch": 0.29822556390977445, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2001, + "step": 9958 + }, + { + "epoch": 0.29828571428571427, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 9959 + }, + { + "epoch": 0.29834586466165414, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2527, + "step": 9960 + }, + { + "epoch": 0.298406015037594, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 9961 + }, + { + "epoch": 0.2984661654135338, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 9962 + }, + { + "epoch": 0.2985263157894737, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.19, + "step": 9963 + }, + { + "epoch": 0.2985864661654135, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2084, + "step": 9964 + }, + { + "epoch": 0.2986466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 9965 + }, + { + "epoch": 0.29870676691729325, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 9966 + }, + { + "epoch": 0.29876691729323307, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 9967 + }, + { + "epoch": 0.29882706766917294, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2657, + "step": 9968 + }, + { + "epoch": 0.29888721804511276, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2132, + "step": 9969 + }, + { + "epoch": 0.29894736842105263, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2277, + "step": 9970 + }, + { + "epoch": 0.2990075187969925, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2566, + "step": 9971 + }, + { + "epoch": 0.2990676691729323, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2165, + "step": 9972 + }, + { + "epoch": 0.2991278195488722, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 9973 + }, + { + "epoch": 0.299187969924812, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 9974 + }, + { + "epoch": 0.2992481203007519, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 9975 + }, + { + "epoch": 0.29930827067669175, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 9976 + }, + { + "epoch": 0.29936842105263156, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 9977 + }, + { + "epoch": 0.29942857142857143, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1541, + "step": 9978 + }, + { + "epoch": 0.2994887218045113, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2203, + "step": 9979 + }, + { + "epoch": 0.2995488721804511, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 9980 + }, + { + "epoch": 0.299609022556391, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 9981 + }, + { + "epoch": 0.2996691729323308, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 9982 + }, + { + "epoch": 0.2997293233082707, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2583, + "step": 9983 + }, + { + "epoch": 0.29978947368421055, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2408, + "step": 9984 + }, + { + "epoch": 0.29984962406015037, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2175, + "step": 9985 + }, + { + "epoch": 0.29990977443609024, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.195, + "step": 9986 + }, + { + "epoch": 0.29996992481203005, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2066, + "step": 9987 + }, + { + "epoch": 0.3000300751879699, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.248, + "step": 9988 + }, + { + "epoch": 0.3000902255639098, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 9989 + }, + { + "epoch": 0.3001503759398496, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.2974, + "step": 9990 + }, + { + "epoch": 0.3002105263157895, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3047, + "step": 9991 + }, + { + "epoch": 0.3002706766917293, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 9992 + }, + { + "epoch": 0.30033082706766917, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2456, + "step": 9993 + }, + { + "epoch": 0.30039097744360904, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 9994 + }, + { + "epoch": 0.30045112781954886, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.16, + "step": 9995 + }, + { + "epoch": 0.30051127819548873, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 9996 + }, + { + "epoch": 0.30057142857142854, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2424, + "step": 9997 + }, + { + "epoch": 0.3006315789473684, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2533, + "step": 9998 + }, + { + "epoch": 0.3006917293233083, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 9999 + }, + { + "epoch": 0.3007518796992481, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 10000 + }, + { + "epoch": 0.300812030075188, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2472, + "step": 10001 + }, + { + "epoch": 0.30087218045112785, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2855, + "step": 10002 + }, + { + "epoch": 0.30093233082706766, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 10003 + }, + { + "epoch": 0.30099248120300753, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 10004 + }, + { + "epoch": 0.30105263157894735, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 10005 + }, + { + "epoch": 0.3011127819548872, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2174, + "step": 10006 + }, + { + "epoch": 0.3011729323308271, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 10007 + }, + { + "epoch": 0.3012330827067669, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 10008 + }, + { + "epoch": 0.3012932330827068, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1868, + "step": 10009 + }, + { + "epoch": 0.3013533834586466, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 10010 + }, + { + "epoch": 0.30141353383458647, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2003, + "step": 10011 + }, + { + "epoch": 0.30147368421052634, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.239, + "step": 10012 + }, + { + "epoch": 0.30153383458646615, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3318, + "step": 10013 + }, + { + "epoch": 0.301593984962406, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2674, + "step": 10014 + }, + { + "epoch": 0.30165413533834584, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2483, + "step": 10015 + }, + { + "epoch": 0.3017142857142857, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 10016 + }, + { + "epoch": 0.3017744360902256, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.209, + "step": 10017 + }, + { + "epoch": 0.3018345864661654, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 10018 + }, + { + "epoch": 0.30189473684210527, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.229, + "step": 10019 + }, + { + "epoch": 0.3019548872180451, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2529, + "step": 10020 + }, + { + "epoch": 0.30201503759398496, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 10021 + }, + { + "epoch": 0.30207518796992483, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2744, + "step": 10022 + }, + { + "epoch": 0.30213533834586465, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2224, + "step": 10023 + }, + { + "epoch": 0.3021954887218045, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2804, + "step": 10024 + }, + { + "epoch": 0.3022556390977444, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 10025 + }, + { + "epoch": 0.3023157894736842, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 10026 + }, + { + "epoch": 0.3023759398496241, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2555, + "step": 10027 + }, + { + "epoch": 0.3024360902255639, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 10028 + }, + { + "epoch": 0.30249624060150376, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1822, + "step": 10029 + }, + { + "epoch": 0.30255639097744363, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.251, + "step": 10030 + }, + { + "epoch": 0.30261654135338345, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 10031 + }, + { + "epoch": 0.3026766917293233, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 10032 + }, + { + "epoch": 0.30273684210526314, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1991, + "step": 10033 + }, + { + "epoch": 0.302796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2408, + "step": 10034 + }, + { + "epoch": 0.3028571428571429, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2326, + "step": 10035 + }, + { + "epoch": 0.3029172932330827, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2358, + "step": 10036 + }, + { + "epoch": 0.30297744360902257, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2193, + "step": 10037 + }, + { + "epoch": 0.3030375939849624, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 10038 + }, + { + "epoch": 0.30309774436090225, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 10039 + }, + { + "epoch": 0.3031578947368421, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2281, + "step": 10040 + }, + { + "epoch": 0.30321804511278194, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2672, + "step": 10041 + }, + { + "epoch": 0.3032781954887218, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3542, + "step": 10042 + }, + { + "epoch": 0.30333834586466163, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1836, + "step": 10043 + }, + { + "epoch": 0.3033984962406015, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2039, + "step": 10044 + }, + { + "epoch": 0.30345864661654137, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2812, + "step": 10045 + }, + { + "epoch": 0.3035187969924812, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2173, + "step": 10046 + }, + { + "epoch": 0.30357894736842106, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1904, + "step": 10047 + }, + { + "epoch": 0.3036390977443609, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2478, + "step": 10048 + }, + { + "epoch": 0.30369924812030075, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 10049 + }, + { + "epoch": 0.3037593984962406, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3788, + "step": 10050 + }, + { + "epoch": 0.30381954887218043, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 10051 + }, + { + "epoch": 0.3038796992481203, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 10052 + }, + { + "epoch": 0.3039398496240602, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2118, + "step": 10053 + }, + { + "epoch": 0.304, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2007, + "step": 10054 + }, + { + "epoch": 0.30406015037593986, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 10055 + }, + { + "epoch": 0.3041203007518797, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2308, + "step": 10056 + }, + { + "epoch": 0.30418045112781955, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 10057 + }, + { + "epoch": 0.3042406015037594, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 10058 + }, + { + "epoch": 0.30430075187969924, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 10059 + }, + { + "epoch": 0.3043609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 10060 + }, + { + "epoch": 0.3044210526315789, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1662, + "step": 10061 + }, + { + "epoch": 0.3044812030075188, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 10062 + }, + { + "epoch": 0.30454135338345867, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 10063 + }, + { + "epoch": 0.3046015037593985, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3272, + "step": 10064 + }, + { + "epoch": 0.30466165413533836, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 10065 + }, + { + "epoch": 0.30472180451127817, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 10066 + }, + { + "epoch": 0.30478195488721804, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2043, + "step": 10067 + }, + { + "epoch": 0.3048421052631579, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1939, + "step": 10068 + }, + { + "epoch": 0.30490225563909773, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 10069 + }, + { + "epoch": 0.3049624060150376, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2005, + "step": 10070 + }, + { + "epoch": 0.3050225563909774, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.25, + "step": 10071 + }, + { + "epoch": 0.3050827067669173, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 10072 + }, + { + "epoch": 0.30514285714285716, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 10073 + }, + { + "epoch": 0.305203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 10074 + }, + { + "epoch": 0.30526315789473685, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.278, + "step": 10075 + }, + { + "epoch": 0.3053233082706767, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 10076 + }, + { + "epoch": 0.30538345864661653, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2345, + "step": 10077 + }, + { + "epoch": 0.3054436090225564, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2246, + "step": 10078 + }, + { + "epoch": 0.3055037593984962, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 10079 + }, + { + "epoch": 0.3055639097744361, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 10080 + }, + { + "epoch": 0.30562406015037596, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 10081 + }, + { + "epoch": 0.3056842105263158, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2061, + "step": 10082 + }, + { + "epoch": 0.30574436090225565, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3326, + "step": 10083 + }, + { + "epoch": 0.30580451127819547, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 10084 + }, + { + "epoch": 0.30586466165413534, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1841, + "step": 10085 + }, + { + "epoch": 0.3059248120300752, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 10086 + }, + { + "epoch": 0.305984962406015, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 10087 + }, + { + "epoch": 0.3060451127819549, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1995, + "step": 10088 + }, + { + "epoch": 0.3061052631578947, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2626, + "step": 10089 + }, + { + "epoch": 0.3061654135338346, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 10090 + }, + { + "epoch": 0.30622556390977446, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2011, + "step": 10091 + }, + { + "epoch": 0.3062857142857143, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 10092 + }, + { + "epoch": 0.30634586466165414, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 10093 + }, + { + "epoch": 0.30640601503759396, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 10094 + }, + { + "epoch": 0.30646616541353383, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.263, + "step": 10095 + }, + { + "epoch": 0.3065263157894737, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 10096 + }, + { + "epoch": 0.3065864661654135, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 10097 + }, + { + "epoch": 0.3066466165413534, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2305, + "step": 10098 + }, + { + "epoch": 0.3067067669172932, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 10099 + }, + { + "epoch": 0.3067669172932331, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2265, + "step": 10100 + }, + { + "epoch": 0.30682706766917295, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 10101 + }, + { + "epoch": 0.30688721804511276, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 10102 + }, + { + "epoch": 0.30694736842105264, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2019, + "step": 10103 + }, + { + "epoch": 0.3070075187969925, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.5077, + "step": 10104 + }, + { + "epoch": 0.3070676691729323, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.5812, + "step": 10105 + }, + { + "epoch": 0.3071278195488722, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.2436, + "step": 10106 + }, + { + "epoch": 0.307187969924812, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 10107 + }, + { + "epoch": 0.3072481203007519, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2769, + "step": 10108 + }, + { + "epoch": 0.30730827067669175, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2552, + "step": 10109 + }, + { + "epoch": 0.30736842105263157, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 10110 + }, + { + "epoch": 0.30742857142857144, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1801, + "step": 10111 + }, + { + "epoch": 0.30748872180451126, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1767, + "step": 10112 + }, + { + "epoch": 0.3075488721804511, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.234, + "step": 10113 + }, + { + "epoch": 0.307609022556391, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 10114 + }, + { + "epoch": 0.3076691729323308, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.2157, + "step": 10115 + }, + { + "epoch": 0.3077293233082707, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 10116 + }, + { + "epoch": 0.3077894736842105, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 10117 + }, + { + "epoch": 0.3078496240601504, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 10118 + }, + { + "epoch": 0.30790977443609024, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1843, + "step": 10119 + }, + { + "epoch": 0.30796992481203006, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 10120 + }, + { + "epoch": 0.30803007518796993, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 10121 + }, + { + "epoch": 0.30809022556390975, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 10122 + }, + { + "epoch": 0.3081503759398496, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2177, + "step": 10123 + }, + { + "epoch": 0.3082105263157895, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.237, + "step": 10124 + }, + { + "epoch": 0.3082706766917293, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 10125 + }, + { + "epoch": 0.3083308270676692, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 10126 + }, + { + "epoch": 0.30839097744360905, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 10127 + }, + { + "epoch": 0.30845112781954886, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2444, + "step": 10128 + }, + { + "epoch": 0.30851127819548874, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2171, + "step": 10129 + }, + { + "epoch": 0.30857142857142855, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 10130 + }, + { + "epoch": 0.3086315789473684, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 10131 + }, + { + "epoch": 0.3086917293233083, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 10132 + }, + { + "epoch": 0.3087518796992481, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 10133 + }, + { + "epoch": 0.308812030075188, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 10134 + }, + { + "epoch": 0.3088721804511278, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 10135 + }, + { + "epoch": 0.30893233082706767, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 10136 + }, + { + "epoch": 0.30899248120300754, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 10137 + }, + { + "epoch": 0.30905263157894736, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.2215, + "step": 10138 + }, + { + "epoch": 0.30911278195488723, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 10139 + }, + { + "epoch": 0.30917293233082704, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3518, + "step": 10140 + }, + { + "epoch": 0.3092330827067669, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.3995, + "step": 10141 + }, + { + "epoch": 0.3092932330827068, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 10142 + }, + { + "epoch": 0.3093533834586466, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 10143 + }, + { + "epoch": 0.3094135338345865, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 10144 + }, + { + "epoch": 0.3094736842105263, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.3026, + "step": 10145 + }, + { + "epoch": 0.30953383458646616, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.273, + "step": 10146 + }, + { + "epoch": 0.30959398496240603, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 10147 + }, + { + "epoch": 0.30965413533834585, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2504, + "step": 10148 + }, + { + "epoch": 0.3097142857142857, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 10149 + }, + { + "epoch": 0.3097744360902256, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 10150 + }, + { + "epoch": 0.3098345864661654, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.3076, + "step": 10151 + }, + { + "epoch": 0.3098947368421053, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2476, + "step": 10152 + }, + { + "epoch": 0.3099548872180451, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 10153 + }, + { + "epoch": 0.31001503759398497, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.285, + "step": 10154 + }, + { + "epoch": 0.31007518796992484, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 10155 + }, + { + "epoch": 0.31013533834586465, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.341, + "step": 10156 + }, + { + "epoch": 0.3101954887218045, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 10157 + }, + { + "epoch": 0.31025563909774434, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 10158 + }, + { + "epoch": 0.3103157894736842, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 10159 + }, + { + "epoch": 0.3103759398496241, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1793, + "step": 10160 + }, + { + "epoch": 0.3104360902255639, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2523, + "step": 10161 + }, + { + "epoch": 0.31049624060150377, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 10162 + }, + { + "epoch": 0.3105563909774436, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 10163 + }, + { + "epoch": 0.31061654135338346, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 10164 + }, + { + "epoch": 0.31067669172932333, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2509, + "step": 10165 + }, + { + "epoch": 0.31073684210526314, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2556, + "step": 10166 + }, + { + "epoch": 0.310796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3206, + "step": 10167 + }, + { + "epoch": 0.31085714285714283, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2022, + "step": 10168 + }, + { + "epoch": 0.3109172932330827, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 10169 + }, + { + "epoch": 0.3109774436090226, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 10170 + }, + { + "epoch": 0.3110375939849624, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 10171 + }, + { + "epoch": 0.31109774436090226, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 10172 + }, + { + "epoch": 0.3111578947368421, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 10173 + }, + { + "epoch": 0.31121804511278195, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 10174 + }, + { + "epoch": 0.3112781954887218, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 10175 + }, + { + "epoch": 0.31133834586466164, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 10176 + }, + { + "epoch": 0.3113984962406015, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.211, + "step": 10177 + }, + { + "epoch": 0.3114586466165414, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 10178 + }, + { + "epoch": 0.3115187969924812, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 10179 + }, + { + "epoch": 0.31157894736842107, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2328, + "step": 10180 + }, + { + "epoch": 0.3116390977443609, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2129, + "step": 10181 + }, + { + "epoch": 0.31169924812030075, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2077, + "step": 10182 + }, + { + "epoch": 0.3117593984962406, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2249, + "step": 10183 + }, + { + "epoch": 0.31181954887218044, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 10184 + }, + { + "epoch": 0.3118796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 10185 + }, + { + "epoch": 0.31193984962406013, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2169, + "step": 10186 + }, + { + "epoch": 0.312, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1845, + "step": 10187 + }, + { + "epoch": 0.31206015037593987, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 10188 + }, + { + "epoch": 0.3121203007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 10189 + }, + { + "epoch": 0.31218045112781956, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.226, + "step": 10190 + }, + { + "epoch": 0.3122406015037594, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 10191 + }, + { + "epoch": 0.31230075187969925, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 10192 + }, + { + "epoch": 0.3123609022556391, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 10193 + }, + { + "epoch": 0.31242105263157893, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1446, + "step": 10194 + }, + { + "epoch": 0.3124812030075188, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2316, + "step": 10195 + }, + { + "epoch": 0.3125413533834586, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 10196 + }, + { + "epoch": 0.3126015037593985, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1912, + "step": 10197 + }, + { + "epoch": 0.31266165413533836, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2391, + "step": 10198 + }, + { + "epoch": 0.3127218045112782, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.195, + "step": 10199 + }, + { + "epoch": 0.31278195488721805, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 10200 + }, + { + "epoch": 0.3128421052631579, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1956, + "step": 10201 + }, + { + "epoch": 0.31290225563909774, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 10202 + }, + { + "epoch": 0.3129624060150376, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 10203 + }, + { + "epoch": 0.3130225563909774, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2615, + "step": 10204 + }, + { + "epoch": 0.3130827067669173, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 10205 + }, + { + "epoch": 0.31314285714285717, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2326, + "step": 10206 + }, + { + "epoch": 0.313203007518797, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2659, + "step": 10207 + }, + { + "epoch": 0.31326315789473685, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.202, + "step": 10208 + }, + { + "epoch": 0.31332330827067667, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1995, + "step": 10209 + }, + { + "epoch": 0.31338345864661654, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.3087, + "step": 10210 + }, + { + "epoch": 0.3134436090225564, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 10211 + }, + { + "epoch": 0.31350375939849623, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 10212 + }, + { + "epoch": 0.3135639097744361, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 10213 + }, + { + "epoch": 0.3136240601503759, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.3459, + "step": 10214 + }, + { + "epoch": 0.3136842105263158, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1983, + "step": 10215 + }, + { + "epoch": 0.31374436090225566, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2006, + "step": 10216 + }, + { + "epoch": 0.3138045112781955, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 10217 + }, + { + "epoch": 0.31386466165413535, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2443, + "step": 10218 + }, + { + "epoch": 0.31392481203007516, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 10219 + }, + { + "epoch": 0.31398496240601503, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2873, + "step": 10220 + }, + { + "epoch": 0.3140451127819549, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 10221 + }, + { + "epoch": 0.3141052631578947, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 10222 + }, + { + "epoch": 0.3141654135338346, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2706, + "step": 10223 + }, + { + "epoch": 0.31422556390977446, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 10224 + }, + { + "epoch": 0.3142857142857143, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2093, + "step": 10225 + }, + { + "epoch": 0.31434586466165415, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 10226 + }, + { + "epoch": 0.31440601503759397, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 10227 + }, + { + "epoch": 0.31446616541353384, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 10228 + }, + { + "epoch": 0.3145263157894737, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2183, + "step": 10229 + }, + { + "epoch": 0.3145864661654135, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 10230 + }, + { + "epoch": 0.3146466165413534, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 10231 + }, + { + "epoch": 0.3147067669172932, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 10232 + }, + { + "epoch": 0.3147669172932331, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 10233 + }, + { + "epoch": 0.31482706766917296, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2531, + "step": 10234 + }, + { + "epoch": 0.31488721804511277, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2099, + "step": 10235 + }, + { + "epoch": 0.31494736842105264, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2056, + "step": 10236 + }, + { + "epoch": 0.31500751879699246, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2407, + "step": 10237 + }, + { + "epoch": 0.31506766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 10238 + }, + { + "epoch": 0.3151278195488722, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 10239 + }, + { + "epoch": 0.315187969924812, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 10240 + }, + { + "epoch": 0.3152481203007519, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 10241 + }, + { + "epoch": 0.3153082706766917, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 10242 + }, + { + "epoch": 0.3153684210526316, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 10243 + }, + { + "epoch": 0.31542857142857145, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 10244 + }, + { + "epoch": 0.31548872180451126, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2309, + "step": 10245 + }, + { + "epoch": 0.31554887218045113, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1731, + "step": 10246 + }, + { + "epoch": 0.31560902255639095, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 10247 + }, + { + "epoch": 0.3156691729323308, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 10248 + }, + { + "epoch": 0.3157293233082707, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 10249 + }, + { + "epoch": 0.3157894736842105, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2405, + "step": 10250 + }, + { + "epoch": 0.3158496240601504, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 10251 + }, + { + "epoch": 0.31590977443609025, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2576, + "step": 10252 + }, + { + "epoch": 0.31596992481203007, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 10253 + }, + { + "epoch": 0.31603007518796994, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2378, + "step": 10254 + }, + { + "epoch": 0.31609022556390975, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2238, + "step": 10255 + }, + { + "epoch": 0.3161503759398496, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2024, + "step": 10256 + }, + { + "epoch": 0.3162105263157895, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2257, + "step": 10257 + }, + { + "epoch": 0.3162706766917293, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2275, + "step": 10258 + }, + { + "epoch": 0.3163308270676692, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 10259 + }, + { + "epoch": 0.316390977443609, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2057, + "step": 10260 + }, + { + "epoch": 0.31645112781954887, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 10261 + }, + { + "epoch": 0.31651127819548874, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2191, + "step": 10262 + }, + { + "epoch": 0.31657142857142856, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 10263 + }, + { + "epoch": 0.31663157894736843, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 10264 + }, + { + "epoch": 0.31669172932330825, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 10265 + }, + { + "epoch": 0.3167518796992481, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2232, + "step": 10266 + }, + { + "epoch": 0.316812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2739, + "step": 10267 + }, + { + "epoch": 0.3168721804511278, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.192, + "step": 10268 + }, + { + "epoch": 0.3169323308270677, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 10269 + }, + { + "epoch": 0.3169924812030075, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1552, + "step": 10270 + }, + { + "epoch": 0.31705263157894736, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 10271 + }, + { + "epoch": 0.31711278195488724, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.2605, + "step": 10272 + }, + { + "epoch": 0.31717293233082705, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 10273 + }, + { + "epoch": 0.3172330827067669, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2605, + "step": 10274 + }, + { + "epoch": 0.3172932330827068, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 10275 + }, + { + "epoch": 0.3173533834586466, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2609, + "step": 10276 + }, + { + "epoch": 0.3174135338345865, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1972, + "step": 10277 + }, + { + "epoch": 0.3174736842105263, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 10278 + }, + { + "epoch": 0.31753383458646617, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2003, + "step": 10279 + }, + { + "epoch": 0.31759398496240604, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2587, + "step": 10280 + }, + { + "epoch": 0.31765413533834586, + "grad_norm": 0.43359375, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 10281 + }, + { + "epoch": 0.3177142857142857, + "grad_norm": 1.6953125, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 10282 + }, + { + "epoch": 0.31777443609022554, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 10283 + }, + { + "epoch": 0.3178345864661654, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 10284 + }, + { + "epoch": 0.3178947368421053, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 10285 + }, + { + "epoch": 0.3179548872180451, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 10286 + }, + { + "epoch": 0.318015037593985, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 10287 + }, + { + "epoch": 0.3180751879699248, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2637, + "step": 10288 + }, + { + "epoch": 0.31813533834586466, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1633, + "step": 10289 + }, + { + "epoch": 0.31819548872180453, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 10290 + }, + { + "epoch": 0.31825563909774435, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1529, + "step": 10291 + }, + { + "epoch": 0.3183157894736842, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2647, + "step": 10292 + }, + { + "epoch": 0.31837593984962403, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2258, + "step": 10293 + }, + { + "epoch": 0.3184360902255639, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1912, + "step": 10294 + }, + { + "epoch": 0.3184962406015038, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.3204, + "step": 10295 + }, + { + "epoch": 0.3185563909774436, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2233, + "step": 10296 + }, + { + "epoch": 0.31861654135338346, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 10297 + }, + { + "epoch": 0.31867669172932334, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 10298 + }, + { + "epoch": 0.31873684210526315, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2093, + "step": 10299 + }, + { + "epoch": 0.318796992481203, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 10300 + }, + { + "epoch": 0.31885714285714284, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 10301 + }, + { + "epoch": 0.3189172932330827, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 10302 + }, + { + "epoch": 0.3189774436090226, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 10303 + }, + { + "epoch": 0.3190375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 10304 + }, + { + "epoch": 0.31909774436090227, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 10305 + }, + { + "epoch": 0.3191578947368421, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 10306 + }, + { + "epoch": 0.31921804511278196, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2757, + "step": 10307 + }, + { + "epoch": 0.3192781954887218, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 10308 + }, + { + "epoch": 0.31933834586466164, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2691, + "step": 10309 + }, + { + "epoch": 0.3193984962406015, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1617, + "step": 10310 + }, + { + "epoch": 0.31945864661654133, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 10311 + }, + { + "epoch": 0.3195187969924812, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2022, + "step": 10312 + }, + { + "epoch": 0.3195789473684211, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 10313 + }, + { + "epoch": 0.3196390977443609, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2301, + "step": 10314 + }, + { + "epoch": 0.31969924812030076, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1617, + "step": 10315 + }, + { + "epoch": 0.3197593984962406, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 10316 + }, + { + "epoch": 0.31981954887218045, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 10317 + }, + { + "epoch": 0.3198796992481203, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 10318 + }, + { + "epoch": 0.31993984962406014, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2139, + "step": 10319 + }, + { + "epoch": 0.32, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 10320 + }, + { + "epoch": 0.3200601503759398, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 10321 + }, + { + "epoch": 0.3201203007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 10322 + }, + { + "epoch": 0.32018045112781957, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 10323 + }, + { + "epoch": 0.3202406015037594, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 10324 + }, + { + "epoch": 0.32030075187969925, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2104, + "step": 10325 + }, + { + "epoch": 0.3203609022556391, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1732, + "step": 10326 + }, + { + "epoch": 0.32042105263157894, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1436, + "step": 10327 + }, + { + "epoch": 0.3204812030075188, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 10328 + }, + { + "epoch": 0.3205413533834586, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2289, + "step": 10329 + }, + { + "epoch": 0.3206015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2782, + "step": 10330 + }, + { + "epoch": 0.32066165413533837, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 10331 + }, + { + "epoch": 0.3207218045112782, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1876, + "step": 10332 + }, + { + "epoch": 0.32078195488721806, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 10333 + }, + { + "epoch": 0.3208421052631579, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 10334 + }, + { + "epoch": 0.32090225563909774, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 10335 + }, + { + "epoch": 0.3209624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2962, + "step": 10336 + }, + { + "epoch": 0.32102255639097743, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 10337 + }, + { + "epoch": 0.3210827067669173, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2141, + "step": 10338 + }, + { + "epoch": 0.3211428571428571, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 10339 + }, + { + "epoch": 0.321203007518797, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 10340 + }, + { + "epoch": 0.32126315789473686, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 10341 + }, + { + "epoch": 0.3213233082706767, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 10342 + }, + { + "epoch": 0.32138345864661655, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3119, + "step": 10343 + }, + { + "epoch": 0.32144360902255636, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.4456, + "step": 10344 + }, + { + "epoch": 0.32150375939849624, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 10345 + }, + { + "epoch": 0.3215639097744361, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 10346 + }, + { + "epoch": 0.3216240601503759, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 10347 + }, + { + "epoch": 0.3216842105263158, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2055, + "step": 10348 + }, + { + "epoch": 0.32174436090225567, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 10349 + }, + { + "epoch": 0.3218045112781955, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2672, + "step": 10350 + }, + { + "epoch": 0.32186466165413535, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 10351 + }, + { + "epoch": 0.32192481203007517, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 10352 + }, + { + "epoch": 0.32198496240601504, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2663, + "step": 10353 + }, + { + "epoch": 0.3220451127819549, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2134, + "step": 10354 + }, + { + "epoch": 0.32210526315789473, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2612, + "step": 10355 + }, + { + "epoch": 0.3221654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2416, + "step": 10356 + }, + { + "epoch": 0.3222255639097744, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2022, + "step": 10357 + }, + { + "epoch": 0.3222857142857143, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 10358 + }, + { + "epoch": 0.32234586466165416, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 10359 + }, + { + "epoch": 0.322406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2314, + "step": 10360 + }, + { + "epoch": 0.32246616541353385, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2691, + "step": 10361 + }, + { + "epoch": 0.32252631578947366, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2579, + "step": 10362 + }, + { + "epoch": 0.32258646616541353, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2616, + "step": 10363 + }, + { + "epoch": 0.3226466165413534, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 10364 + }, + { + "epoch": 0.3227067669172932, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2866, + "step": 10365 + }, + { + "epoch": 0.3227669172932331, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2768, + "step": 10366 + }, + { + "epoch": 0.3228270676691729, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3478, + "step": 10367 + }, + { + "epoch": 0.3228872180451128, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 10368 + }, + { + "epoch": 0.32294736842105265, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 10369 + }, + { + "epoch": 0.32300751879699247, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2343, + "step": 10370 + }, + { + "epoch": 0.32306766917293234, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 10371 + }, + { + "epoch": 0.32312781954887215, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 10372 + }, + { + "epoch": 0.323187969924812, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 10373 + }, + { + "epoch": 0.3232481203007519, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2456, + "step": 10374 + }, + { + "epoch": 0.3233082706766917, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 10375 + }, + { + "epoch": 0.3233684210526316, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.3752, + "step": 10376 + }, + { + "epoch": 0.32342857142857145, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1903, + "step": 10377 + }, + { + "epoch": 0.32348872180451127, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 10378 + }, + { + "epoch": 0.32354887218045114, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2279, + "step": 10379 + }, + { + "epoch": 0.32360902255639096, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.255, + "step": 10380 + }, + { + "epoch": 0.32366917293233083, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 10381 + }, + { + "epoch": 0.3237293233082707, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 10382 + }, + { + "epoch": 0.3237894736842105, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.2515, + "step": 10383 + }, + { + "epoch": 0.3238496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 10384 + }, + { + "epoch": 0.3239097744360902, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 10385 + }, + { + "epoch": 0.3239699248120301, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.2808, + "step": 10386 + }, + { + "epoch": 0.32403007518796995, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1814, + "step": 10387 + }, + { + "epoch": 0.32409022556390976, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 10388 + }, + { + "epoch": 0.32415037593984963, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 10389 + }, + { + "epoch": 0.32421052631578945, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2002, + "step": 10390 + }, + { + "epoch": 0.3242706766917293, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.246, + "step": 10391 + }, + { + "epoch": 0.3243308270676692, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 10392 + }, + { + "epoch": 0.324390977443609, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2583, + "step": 10393 + }, + { + "epoch": 0.3244511278195489, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 10394 + }, + { + "epoch": 0.3245112781954887, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 10395 + }, + { + "epoch": 0.32457142857142857, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 10396 + }, + { + "epoch": 0.32463157894736844, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 10397 + }, + { + "epoch": 0.32469172932330825, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 10398 + }, + { + "epoch": 0.3247518796992481, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 10399 + }, + { + "epoch": 0.324812030075188, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 10400 + }, + { + "epoch": 0.3248721804511278, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 10401 + }, + { + "epoch": 0.3249323308270677, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2176, + "step": 10402 + }, + { + "epoch": 0.3249924812030075, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2812, + "step": 10403 + }, + { + "epoch": 0.32505263157894737, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2156, + "step": 10404 + }, + { + "epoch": 0.32511278195488724, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 10405 + }, + { + "epoch": 0.32517293233082706, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1658, + "step": 10406 + }, + { + "epoch": 0.32523308270676693, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2745, + "step": 10407 + }, + { + "epoch": 0.32529323308270675, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2357, + "step": 10408 + }, + { + "epoch": 0.3253533834586466, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 10409 + }, + { + "epoch": 0.3254135338345865, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2539, + "step": 10410 + }, + { + "epoch": 0.3254736842105263, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 10411 + }, + { + "epoch": 0.3255338345864662, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 10412 + }, + { + "epoch": 0.325593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 10413 + }, + { + "epoch": 0.32565413533834586, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 10414 + }, + { + "epoch": 0.32571428571428573, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 10415 + }, + { + "epoch": 0.32577443609022555, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 10416 + }, + { + "epoch": 0.3258345864661654, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.3086, + "step": 10417 + }, + { + "epoch": 0.32589473684210524, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 10418 + }, + { + "epoch": 0.3259548872180451, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 10419 + }, + { + "epoch": 0.326015037593985, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 10420 + }, + { + "epoch": 0.3260751879699248, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 10421 + }, + { + "epoch": 0.32613533834586467, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2501, + "step": 10422 + }, + { + "epoch": 0.32619548872180454, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 10423 + }, + { + "epoch": 0.32625563909774435, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2213, + "step": 10424 + }, + { + "epoch": 0.3263157894736842, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 10425 + }, + { + "epoch": 0.32637593984962404, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 10426 + }, + { + "epoch": 0.3264360902255639, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 10427 + }, + { + "epoch": 0.3264962406015038, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3338, + "step": 10428 + }, + { + "epoch": 0.3265563909774436, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 10429 + }, + { + "epoch": 0.32661654135338347, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2161, + "step": 10430 + }, + { + "epoch": 0.3266766917293233, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 10431 + }, + { + "epoch": 0.32673684210526316, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 10432 + }, + { + "epoch": 0.32679699248120303, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2183, + "step": 10433 + }, + { + "epoch": 0.32685714285714285, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2029, + "step": 10434 + }, + { + "epoch": 0.3269172932330827, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1635, + "step": 10435 + }, + { + "epoch": 0.32697744360902253, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2569, + "step": 10436 + }, + { + "epoch": 0.3270375939849624, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.192, + "step": 10437 + }, + { + "epoch": 0.3270977443609023, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2616, + "step": 10438 + }, + { + "epoch": 0.3271578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 10439 + }, + { + "epoch": 0.32721804511278196, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 10440 + }, + { + "epoch": 0.3272781954887218, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.267, + "step": 10441 + }, + { + "epoch": 0.32733834586466165, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 10442 + }, + { + "epoch": 0.3273984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 10443 + }, + { + "epoch": 0.32745864661654134, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2028, + "step": 10444 + }, + { + "epoch": 0.3275187969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2871, + "step": 10445 + }, + { + "epoch": 0.327578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 10446 + }, + { + "epoch": 0.3276390977443609, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 10447 + }, + { + "epoch": 0.32769924812030077, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2115, + "step": 10448 + }, + { + "epoch": 0.3277593984962406, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1929, + "step": 10449 + }, + { + "epoch": 0.32781954887218046, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2139, + "step": 10450 + }, + { + "epoch": 0.3278796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 10451 + }, + { + "epoch": 0.32793984962406014, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 10452 + }, + { + "epoch": 0.328, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.25, + "step": 10453 + }, + { + "epoch": 0.32806015037593983, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 10454 + }, + { + "epoch": 0.3281203007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 10455 + }, + { + "epoch": 0.3281804511278196, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2304, + "step": 10456 + }, + { + "epoch": 0.3282406015037594, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 10457 + }, + { + "epoch": 0.32830075187969926, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2364, + "step": 10458 + }, + { + "epoch": 0.3283609022556391, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 10459 + }, + { + "epoch": 0.32842105263157895, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1996, + "step": 10460 + }, + { + "epoch": 0.3284812030075188, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2888, + "step": 10461 + }, + { + "epoch": 0.32854135338345863, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1733, + "step": 10462 + }, + { + "epoch": 0.3286015037593985, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 10463 + }, + { + "epoch": 0.3286616541353383, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 10464 + }, + { + "epoch": 0.3287218045112782, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 10465 + }, + { + "epoch": 0.32878195488721806, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1773, + "step": 10466 + }, + { + "epoch": 0.3288421052631579, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 10467 + }, + { + "epoch": 0.32890225563909775, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 10468 + }, + { + "epoch": 0.32896240601503757, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2005, + "step": 10469 + }, + { + "epoch": 0.32902255639097744, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2038, + "step": 10470 + }, + { + "epoch": 0.3290827067669173, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2179, + "step": 10471 + }, + { + "epoch": 0.3291428571428571, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 10472 + }, + { + "epoch": 0.329203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2507, + "step": 10473 + }, + { + "epoch": 0.32926315789473687, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1988, + "step": 10474 + }, + { + "epoch": 0.3293233082706767, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2829, + "step": 10475 + }, + { + "epoch": 0.32938345864661656, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2685, + "step": 10476 + }, + { + "epoch": 0.32944360902255637, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 10477 + }, + { + "epoch": 0.32950375939849624, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2423, + "step": 10478 + }, + { + "epoch": 0.3295639097744361, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 10479 + }, + { + "epoch": 0.32962406015037593, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 10480 + }, + { + "epoch": 0.3296842105263158, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.252, + "step": 10481 + }, + { + "epoch": 0.3297443609022556, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.24, + "step": 10482 + }, + { + "epoch": 0.3298045112781955, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 10483 + }, + { + "epoch": 0.32986466165413536, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 10484 + }, + { + "epoch": 0.3299248120300752, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 10485 + }, + { + "epoch": 0.32998496240601505, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 10486 + }, + { + "epoch": 0.33004511278195486, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 10487 + }, + { + "epoch": 0.33010526315789473, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 10488 + }, + { + "epoch": 0.3301654135338346, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 10489 + }, + { + "epoch": 0.3302255639097744, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 10490 + }, + { + "epoch": 0.3302857142857143, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 10491 + }, + { + "epoch": 0.3303458646616541, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2143, + "step": 10492 + }, + { + "epoch": 0.330406015037594, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 10493 + }, + { + "epoch": 0.33046616541353385, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 10494 + }, + { + "epoch": 0.33052631578947367, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 10495 + }, + { + "epoch": 0.33058646616541354, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3779, + "step": 10496 + }, + { + "epoch": 0.3306466165413534, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 10497 + }, + { + "epoch": 0.3307067669172932, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 10498 + }, + { + "epoch": 0.3307669172932331, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 10499 + }, + { + "epoch": 0.3308270676691729, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1951, + "step": 10500 + }, + { + "epoch": 0.3308872180451128, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 10501 + }, + { + "epoch": 0.33094736842105266, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 10502 + }, + { + "epoch": 0.3310075187969925, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2242, + "step": 10503 + }, + { + "epoch": 0.33106766917293234, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 10504 + }, + { + "epoch": 0.33112781954887216, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 10505 + }, + { + "epoch": 0.33118796992481203, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 10506 + }, + { + "epoch": 0.3312481203007519, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 10507 + }, + { + "epoch": 0.3313082706766917, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1915, + "step": 10508 + }, + { + "epoch": 0.3313684210526316, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 10509 + }, + { + "epoch": 0.3314285714285714, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2775, + "step": 10510 + }, + { + "epoch": 0.3314887218045113, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 10511 + }, + { + "epoch": 0.33154887218045115, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2525, + "step": 10512 + }, + { + "epoch": 0.33160902255639096, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 10513 + }, + { + "epoch": 0.33166917293233084, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 10514 + }, + { + "epoch": 0.33172932330827065, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 10515 + }, + { + "epoch": 0.3317894736842105, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 10516 + }, + { + "epoch": 0.3318496240601504, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 10517 + }, + { + "epoch": 0.3319097744360902, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1966, + "step": 10518 + }, + { + "epoch": 0.3319699248120301, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 10519 + }, + { + "epoch": 0.3320300751879699, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 10520 + }, + { + "epoch": 0.33209022556390977, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 10521 + }, + { + "epoch": 0.33215037593984964, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 10522 + }, + { + "epoch": 0.33221052631578946, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2687, + "step": 10523 + }, + { + "epoch": 0.3322706766917293, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 10524 + }, + { + "epoch": 0.3323308270676692, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2052, + "step": 10525 + }, + { + "epoch": 0.332390977443609, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 10526 + }, + { + "epoch": 0.3324511278195489, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 10527 + }, + { + "epoch": 0.3325112781954887, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1923, + "step": 10528 + }, + { + "epoch": 0.3325714285714286, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2213, + "step": 10529 + }, + { + "epoch": 0.33263157894736844, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 10530 + }, + { + "epoch": 0.33269172932330826, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 10531 + }, + { + "epoch": 0.33275187969924813, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 10532 + }, + { + "epoch": 0.33281203007518795, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1933, + "step": 10533 + }, + { + "epoch": 0.3328721804511278, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 10534 + }, + { + "epoch": 0.3329323308270677, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 10535 + }, + { + "epoch": 0.3329924812030075, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 10536 + }, + { + "epoch": 0.3330526315789474, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2691, + "step": 10537 + }, + { + "epoch": 0.3331127819548872, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 10538 + }, + { + "epoch": 0.33317293233082707, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 10539 + }, + { + "epoch": 0.33323308270676694, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2686, + "step": 10540 + }, + { + "epoch": 0.33329323308270675, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 10541 + }, + { + "epoch": 0.3333533834586466, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2366, + "step": 10542 + }, + { + "epoch": 0.33341353383458644, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 10543 + }, + { + "epoch": 0.3334736842105263, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 10544 + }, + { + "epoch": 0.3335338345864662, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 10545 + }, + { + "epoch": 0.333593984962406, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 10546 + }, + { + "epoch": 0.33365413533834587, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.2727, + "step": 10547 + }, + { + "epoch": 0.33371428571428574, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 10548 + }, + { + "epoch": 0.33377443609022556, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2092, + "step": 10549 + }, + { + "epoch": 0.33383458646616543, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2254, + "step": 10550 + }, + { + "epoch": 0.33389473684210524, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 10551 + }, + { + "epoch": 0.3339548872180451, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 10552 + }, + { + "epoch": 0.334015037593985, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.3143, + "step": 10553 + }, + { + "epoch": 0.3340751879699248, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 10554 + }, + { + "epoch": 0.3341353383458647, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2151, + "step": 10555 + }, + { + "epoch": 0.3341954887218045, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2536, + "step": 10556 + }, + { + "epoch": 0.33425563909774436, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2056, + "step": 10557 + }, + { + "epoch": 0.33431578947368423, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2726, + "step": 10558 + }, + { + "epoch": 0.33437593984962405, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 10559 + }, + { + "epoch": 0.3344360902255639, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 10560 + }, + { + "epoch": 0.33449624060150374, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2075, + "step": 10561 + }, + { + "epoch": 0.3345563909774436, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 10562 + }, + { + "epoch": 0.3346165413533835, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2369, + "step": 10563 + }, + { + "epoch": 0.3346766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 10564 + }, + { + "epoch": 0.33473684210526317, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2552, + "step": 10565 + }, + { + "epoch": 0.334796992481203, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1271, + "step": 10566 + }, + { + "epoch": 0.33485714285714285, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2484, + "step": 10567 + }, + { + "epoch": 0.3349172932330827, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2042, + "step": 10568 + }, + { + "epoch": 0.33497744360902254, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 10569 + }, + { + "epoch": 0.3350375939849624, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.2211, + "step": 10570 + }, + { + "epoch": 0.3350977443609023, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.3053, + "step": 10571 + }, + { + "epoch": 0.3351578947368421, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 10572 + }, + { + "epoch": 0.33521804511278197, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 10573 + }, + { + "epoch": 0.3352781954887218, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 10574 + }, + { + "epoch": 0.33533834586466166, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 10575 + }, + { + "epoch": 0.33539849624060153, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 10576 + }, + { + "epoch": 0.33545864661654134, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2001, + "step": 10577 + }, + { + "epoch": 0.3355187969924812, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2331, + "step": 10578 + }, + { + "epoch": 0.33557894736842103, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2299, + "step": 10579 + }, + { + "epoch": 0.3356390977443609, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 10580 + }, + { + "epoch": 0.3356992481203008, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 10581 + }, + { + "epoch": 0.3357593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 10582 + }, + { + "epoch": 0.33581954887218046, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1705, + "step": 10583 + }, + { + "epoch": 0.3358796992481203, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1651, + "step": 10584 + }, + { + "epoch": 0.33593984962406015, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 10585 + }, + { + "epoch": 0.336, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 10586 + }, + { + "epoch": 0.33606015037593984, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2203, + "step": 10587 + }, + { + "epoch": 0.3361203007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 10588 + }, + { + "epoch": 0.3361804511278195, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1841, + "step": 10589 + }, + { + "epoch": 0.3362406015037594, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 10590 + }, + { + "epoch": 0.33630075187969927, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 10591 + }, + { + "epoch": 0.3363609022556391, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1916, + "step": 10592 + }, + { + "epoch": 0.33642105263157895, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 10593 + }, + { + "epoch": 0.33648120300751877, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2415, + "step": 10594 + }, + { + "epoch": 0.33654135338345864, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 10595 + }, + { + "epoch": 0.3366015037593985, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 10596 + }, + { + "epoch": 0.33666165413533833, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 10597 + }, + { + "epoch": 0.3367218045112782, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 10598 + }, + { + "epoch": 0.33678195488721807, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 10599 + }, + { + "epoch": 0.3368421052631579, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2823, + "step": 10600 + }, + { + "epoch": 0.33690225563909776, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1879, + "step": 10601 + }, + { + "epoch": 0.3369624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 10602 + }, + { + "epoch": 0.33702255639097745, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 10603 + }, + { + "epoch": 0.3370827067669173, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 10604 + }, + { + "epoch": 0.33714285714285713, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 10605 + }, + { + "epoch": 0.337203007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 10606 + }, + { + "epoch": 0.3372631578947368, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 10607 + }, + { + "epoch": 0.3373233082706767, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2033, + "step": 10608 + }, + { + "epoch": 0.33738345864661656, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 10609 + }, + { + "epoch": 0.3374436090225564, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 10610 + }, + { + "epoch": 0.33750375939849625, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 10611 + }, + { + "epoch": 0.33756390977443607, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3148, + "step": 10612 + }, + { + "epoch": 0.33762406015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2782, + "step": 10613 + }, + { + "epoch": 0.3376842105263158, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 10614 + }, + { + "epoch": 0.3377443609022556, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 10615 + }, + { + "epoch": 0.3378045112781955, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2645, + "step": 10616 + }, + { + "epoch": 0.3378646616541353, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 10617 + }, + { + "epoch": 0.3379248120300752, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 10618 + }, + { + "epoch": 0.33798496240601505, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2036, + "step": 10619 + }, + { + "epoch": 0.33804511278195487, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2037, + "step": 10620 + }, + { + "epoch": 0.33810526315789474, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 10621 + }, + { + "epoch": 0.3381654135338346, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1872, + "step": 10622 + }, + { + "epoch": 0.33822556390977443, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 10623 + }, + { + "epoch": 0.3382857142857143, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2631, + "step": 10624 + }, + { + "epoch": 0.3383458646616541, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 10625 + }, + { + "epoch": 0.338406015037594, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 10626 + }, + { + "epoch": 0.33846616541353386, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 10627 + }, + { + "epoch": 0.3385263157894737, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1971, + "step": 10628 + }, + { + "epoch": 0.33858646616541355, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 10629 + }, + { + "epoch": 0.33864661654135336, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 10630 + }, + { + "epoch": 0.33870676691729323, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 10631 + }, + { + "epoch": 0.3387669172932331, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1707, + "step": 10632 + }, + { + "epoch": 0.3388270676691729, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 10633 + }, + { + "epoch": 0.3388872180451128, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3211, + "step": 10634 + }, + { + "epoch": 0.3389473684210526, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 10635 + }, + { + "epoch": 0.3390075187969925, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 10636 + }, + { + "epoch": 0.33906766917293235, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 10637 + }, + { + "epoch": 0.33912781954887217, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 10638 + }, + { + "epoch": 0.33918796992481204, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2422, + "step": 10639 + }, + { + "epoch": 0.33924812030075185, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 10640 + }, + { + "epoch": 0.3393082706766917, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 10641 + }, + { + "epoch": 0.3393684210526316, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 10642 + }, + { + "epoch": 0.3394285714285714, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 10643 + }, + { + "epoch": 0.3394887218045113, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 10644 + }, + { + "epoch": 0.3395488721804511, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 10645 + }, + { + "epoch": 0.33960902255639097, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3489, + "step": 10646 + }, + { + "epoch": 0.33966917293233084, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 10647 + }, + { + "epoch": 0.33972932330827066, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2482, + "step": 10648 + }, + { + "epoch": 0.33978947368421053, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 10649 + }, + { + "epoch": 0.3398496240601504, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 10650 + }, + { + "epoch": 0.3399097744360902, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 10651 + }, + { + "epoch": 0.3399699248120301, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 10652 + }, + { + "epoch": 0.3400300751879699, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 10653 + }, + { + "epoch": 0.3400902255639098, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 10654 + }, + { + "epoch": 0.34015037593984965, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1719, + "step": 10655 + }, + { + "epoch": 0.34021052631578946, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.22, + "step": 10656 + }, + { + "epoch": 0.34027067669172933, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 10657 + }, + { + "epoch": 0.34033082706766915, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 10658 + }, + { + "epoch": 0.340390977443609, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 10659 + }, + { + "epoch": 0.3404511278195489, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2616, + "step": 10660 + }, + { + "epoch": 0.3405112781954887, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 10661 + }, + { + "epoch": 0.3405714285714286, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1494, + "step": 10662 + }, + { + "epoch": 0.3406315789473684, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 10663 + }, + { + "epoch": 0.34069172932330827, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2501, + "step": 10664 + }, + { + "epoch": 0.34075187969924814, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1915, + "step": 10665 + }, + { + "epoch": 0.34081203007518796, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2757, + "step": 10666 + }, + { + "epoch": 0.3408721804511278, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 10667 + }, + { + "epoch": 0.34093233082706764, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2317, + "step": 10668 + }, + { + "epoch": 0.3409924812030075, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2369, + "step": 10669 + }, + { + "epoch": 0.3410526315789474, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.2357, + "step": 10670 + }, + { + "epoch": 0.3411127819548872, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 10671 + }, + { + "epoch": 0.3411729323308271, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 10672 + }, + { + "epoch": 0.34123308270676694, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1604, + "step": 10673 + }, + { + "epoch": 0.34129323308270676, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2623, + "step": 10674 + }, + { + "epoch": 0.34135338345864663, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 10675 + }, + { + "epoch": 0.34141353383458645, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2544, + "step": 10676 + }, + { + "epoch": 0.3414736842105263, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 10677 + }, + { + "epoch": 0.3415338345864662, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 10678 + }, + { + "epoch": 0.341593984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2405, + "step": 10679 + }, + { + "epoch": 0.3416541353383459, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 10680 + }, + { + "epoch": 0.3417142857142857, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2591, + "step": 10681 + }, + { + "epoch": 0.34177443609022556, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2346, + "step": 10682 + }, + { + "epoch": 0.34183458646616544, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2656, + "step": 10683 + }, + { + "epoch": 0.34189473684210525, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2815, + "step": 10684 + }, + { + "epoch": 0.3419548872180451, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 10685 + }, + { + "epoch": 0.34201503759398494, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1464, + "step": 10686 + }, + { + "epoch": 0.3420751879699248, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2634, + "step": 10687 + }, + { + "epoch": 0.3421353383458647, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2021, + "step": 10688 + }, + { + "epoch": 0.3421954887218045, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 10689 + }, + { + "epoch": 0.34225563909774437, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2399, + "step": 10690 + }, + { + "epoch": 0.3423157894736842, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 10691 + }, + { + "epoch": 0.34237593984962406, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2275, + "step": 10692 + }, + { + "epoch": 0.3424360902255639, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.216, + "step": 10693 + }, + { + "epoch": 0.34249624060150374, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 10694 + }, + { + "epoch": 0.3425563909774436, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2309, + "step": 10695 + }, + { + "epoch": 0.3426165413533835, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 10696 + }, + { + "epoch": 0.3426766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 10697 + }, + { + "epoch": 0.3427368421052632, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 10698 + }, + { + "epoch": 0.342796992481203, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 10699 + }, + { + "epoch": 0.34285714285714286, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 10700 + }, + { + "epoch": 0.34291729323308273, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1938, + "step": 10701 + }, + { + "epoch": 0.34297744360902255, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2224, + "step": 10702 + }, + { + "epoch": 0.3430375939849624, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 10703 + }, + { + "epoch": 0.34309774436090223, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.3014, + "step": 10704 + }, + { + "epoch": 0.3431578947368421, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1983, + "step": 10705 + }, + { + "epoch": 0.343218045112782, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 10706 + }, + { + "epoch": 0.3432781954887218, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 10707 + }, + { + "epoch": 0.34333834586466166, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 10708 + }, + { + "epoch": 0.3433984962406015, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1917, + "step": 10709 + }, + { + "epoch": 0.34345864661654135, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 10710 + }, + { + "epoch": 0.3435187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 10711 + }, + { + "epoch": 0.34357894736842104, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2212, + "step": 10712 + }, + { + "epoch": 0.3436390977443609, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 10713 + }, + { + "epoch": 0.3436992481203007, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2735, + "step": 10714 + }, + { + "epoch": 0.3437593984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 10715 + }, + { + "epoch": 0.34381954887218047, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 10716 + }, + { + "epoch": 0.3438796992481203, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 10717 + }, + { + "epoch": 0.34393984962406016, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.3037, + "step": 10718 + }, + { + "epoch": 0.344, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 10719 + }, + { + "epoch": 0.34406015037593984, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 10720 + }, + { + "epoch": 0.3441203007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 10721 + }, + { + "epoch": 0.34418045112781953, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1928, + "step": 10722 + }, + { + "epoch": 0.3442406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 10723 + }, + { + "epoch": 0.3443007518796993, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 10724 + }, + { + "epoch": 0.3443609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2101, + "step": 10725 + }, + { + "epoch": 0.34442105263157896, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 10726 + }, + { + "epoch": 0.3444812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 10727 + }, + { + "epoch": 0.34454135338345865, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2841, + "step": 10728 + }, + { + "epoch": 0.3446015037593985, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 10729 + }, + { + "epoch": 0.34466165413533834, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 10730 + }, + { + "epoch": 0.3447218045112782, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 10731 + }, + { + "epoch": 0.344781954887218, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 10732 + }, + { + "epoch": 0.3448421052631579, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1781, + "step": 10733 + }, + { + "epoch": 0.34490225563909777, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2399, + "step": 10734 + }, + { + "epoch": 0.3449624060150376, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 10735 + }, + { + "epoch": 0.34502255639097745, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 10736 + }, + { + "epoch": 0.34508270676691727, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 10737 + }, + { + "epoch": 0.34514285714285714, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 10738 + }, + { + "epoch": 0.345203007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2064, + "step": 10739 + }, + { + "epoch": 0.3452631578947368, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 10740 + }, + { + "epoch": 0.3453233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.25, + "step": 10741 + }, + { + "epoch": 0.3453834586466165, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2642, + "step": 10742 + }, + { + "epoch": 0.3454436090225564, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2034, + "step": 10743 + }, + { + "epoch": 0.34550375939849626, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 10744 + }, + { + "epoch": 0.3455639097744361, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 10745 + }, + { + "epoch": 0.34562406015037594, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 10746 + }, + { + "epoch": 0.3456842105263158, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2014, + "step": 10747 + }, + { + "epoch": 0.34574436090225563, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 10748 + }, + { + "epoch": 0.3458045112781955, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2182, + "step": 10749 + }, + { + "epoch": 0.3458646616541353, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 10750 + }, + { + "epoch": 0.3459248120300752, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 10751 + }, + { + "epoch": 0.34598496240601506, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 10752 + }, + { + "epoch": 0.3460451127819549, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2337, + "step": 10753 + }, + { + "epoch": 0.34610526315789475, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.199, + "step": 10754 + }, + { + "epoch": 0.34616541353383457, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1937, + "step": 10755 + }, + { + "epoch": 0.34622556390977444, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 10756 + }, + { + "epoch": 0.3462857142857143, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.302, + "step": 10757 + }, + { + "epoch": 0.3463458646616541, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2227, + "step": 10758 + }, + { + "epoch": 0.346406015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 10759 + }, + { + "epoch": 0.3464661654135338, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 10760 + }, + { + "epoch": 0.3465263157894737, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3044, + "step": 10761 + }, + { + "epoch": 0.34658646616541355, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.3031, + "step": 10762 + }, + { + "epoch": 0.34664661654135337, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 10763 + }, + { + "epoch": 0.34670676691729324, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 10764 + }, + { + "epoch": 0.34676691729323306, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2663, + "step": 10765 + }, + { + "epoch": 0.34682706766917293, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 10766 + }, + { + "epoch": 0.3468872180451128, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 10767 + }, + { + "epoch": 0.3469473684210526, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2184, + "step": 10768 + }, + { + "epoch": 0.3470075187969925, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 10769 + }, + { + "epoch": 0.34706766917293236, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 10770 + }, + { + "epoch": 0.3471278195488722, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 10771 + }, + { + "epoch": 0.34718796992481205, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1923, + "step": 10772 + }, + { + "epoch": 0.34724812030075186, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3681, + "step": 10773 + }, + { + "epoch": 0.34730827067669173, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1332, + "step": 10774 + }, + { + "epoch": 0.3473684210526316, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 10775 + }, + { + "epoch": 0.3474285714285714, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1662, + "step": 10776 + }, + { + "epoch": 0.3474887218045113, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 10777 + }, + { + "epoch": 0.3475488721804511, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 10778 + }, + { + "epoch": 0.347609022556391, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 10779 + }, + { + "epoch": 0.34766917293233085, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 10780 + }, + { + "epoch": 0.34772932330827067, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 10781 + }, + { + "epoch": 0.34778947368421054, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1911, + "step": 10782 + }, + { + "epoch": 0.34784962406015035, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2186, + "step": 10783 + }, + { + "epoch": 0.3479097744360902, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1831, + "step": 10784 + }, + { + "epoch": 0.3479699248120301, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1905, + "step": 10785 + }, + { + "epoch": 0.3480300751879699, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2254, + "step": 10786 + }, + { + "epoch": 0.3480902255639098, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 10787 + }, + { + "epoch": 0.3481503759398496, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 10788 + }, + { + "epoch": 0.34821052631578947, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3279, + "step": 10789 + }, + { + "epoch": 0.34827067669172934, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 10790 + }, + { + "epoch": 0.34833082706766916, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2634, + "step": 10791 + }, + { + "epoch": 0.34839097744360903, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 10792 + }, + { + "epoch": 0.34845112781954884, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 10793 + }, + { + "epoch": 0.3485112781954887, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 10794 + }, + { + "epoch": 0.3485714285714286, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 10795 + }, + { + "epoch": 0.3486315789473684, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 10796 + }, + { + "epoch": 0.3486917293233083, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 10797 + }, + { + "epoch": 0.34875187969924815, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1702, + "step": 10798 + }, + { + "epoch": 0.34881203007518796, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2468, + "step": 10799 + }, + { + "epoch": 0.34887218045112783, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 10800 + }, + { + "epoch": 0.34893233082706765, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2131, + "step": 10801 + }, + { + "epoch": 0.3489924812030075, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2341, + "step": 10802 + }, + { + "epoch": 0.3490526315789474, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 10803 + }, + { + "epoch": 0.3491127819548872, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 10804 + }, + { + "epoch": 0.3491729323308271, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2396, + "step": 10805 + }, + { + "epoch": 0.3492330827067669, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 10806 + }, + { + "epoch": 0.34929323308270677, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 10807 + }, + { + "epoch": 0.34935338345864664, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2844, + "step": 10808 + }, + { + "epoch": 0.34941353383458645, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 10809 + }, + { + "epoch": 0.3494736842105263, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 10810 + }, + { + "epoch": 0.34953383458646614, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1526, + "step": 10811 + }, + { + "epoch": 0.349593984962406, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 10812 + }, + { + "epoch": 0.3496541353383459, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2268, + "step": 10813 + }, + { + "epoch": 0.3497142857142857, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2047, + "step": 10814 + }, + { + "epoch": 0.34977443609022557, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 10815 + }, + { + "epoch": 0.3498345864661654, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2581, + "step": 10816 + }, + { + "epoch": 0.34989473684210526, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 10817 + }, + { + "epoch": 0.34995488721804513, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 10818 + }, + { + "epoch": 0.35001503759398495, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 10819 + }, + { + "epoch": 0.3500751879699248, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 10820 + }, + { + "epoch": 0.3501353383458647, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 10821 + }, + { + "epoch": 0.3501954887218045, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2603, + "step": 10822 + }, + { + "epoch": 0.3502556390977444, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2225, + "step": 10823 + }, + { + "epoch": 0.3503157894736842, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 10824 + }, + { + "epoch": 0.35037593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1833, + "step": 10825 + }, + { + "epoch": 0.35043609022556393, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 10826 + }, + { + "epoch": 0.35049624060150375, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 10827 + }, + { + "epoch": 0.3505563909774436, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 10828 + }, + { + "epoch": 0.35061654135338344, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 10829 + }, + { + "epoch": 0.3506766917293233, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 10830 + }, + { + "epoch": 0.3507368421052632, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 10831 + }, + { + "epoch": 0.350796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 10832 + }, + { + "epoch": 0.35085714285714287, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2294, + "step": 10833 + }, + { + "epoch": 0.3509172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2717, + "step": 10834 + }, + { + "epoch": 0.35097744360902255, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1751, + "step": 10835 + }, + { + "epoch": 0.3510375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1962, + "step": 10836 + }, + { + "epoch": 0.35109774436090224, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 10837 + }, + { + "epoch": 0.3511578947368421, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1722, + "step": 10838 + }, + { + "epoch": 0.35121804511278193, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 10839 + }, + { + "epoch": 0.3512781954887218, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1886, + "step": 10840 + }, + { + "epoch": 0.35133834586466167, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 10841 + }, + { + "epoch": 0.3513984962406015, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 10842 + }, + { + "epoch": 0.35145864661654136, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2296, + "step": 10843 + }, + { + "epoch": 0.35151879699248123, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 10844 + }, + { + "epoch": 0.35157894736842105, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 10845 + }, + { + "epoch": 0.3516390977443609, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1934, + "step": 10846 + }, + { + "epoch": 0.35169924812030073, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 10847 + }, + { + "epoch": 0.3517593984962406, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2117, + "step": 10848 + }, + { + "epoch": 0.3518195488721805, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.5486, + "step": 10849 + }, + { + "epoch": 0.3518796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 10850 + }, + { + "epoch": 0.35193984962406016, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 10851 + }, + { + "epoch": 0.352, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3736, + "step": 10852 + }, + { + "epoch": 0.35206015037593985, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 10853 + }, + { + "epoch": 0.3521203007518797, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2194, + "step": 10854 + }, + { + "epoch": 0.35218045112781954, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 10855 + }, + { + "epoch": 0.3522406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2243, + "step": 10856 + }, + { + "epoch": 0.3523007518796992, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2004, + "step": 10857 + }, + { + "epoch": 0.3523609022556391, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 10858 + }, + { + "epoch": 0.35242105263157897, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1637, + "step": 10859 + }, + { + "epoch": 0.3524812030075188, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2398, + "step": 10860 + }, + { + "epoch": 0.35254135338345866, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2175, + "step": 10861 + }, + { + "epoch": 0.35260150375939847, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 10862 + }, + { + "epoch": 0.35266165413533834, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 10863 + }, + { + "epoch": 0.3527218045112782, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 10864 + }, + { + "epoch": 0.35278195488721803, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1836, + "step": 10865 + }, + { + "epoch": 0.3528421052631579, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1995, + "step": 10866 + }, + { + "epoch": 0.3529022556390977, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1763, + "step": 10867 + }, + { + "epoch": 0.3529624060150376, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 10868 + }, + { + "epoch": 0.35302255639097746, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 10869 + }, + { + "epoch": 0.3530827067669173, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 10870 + }, + { + "epoch": 0.35314285714285715, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2365, + "step": 10871 + }, + { + "epoch": 0.353203007518797, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 10872 + }, + { + "epoch": 0.35326315789473683, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2398, + "step": 10873 + }, + { + "epoch": 0.3533233082706767, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 10874 + }, + { + "epoch": 0.3533834586466165, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 10875 + }, + { + "epoch": 0.3534436090225564, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 10876 + }, + { + "epoch": 0.35350375939849626, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 10877 + }, + { + "epoch": 0.3535639097744361, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2676, + "step": 10878 + }, + { + "epoch": 0.35362406015037595, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 10879 + }, + { + "epoch": 0.35368421052631577, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2559, + "step": 10880 + }, + { + "epoch": 0.35374436090225564, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.271, + "step": 10881 + }, + { + "epoch": 0.3538045112781955, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 10882 + }, + { + "epoch": 0.3538646616541353, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 10883 + }, + { + "epoch": 0.3539248120300752, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 10884 + }, + { + "epoch": 0.353984962406015, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2277, + "step": 10885 + }, + { + "epoch": 0.3540451127819549, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 10886 + }, + { + "epoch": 0.35410526315789476, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 10887 + }, + { + "epoch": 0.35416541353383457, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2083, + "step": 10888 + }, + { + "epoch": 0.35422556390977444, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2877, + "step": 10889 + }, + { + "epoch": 0.35428571428571426, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 10890 + }, + { + "epoch": 0.35434586466165413, + "grad_norm": 1.3671875, + "learning_rate": 3e-05, + "loss": 1.2702, + "step": 10891 + }, + { + "epoch": 0.354406015037594, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.312, + "step": 10892 + }, + { + "epoch": 0.3544661654135338, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 10893 + }, + { + "epoch": 0.3545263157894737, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1905, + "step": 10894 + }, + { + "epoch": 0.35458646616541356, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 10895 + }, + { + "epoch": 0.3546466165413534, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 10896 + }, + { + "epoch": 0.35470676691729325, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 10897 + }, + { + "epoch": 0.35476691729323306, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 10898 + }, + { + "epoch": 0.35482706766917294, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 10899 + }, + { + "epoch": 0.3548872180451128, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1936, + "step": 10900 + }, + { + "epoch": 0.3549473684210526, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1811, + "step": 10901 + }, + { + "epoch": 0.3550075187969925, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 10902 + }, + { + "epoch": 0.3550676691729323, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.3244, + "step": 10903 + }, + { + "epoch": 0.3551278195488722, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 10904 + }, + { + "epoch": 0.35518796992481205, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2156, + "step": 10905 + }, + { + "epoch": 0.35524812030075187, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 10906 + }, + { + "epoch": 0.35530827067669174, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 10907 + }, + { + "epoch": 0.35536842105263156, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 10908 + }, + { + "epoch": 0.3554285714285714, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1896, + "step": 10909 + }, + { + "epoch": 0.3554887218045113, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.3411, + "step": 10910 + }, + { + "epoch": 0.3555488721804511, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 10911 + }, + { + "epoch": 0.355609022556391, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1774, + "step": 10912 + }, + { + "epoch": 0.3556691729323308, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 10913 + }, + { + "epoch": 0.3557293233082707, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2438, + "step": 10914 + }, + { + "epoch": 0.35578947368421054, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 10915 + }, + { + "epoch": 0.35584962406015036, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 10916 + }, + { + "epoch": 0.35590977443609023, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2767, + "step": 10917 + }, + { + "epoch": 0.35596992481203005, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 10918 + }, + { + "epoch": 0.3560300751879699, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 10919 + }, + { + "epoch": 0.3560902255639098, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.2046, + "step": 10920 + }, + { + "epoch": 0.3561503759398496, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2315, + "step": 10921 + }, + { + "epoch": 0.3562105263157895, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2895, + "step": 10922 + }, + { + "epoch": 0.35627067669172935, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 10923 + }, + { + "epoch": 0.35633082706766916, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 10924 + }, + { + "epoch": 0.35639097744360904, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2409, + "step": 10925 + }, + { + "epoch": 0.35645112781954885, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 10926 + }, + { + "epoch": 0.3565112781954887, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2499, + "step": 10927 + }, + { + "epoch": 0.3565714285714286, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 10928 + }, + { + "epoch": 0.3566315789473684, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 10929 + }, + { + "epoch": 0.3566917293233083, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 10930 + }, + { + "epoch": 0.3567518796992481, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2551, + "step": 10931 + }, + { + "epoch": 0.35681203007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 10932 + }, + { + "epoch": 0.35687218045112784, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 10933 + }, + { + "epoch": 0.35693233082706766, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.2699, + "step": 10934 + }, + { + "epoch": 0.35699248120300753, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 10935 + }, + { + "epoch": 0.35705263157894734, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2648, + "step": 10936 + }, + { + "epoch": 0.3571127819548872, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 10937 + }, + { + "epoch": 0.3571729323308271, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 10938 + }, + { + "epoch": 0.3572330827067669, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 10939 + }, + { + "epoch": 0.3572932330827068, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 10940 + }, + { + "epoch": 0.3573533834586466, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2598, + "step": 10941 + }, + { + "epoch": 0.35741353383458646, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2254, + "step": 10942 + }, + { + "epoch": 0.35747368421052633, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 10943 + }, + { + "epoch": 0.35753383458646615, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2858, + "step": 10944 + }, + { + "epoch": 0.357593984962406, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2221, + "step": 10945 + }, + { + "epoch": 0.3576541353383459, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3018, + "step": 10946 + }, + { + "epoch": 0.3577142857142857, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 10947 + }, + { + "epoch": 0.3577744360902256, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 10948 + }, + { + "epoch": 0.3578345864661654, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 10949 + }, + { + "epoch": 0.35789473684210527, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 10950 + }, + { + "epoch": 0.35795488721804514, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 10951 + }, + { + "epoch": 0.35801503759398495, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.3547, + "step": 10952 + }, + { + "epoch": 0.3580751879699248, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1923, + "step": 10953 + }, + { + "epoch": 0.35813533834586464, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1962, + "step": 10954 + }, + { + "epoch": 0.3581954887218045, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1859, + "step": 10955 + }, + { + "epoch": 0.3582556390977444, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 10956 + }, + { + "epoch": 0.3583157894736842, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.189, + "step": 10957 + }, + { + "epoch": 0.35837593984962407, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 10958 + }, + { + "epoch": 0.3584360902255639, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 10959 + }, + { + "epoch": 0.35849624060150376, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 10960 + }, + { + "epoch": 0.35855639097744363, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2874, + "step": 10961 + }, + { + "epoch": 0.35861654135338344, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 10962 + }, + { + "epoch": 0.3586766917293233, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 10963 + }, + { + "epoch": 0.35873684210526313, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1817, + "step": 10964 + }, + { + "epoch": 0.358796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2079, + "step": 10965 + }, + { + "epoch": 0.3588571428571429, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 10966 + }, + { + "epoch": 0.3589172932330827, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 10967 + }, + { + "epoch": 0.35897744360902256, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3017, + "step": 10968 + }, + { + "epoch": 0.35903759398496243, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 10969 + }, + { + "epoch": 0.35909774436090225, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1738, + "step": 10970 + }, + { + "epoch": 0.3591578947368421, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 10971 + }, + { + "epoch": 0.35921804511278194, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2464, + "step": 10972 + }, + { + "epoch": 0.3592781954887218, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2695, + "step": 10973 + }, + { + "epoch": 0.3593383458646617, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3246, + "step": 10974 + }, + { + "epoch": 0.3593984962406015, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 10975 + }, + { + "epoch": 0.35945864661654137, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2372, + "step": 10976 + }, + { + "epoch": 0.3595187969924812, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2124, + "step": 10977 + }, + { + "epoch": 0.35957894736842105, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2263, + "step": 10978 + }, + { + "epoch": 0.3596390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 10979 + }, + { + "epoch": 0.35969924812030074, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1719, + "step": 10980 + }, + { + "epoch": 0.3597593984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2316, + "step": 10981 + }, + { + "epoch": 0.35981954887218043, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1909, + "step": 10982 + }, + { + "epoch": 0.3598796992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 10983 + }, + { + "epoch": 0.35993984962406017, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 10984 + }, + { + "epoch": 0.36, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 10985 + }, + { + "epoch": 0.36006015037593986, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2527, + "step": 10986 + }, + { + "epoch": 0.3601203007518797, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1653, + "step": 10987 + }, + { + "epoch": 0.36018045112781955, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3576, + "step": 10988 + }, + { + "epoch": 0.3602406015037594, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2257, + "step": 10989 + }, + { + "epoch": 0.36030075187969923, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 10990 + }, + { + "epoch": 0.3603609022556391, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1845, + "step": 10991 + }, + { + "epoch": 0.3604210526315789, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2149, + "step": 10992 + }, + { + "epoch": 0.3604812030075188, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2332, + "step": 10993 + }, + { + "epoch": 0.36054135338345866, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 10994 + }, + { + "epoch": 0.3606015037593985, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.239, + "step": 10995 + }, + { + "epoch": 0.36066165413533835, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2348, + "step": 10996 + }, + { + "epoch": 0.3607218045112782, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 10997 + }, + { + "epoch": 0.36078195488721804, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 10998 + }, + { + "epoch": 0.3608421052631579, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 10999 + }, + { + "epoch": 0.3609022556390977, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2963, + "step": 11000 + }, + { + "epoch": 0.3609624060150376, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 11001 + }, + { + "epoch": 0.36102255639097747, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3848, + "step": 11002 + }, + { + "epoch": 0.3610827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 11003 + }, + { + "epoch": 0.36114285714285715, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 11004 + }, + { + "epoch": 0.36120300751879697, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 11005 + }, + { + "epoch": 0.36126315789473684, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.261, + "step": 11006 + }, + { + "epoch": 0.3613233082706767, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 11007 + }, + { + "epoch": 0.36138345864661653, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2032, + "step": 11008 + }, + { + "epoch": 0.3614436090225564, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 11009 + }, + { + "epoch": 0.3615037593984962, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2205, + "step": 11010 + }, + { + "epoch": 0.3615639097744361, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 11011 + }, + { + "epoch": 0.36162406015037596, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 11012 + }, + { + "epoch": 0.3616842105263158, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 11013 + }, + { + "epoch": 0.36174436090225565, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 11014 + }, + { + "epoch": 0.36180451127819546, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2531, + "step": 11015 + }, + { + "epoch": 0.36186466165413533, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2113, + "step": 11016 + }, + { + "epoch": 0.3619248120300752, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 11017 + }, + { + "epoch": 0.361984962406015, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 11018 + }, + { + "epoch": 0.3620451127819549, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 11019 + }, + { + "epoch": 0.36210526315789476, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 11020 + }, + { + "epoch": 0.3621654135338346, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 11021 + }, + { + "epoch": 0.36222556390977445, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 11022 + }, + { + "epoch": 0.36228571428571427, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1719, + "step": 11023 + }, + { + "epoch": 0.36234586466165414, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 11024 + }, + { + "epoch": 0.362406015037594, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.3797, + "step": 11025 + }, + { + "epoch": 0.3624661654135338, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2465, + "step": 11026 + }, + { + "epoch": 0.3625263157894737, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 11027 + }, + { + "epoch": 0.3625864661654135, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 11028 + }, + { + "epoch": 0.3626466165413534, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1956, + "step": 11029 + }, + { + "epoch": 0.36270676691729326, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 11030 + }, + { + "epoch": 0.36276691729323307, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 11031 + }, + { + "epoch": 0.36282706766917294, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 11032 + }, + { + "epoch": 0.36288721804511276, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 11033 + }, + { + "epoch": 0.36294736842105263, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 11034 + }, + { + "epoch": 0.3630075187969925, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 11035 + }, + { + "epoch": 0.3630676691729323, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 11036 + }, + { + "epoch": 0.3631278195488722, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 11037 + }, + { + "epoch": 0.363187969924812, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 11038 + }, + { + "epoch": 0.3632481203007519, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1998, + "step": 11039 + }, + { + "epoch": 0.36330827067669175, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 11040 + }, + { + "epoch": 0.36336842105263156, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 11041 + }, + { + "epoch": 0.36342857142857143, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 11042 + }, + { + "epoch": 0.3634887218045113, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 11043 + }, + { + "epoch": 0.3635488721804511, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 11044 + }, + { + "epoch": 0.363609022556391, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 11045 + }, + { + "epoch": 0.3636691729323308, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 11046 + }, + { + "epoch": 0.3637293233082707, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1667, + "step": 11047 + }, + { + "epoch": 0.36378947368421055, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 11048 + }, + { + "epoch": 0.36384962406015037, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 11049 + }, + { + "epoch": 0.36390977443609024, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2296, + "step": 11050 + }, + { + "epoch": 0.36396992481203005, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3187, + "step": 11051 + }, + { + "epoch": 0.3640300751879699, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 11052 + }, + { + "epoch": 0.3640902255639098, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2611, + "step": 11053 + }, + { + "epoch": 0.3641503759398496, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 11054 + }, + { + "epoch": 0.3642105263157895, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 11055 + }, + { + "epoch": 0.3642706766917293, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 11056 + }, + { + "epoch": 0.36433082706766917, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1917, + "step": 11057 + }, + { + "epoch": 0.36439097744360904, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 11058 + }, + { + "epoch": 0.36445112781954886, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2309, + "step": 11059 + }, + { + "epoch": 0.36451127819548873, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2144, + "step": 11060 + }, + { + "epoch": 0.36457142857142855, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3015, + "step": 11061 + }, + { + "epoch": 0.3646315789473684, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 11062 + }, + { + "epoch": 0.3646917293233083, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2964, + "step": 11063 + }, + { + "epoch": 0.3647518796992481, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2053, + "step": 11064 + }, + { + "epoch": 0.364812030075188, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2557, + "step": 11065 + }, + { + "epoch": 0.3648721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1954, + "step": 11066 + }, + { + "epoch": 0.36493233082706766, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 11067 + }, + { + "epoch": 0.36499248120300753, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2173, + "step": 11068 + }, + { + "epoch": 0.36505263157894735, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 11069 + }, + { + "epoch": 0.3651127819548872, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 11070 + }, + { + "epoch": 0.3651729323308271, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 11071 + }, + { + "epoch": 0.3652330827067669, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 11072 + }, + { + "epoch": 0.3652932330827068, + "grad_norm": 1.765625, + "learning_rate": 3e-05, + "loss": 1.2545, + "step": 11073 + }, + { + "epoch": 0.3653533834586466, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 11074 + }, + { + "epoch": 0.36541353383458647, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 11075 + }, + { + "epoch": 0.36547368421052634, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1512, + "step": 11076 + }, + { + "epoch": 0.36553383458646616, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 11077 + }, + { + "epoch": 0.365593984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 11078 + }, + { + "epoch": 0.36565413533834584, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2224, + "step": 11079 + }, + { + "epoch": 0.3657142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 11080 + }, + { + "epoch": 0.3657744360902256, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1923, + "step": 11081 + }, + { + "epoch": 0.3658345864661654, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2309, + "step": 11082 + }, + { + "epoch": 0.3658947368421053, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 11083 + }, + { + "epoch": 0.3659548872180451, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 11084 + }, + { + "epoch": 0.36601503759398496, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 11085 + }, + { + "epoch": 0.36607518796992483, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2717, + "step": 11086 + }, + { + "epoch": 0.36613533834586465, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1998, + "step": 11087 + }, + { + "epoch": 0.3661954887218045, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1651, + "step": 11088 + }, + { + "epoch": 0.36625563909774433, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 11089 + }, + { + "epoch": 0.3663157894736842, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 11090 + }, + { + "epoch": 0.3663759398496241, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 11091 + }, + { + "epoch": 0.3664360902255639, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 11092 + }, + { + "epoch": 0.36649624060150376, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2245, + "step": 11093 + }, + { + "epoch": 0.36655639097744364, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 11094 + }, + { + "epoch": 0.36661654135338345, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1766, + "step": 11095 + }, + { + "epoch": 0.3666766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 11096 + }, + { + "epoch": 0.36673684210526314, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 11097 + }, + { + "epoch": 0.366796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 11098 + }, + { + "epoch": 0.3668571428571429, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 11099 + }, + { + "epoch": 0.3669172932330827, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 11100 + }, + { + "epoch": 0.36697744360902257, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 11101 + }, + { + "epoch": 0.3670375939849624, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1868, + "step": 11102 + }, + { + "epoch": 0.36709774436090226, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 11103 + }, + { + "epoch": 0.3671578947368421, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 11104 + }, + { + "epoch": 0.36721804511278194, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1918, + "step": 11105 + }, + { + "epoch": 0.3672781954887218, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 11106 + }, + { + "epoch": 0.36733834586466163, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 11107 + }, + { + "epoch": 0.3673984962406015, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2875, + "step": 11108 + }, + { + "epoch": 0.3674586466165414, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2654, + "step": 11109 + }, + { + "epoch": 0.3675187969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.195, + "step": 11110 + }, + { + "epoch": 0.36757894736842106, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 11111 + }, + { + "epoch": 0.3676390977443609, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.213, + "step": 11112 + }, + { + "epoch": 0.36769924812030075, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2196, + "step": 11113 + }, + { + "epoch": 0.3677593984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 11114 + }, + { + "epoch": 0.36781954887218044, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 11115 + }, + { + "epoch": 0.3678796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 11116 + }, + { + "epoch": 0.3679398496240602, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 11117 + }, + { + "epoch": 0.368, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.3055, + "step": 11118 + }, + { + "epoch": 0.36806015037593987, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 11119 + }, + { + "epoch": 0.3681203007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2375, + "step": 11120 + }, + { + "epoch": 0.36818045112781955, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 11121 + }, + { + "epoch": 0.3682406015037594, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 11122 + }, + { + "epoch": 0.36830075187969924, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 11123 + }, + { + "epoch": 0.3683609022556391, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.2092, + "step": 11124 + }, + { + "epoch": 0.3684210526315789, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.244, + "step": 11125 + }, + { + "epoch": 0.3684812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 11126 + }, + { + "epoch": 0.36854135338345867, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 11127 + }, + { + "epoch": 0.3686015037593985, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 11128 + }, + { + "epoch": 0.36866165413533836, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2772, + "step": 11129 + }, + { + "epoch": 0.3687218045112782, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.199, + "step": 11130 + }, + { + "epoch": 0.36878195488721804, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 11131 + }, + { + "epoch": 0.3688421052631579, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 11132 + }, + { + "epoch": 0.36890225563909773, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 11133 + }, + { + "epoch": 0.3689624060150376, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2214, + "step": 11134 + }, + { + "epoch": 0.3690225563909774, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2233, + "step": 11135 + }, + { + "epoch": 0.3690827067669173, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.2776, + "step": 11136 + }, + { + "epoch": 0.36914285714285716, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3129, + "step": 11137 + }, + { + "epoch": 0.369203007518797, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 11138 + }, + { + "epoch": 0.36926315789473685, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 11139 + }, + { + "epoch": 0.36932330827067666, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.3103, + "step": 11140 + }, + { + "epoch": 0.36938345864661654, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 11141 + }, + { + "epoch": 0.3694436090225564, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 11142 + }, + { + "epoch": 0.3695037593984962, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2596, + "step": 11143 + }, + { + "epoch": 0.3695639097744361, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2462, + "step": 11144 + }, + { + "epoch": 0.36962406015037597, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 11145 + }, + { + "epoch": 0.3696842105263158, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1966, + "step": 11146 + }, + { + "epoch": 0.36974436090225565, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 11147 + }, + { + "epoch": 0.36980451127819547, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 11148 + }, + { + "epoch": 0.36986466165413534, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 11149 + }, + { + "epoch": 0.3699248120300752, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2115, + "step": 11150 + }, + { + "epoch": 0.369984962406015, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2033, + "step": 11151 + }, + { + "epoch": 0.3700451127819549, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 11152 + }, + { + "epoch": 0.3701052631578947, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 11153 + }, + { + "epoch": 0.3701654135338346, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2036, + "step": 11154 + }, + { + "epoch": 0.37022556390977446, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2979, + "step": 11155 + }, + { + "epoch": 0.3702857142857143, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 11156 + }, + { + "epoch": 0.37034586466165414, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 11157 + }, + { + "epoch": 0.37040601503759396, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 11158 + }, + { + "epoch": 0.37046616541353383, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 11159 + }, + { + "epoch": 0.3705263157894737, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 11160 + }, + { + "epoch": 0.3705864661654135, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 11161 + }, + { + "epoch": 0.3706466165413534, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 11162 + }, + { + "epoch": 0.3707067669172932, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2104, + "step": 11163 + }, + { + "epoch": 0.3707669172932331, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 11164 + }, + { + "epoch": 0.37082706766917295, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 11165 + }, + { + "epoch": 0.37088721804511277, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1731, + "step": 11166 + }, + { + "epoch": 0.37094736842105264, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 11167 + }, + { + "epoch": 0.3710075187969925, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 11168 + }, + { + "epoch": 0.3710676691729323, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1545, + "step": 11169 + }, + { + "epoch": 0.3711278195488722, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1962, + "step": 11170 + }, + { + "epoch": 0.371187969924812, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2545, + "step": 11171 + }, + { + "epoch": 0.3712481203007519, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 11172 + }, + { + "epoch": 0.37130827067669175, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 11173 + }, + { + "epoch": 0.37136842105263157, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 11174 + }, + { + "epoch": 0.37142857142857144, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 11175 + }, + { + "epoch": 0.37148872180451126, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 11176 + }, + { + "epoch": 0.37154887218045113, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 11177 + }, + { + "epoch": 0.371609022556391, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 11178 + }, + { + "epoch": 0.3716691729323308, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2634, + "step": 11179 + }, + { + "epoch": 0.3717293233082707, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 11180 + }, + { + "epoch": 0.3717894736842105, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 11181 + }, + { + "epoch": 0.3718496240601504, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 11182 + }, + { + "epoch": 0.37190977443609025, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 11183 + }, + { + "epoch": 0.37196992481203006, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3007, + "step": 11184 + }, + { + "epoch": 0.37203007518796993, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 11185 + }, + { + "epoch": 0.37209022556390975, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 11186 + }, + { + "epoch": 0.3721503759398496, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2183, + "step": 11187 + }, + { + "epoch": 0.3722105263157895, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 11188 + }, + { + "epoch": 0.3722706766917293, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 11189 + }, + { + "epoch": 0.3723308270676692, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2376, + "step": 11190 + }, + { + "epoch": 0.37239097744360905, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1464, + "step": 11191 + }, + { + "epoch": 0.37245112781954887, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1998, + "step": 11192 + }, + { + "epoch": 0.37251127819548874, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 11193 + }, + { + "epoch": 0.37257142857142855, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 11194 + }, + { + "epoch": 0.3726315789473684, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 11195 + }, + { + "epoch": 0.3726917293233083, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 11196 + }, + { + "epoch": 0.3727518796992481, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 11197 + }, + { + "epoch": 0.372812030075188, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.3185, + "step": 11198 + }, + { + "epoch": 0.3728721804511278, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1651, + "step": 11199 + }, + { + "epoch": 0.37293233082706767, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 11200 + }, + { + "epoch": 0.37299248120300754, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 11201 + }, + { + "epoch": 0.37305263157894736, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 11202 + }, + { + "epoch": 0.37311278195488723, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2412, + "step": 11203 + }, + { + "epoch": 0.37317293233082705, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 11204 + }, + { + "epoch": 0.3732330827067669, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 11205 + }, + { + "epoch": 0.3732932330827068, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1979, + "step": 11206 + }, + { + "epoch": 0.3733533834586466, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2702, + "step": 11207 + }, + { + "epoch": 0.3734135338345865, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2173, + "step": 11208 + }, + { + "epoch": 0.3734736842105263, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 11209 + }, + { + "epoch": 0.37353383458646616, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 11210 + }, + { + "epoch": 0.37359398496240603, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 11211 + }, + { + "epoch": 0.37365413533834585, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2372, + "step": 11212 + }, + { + "epoch": 0.3737142857142857, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2053, + "step": 11213 + }, + { + "epoch": 0.37377443609022554, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2681, + "step": 11214 + }, + { + "epoch": 0.3738345864661654, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.214, + "step": 11215 + }, + { + "epoch": 0.3738947368421053, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1597, + "step": 11216 + }, + { + "epoch": 0.3739548872180451, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3269, + "step": 11217 + }, + { + "epoch": 0.37401503759398497, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 11218 + }, + { + "epoch": 0.37407518796992484, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 11219 + }, + { + "epoch": 0.37413533834586465, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 11220 + }, + { + "epoch": 0.3741954887218045, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1643, + "step": 11221 + }, + { + "epoch": 0.37425563909774434, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 11222 + }, + { + "epoch": 0.3743157894736842, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 11223 + }, + { + "epoch": 0.3743759398496241, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 11224 + }, + { + "epoch": 0.3744360902255639, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1778, + "step": 11225 + }, + { + "epoch": 0.37449624060150377, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2065, + "step": 11226 + }, + { + "epoch": 0.3745563909774436, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 11227 + }, + { + "epoch": 0.37461654135338346, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2174, + "step": 11228 + }, + { + "epoch": 0.37467669172932333, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 11229 + }, + { + "epoch": 0.37473684210526315, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2182, + "step": 11230 + }, + { + "epoch": 0.374796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 11231 + }, + { + "epoch": 0.37485714285714283, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 11232 + }, + { + "epoch": 0.3749172932330827, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 11233 + }, + { + "epoch": 0.3749774436090226, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 11234 + }, + { + "epoch": 0.3750375939849624, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 11235 + }, + { + "epoch": 0.37509774436090226, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 11236 + }, + { + "epoch": 0.3751578947368421, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 11237 + }, + { + "epoch": 0.37521804511278195, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 11238 + }, + { + "epoch": 0.3752781954887218, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1955, + "step": 11239 + }, + { + "epoch": 0.37533834586466164, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1936, + "step": 11240 + }, + { + "epoch": 0.3753984962406015, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 11241 + }, + { + "epoch": 0.3754586466165414, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1671, + "step": 11242 + }, + { + "epoch": 0.3755187969924812, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2424, + "step": 11243 + }, + { + "epoch": 0.37557894736842107, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2604, + "step": 11244 + }, + { + "epoch": 0.3756390977443609, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 11245 + }, + { + "epoch": 0.37569924812030076, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 11246 + }, + { + "epoch": 0.3757593984962406, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2857, + "step": 11247 + }, + { + "epoch": 0.37581954887218044, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 11248 + }, + { + "epoch": 0.3758796992481203, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 11249 + }, + { + "epoch": 0.37593984962406013, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 11250 + }, + { + "epoch": 0.376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2619, + "step": 11251 + }, + { + "epoch": 0.3760601503759399, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 11252 + }, + { + "epoch": 0.3761203007518797, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 11253 + }, + { + "epoch": 0.37618045112781956, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 11254 + }, + { + "epoch": 0.3762406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1692, + "step": 11255 + }, + { + "epoch": 0.37630075187969925, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 11256 + }, + { + "epoch": 0.3763609022556391, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2468, + "step": 11257 + }, + { + "epoch": 0.37642105263157893, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2932, + "step": 11258 + }, + { + "epoch": 0.3764812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.379, + "step": 11259 + }, + { + "epoch": 0.3765413533834586, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1746, + "step": 11260 + }, + { + "epoch": 0.3766015037593985, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2481, + "step": 11261 + }, + { + "epoch": 0.37666165413533836, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 11262 + }, + { + "epoch": 0.3767218045112782, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 11263 + }, + { + "epoch": 0.37678195488721805, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2696, + "step": 11264 + }, + { + "epoch": 0.37684210526315787, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1073, + "step": 11265 + }, + { + "epoch": 0.37690225563909774, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 11266 + }, + { + "epoch": 0.3769624060150376, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 11267 + }, + { + "epoch": 0.3770225563909774, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 11268 + }, + { + "epoch": 0.3770827067669173, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2328, + "step": 11269 + }, + { + "epoch": 0.37714285714285717, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 11270 + }, + { + "epoch": 0.377203007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 11271 + }, + { + "epoch": 0.37726315789473686, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 11272 + }, + { + "epoch": 0.37732330827067667, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.3641, + "step": 11273 + }, + { + "epoch": 0.37738345864661654, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 11274 + }, + { + "epoch": 0.3774436090225564, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2101, + "step": 11275 + }, + { + "epoch": 0.37750375939849623, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2102, + "step": 11276 + }, + { + "epoch": 0.3775639097744361, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 11277 + }, + { + "epoch": 0.3776240601503759, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 11278 + }, + { + "epoch": 0.3776842105263158, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2527, + "step": 11279 + }, + { + "epoch": 0.37774436090225566, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 11280 + }, + { + "epoch": 0.3778045112781955, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 11281 + }, + { + "epoch": 0.37786466165413535, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 11282 + }, + { + "epoch": 0.37792481203007516, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1727, + "step": 11283 + }, + { + "epoch": 0.37798496240601503, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2001, + "step": 11284 + }, + { + "epoch": 0.3780451127819549, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 11285 + }, + { + "epoch": 0.3781052631578947, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 11286 + }, + { + "epoch": 0.3781654135338346, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2471, + "step": 11287 + }, + { + "epoch": 0.3782255639097744, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 11288 + }, + { + "epoch": 0.3782857142857143, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 11289 + }, + { + "epoch": 0.37834586466165415, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2264, + "step": 11290 + }, + { + "epoch": 0.37840601503759397, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 11291 + }, + { + "epoch": 0.37846616541353384, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 11292 + }, + { + "epoch": 0.3785263157894737, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.1552, + "step": 11293 + }, + { + "epoch": 0.3785864661654135, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 11294 + }, + { + "epoch": 0.3786466165413534, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 11295 + }, + { + "epoch": 0.3787067669172932, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 11296 + }, + { + "epoch": 0.3787669172932331, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 11297 + }, + { + "epoch": 0.37882706766917296, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 11298 + }, + { + "epoch": 0.3788872180451128, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2033, + "step": 11299 + }, + { + "epoch": 0.37894736842105264, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2166, + "step": 11300 + }, + { + "epoch": 0.37900751879699246, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2019, + "step": 11301 + }, + { + "epoch": 0.37906766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2435, + "step": 11302 + }, + { + "epoch": 0.3791278195488722, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 11303 + }, + { + "epoch": 0.379187969924812, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 11304 + }, + { + "epoch": 0.3792481203007519, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 11305 + }, + { + "epoch": 0.3793082706766917, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1604, + "step": 11306 + }, + { + "epoch": 0.3793684210526316, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2044, + "step": 11307 + }, + { + "epoch": 0.37942857142857145, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 11308 + }, + { + "epoch": 0.37948872180451126, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 11309 + }, + { + "epoch": 0.37954887218045114, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2655, + "step": 11310 + }, + { + "epoch": 0.37960902255639095, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 11311 + }, + { + "epoch": 0.3796691729323308, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 11312 + }, + { + "epoch": 0.3797293233082707, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 11313 + }, + { + "epoch": 0.3797894736842105, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 11314 + }, + { + "epoch": 0.3798496240601504, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2131, + "step": 11315 + }, + { + "epoch": 0.37990977443609025, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 11316 + }, + { + "epoch": 0.37996992481203007, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 11317 + }, + { + "epoch": 0.38003007518796994, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 11318 + }, + { + "epoch": 0.38009022556390976, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 11319 + }, + { + "epoch": 0.3801503759398496, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 11320 + }, + { + "epoch": 0.3802105263157895, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 11321 + }, + { + "epoch": 0.3802706766917293, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 11322 + }, + { + "epoch": 0.3803308270676692, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 11323 + }, + { + "epoch": 0.380390977443609, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.3403, + "step": 11324 + }, + { + "epoch": 0.3804511278195489, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2393, + "step": 11325 + }, + { + "epoch": 0.38051127819548874, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 11326 + }, + { + "epoch": 0.38057142857142856, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 11327 + }, + { + "epoch": 0.38063157894736843, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 11328 + }, + { + "epoch": 0.38069172932330825, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 11329 + }, + { + "epoch": 0.3807518796992481, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 11330 + }, + { + "epoch": 0.380812030075188, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 11331 + }, + { + "epoch": 0.3808721804511278, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 11332 + }, + { + "epoch": 0.3809323308270677, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 11333 + }, + { + "epoch": 0.3809924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2774, + "step": 11334 + }, + { + "epoch": 0.38105263157894737, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 11335 + }, + { + "epoch": 0.38111278195488724, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 11336 + }, + { + "epoch": 0.38117293233082705, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 11337 + }, + { + "epoch": 0.3812330827067669, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 11338 + }, + { + "epoch": 0.38129323308270674, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 11339 + }, + { + "epoch": 0.3813533834586466, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1436, + "step": 11340 + }, + { + "epoch": 0.3814135338345865, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 11341 + }, + { + "epoch": 0.3814736842105263, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 11342 + }, + { + "epoch": 0.38153383458646617, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 11343 + }, + { + "epoch": 0.38159398496240604, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 11344 + }, + { + "epoch": 0.38165413533834586, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2022, + "step": 11345 + }, + { + "epoch": 0.38171428571428573, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 11346 + }, + { + "epoch": 0.38177443609022554, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 11347 + }, + { + "epoch": 0.3818345864661654, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.2279, + "step": 11348 + }, + { + "epoch": 0.3818947368421053, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 11349 + }, + { + "epoch": 0.3819548872180451, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 11350 + }, + { + "epoch": 0.382015037593985, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 11351 + }, + { + "epoch": 0.3820751879699248, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 11352 + }, + { + "epoch": 0.38213533834586466, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1543, + "step": 11353 + }, + { + "epoch": 0.38219548872180453, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1882, + "step": 11354 + }, + { + "epoch": 0.38225563909774435, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 11355 + }, + { + "epoch": 0.3823157894736842, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 11356 + }, + { + "epoch": 0.38237593984962404, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 11357 + }, + { + "epoch": 0.3824360902255639, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 11358 + }, + { + "epoch": 0.3824962406015038, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 11359 + }, + { + "epoch": 0.3825563909774436, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 11360 + }, + { + "epoch": 0.38261654135338347, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2743, + "step": 11361 + }, + { + "epoch": 0.3826766917293233, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.289, + "step": 11362 + }, + { + "epoch": 0.38273684210526315, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 11363 + }, + { + "epoch": 0.382796992481203, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.3335, + "step": 11364 + }, + { + "epoch": 0.38285714285714284, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 11365 + }, + { + "epoch": 0.3829172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2052, + "step": 11366 + }, + { + "epoch": 0.3829774436090226, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1623, + "step": 11367 + }, + { + "epoch": 0.3830375939849624, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 11368 + }, + { + "epoch": 0.38309774436090227, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 11369 + }, + { + "epoch": 0.3831578947368421, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1465, + "step": 11370 + }, + { + "epoch": 0.38321804511278196, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 11371 + }, + { + "epoch": 0.38327819548872183, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 11372 + }, + { + "epoch": 0.38333834586466164, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1919, + "step": 11373 + }, + { + "epoch": 0.3833984962406015, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 11374 + }, + { + "epoch": 0.38345864661654133, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 11375 + }, + { + "epoch": 0.3835187969924812, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 11376 + }, + { + "epoch": 0.3835789473684211, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 11377 + }, + { + "epoch": 0.3836390977443609, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 11378 + }, + { + "epoch": 0.38369924812030076, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 11379 + }, + { + "epoch": 0.3837593984962406, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 11380 + }, + { + "epoch": 0.38381954887218045, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.245, + "step": 11381 + }, + { + "epoch": 0.3838796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2291, + "step": 11382 + }, + { + "epoch": 0.38393984962406014, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 11383 + }, + { + "epoch": 0.384, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 11384 + }, + { + "epoch": 0.3840601503759398, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 11385 + }, + { + "epoch": 0.3841203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 11386 + }, + { + "epoch": 0.38418045112781957, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 11387 + }, + { + "epoch": 0.3842406015037594, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 11388 + }, + { + "epoch": 0.38430075187969925, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2088, + "step": 11389 + }, + { + "epoch": 0.3843609022556391, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2006, + "step": 11390 + }, + { + "epoch": 0.38442105263157894, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1735, + "step": 11391 + }, + { + "epoch": 0.3844812030075188, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 11392 + }, + { + "epoch": 0.38454135338345863, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 11393 + }, + { + "epoch": 0.3846015037593985, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2413, + "step": 11394 + }, + { + "epoch": 0.38466165413533837, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 11395 + }, + { + "epoch": 0.3847218045112782, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 11396 + }, + { + "epoch": 0.38478195488721806, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 11397 + }, + { + "epoch": 0.3848421052631579, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 11398 + }, + { + "epoch": 0.38490225563909775, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 11399 + }, + { + "epoch": 0.3849624060150376, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 11400 + }, + { + "epoch": 0.38502255639097743, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 11401 + }, + { + "epoch": 0.3850827067669173, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 11402 + }, + { + "epoch": 0.3851428571428571, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 11403 + }, + { + "epoch": 0.385203007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 11404 + }, + { + "epoch": 0.38526315789473686, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 11405 + }, + { + "epoch": 0.3853233082706767, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2245, + "step": 11406 + }, + { + "epoch": 0.38538345864661655, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2512, + "step": 11407 + }, + { + "epoch": 0.38544360902255637, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 11408 + }, + { + "epoch": 0.38550375939849624, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 11409 + }, + { + "epoch": 0.3855639097744361, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2723, + "step": 11410 + }, + { + "epoch": 0.3856240601503759, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2294, + "step": 11411 + }, + { + "epoch": 0.3856842105263158, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 11412 + }, + { + "epoch": 0.3857443609022556, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 11413 + }, + { + "epoch": 0.3858045112781955, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 11414 + }, + { + "epoch": 0.38586466165413535, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 11415 + }, + { + "epoch": 0.38592481203007517, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 11416 + }, + { + "epoch": 0.38598496240601504, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2677, + "step": 11417 + }, + { + "epoch": 0.3860451127819549, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 11418 + }, + { + "epoch": 0.38610526315789473, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 11419 + }, + { + "epoch": 0.3861654135338346, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2283, + "step": 11420 + }, + { + "epoch": 0.3862255639097744, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 11421 + }, + { + "epoch": 0.3862857142857143, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 11422 + }, + { + "epoch": 0.38634586466165416, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1518, + "step": 11423 + }, + { + "epoch": 0.386406015037594, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 11424 + }, + { + "epoch": 0.38646616541353385, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 11425 + }, + { + "epoch": 0.38652631578947366, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 11426 + }, + { + "epoch": 0.38658646616541353, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.3135, + "step": 11427 + }, + { + "epoch": 0.3866466165413534, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 11428 + }, + { + "epoch": 0.3867067669172932, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 11429 + }, + { + "epoch": 0.3867669172932331, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 11430 + }, + { + "epoch": 0.3868270676691729, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2499, + "step": 11431 + }, + { + "epoch": 0.3868872180451128, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 11432 + }, + { + "epoch": 0.38694736842105265, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 11433 + }, + { + "epoch": 0.38700751879699247, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 11434 + }, + { + "epoch": 0.38706766917293234, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2025, + "step": 11435 + }, + { + "epoch": 0.38712781954887215, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 11436 + }, + { + "epoch": 0.387187969924812, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 11437 + }, + { + "epoch": 0.3872481203007519, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1662, + "step": 11438 + }, + { + "epoch": 0.3873082706766917, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2011, + "step": 11439 + }, + { + "epoch": 0.3873684210526316, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.5154, + "step": 11440 + }, + { + "epoch": 0.38742857142857146, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 11441 + }, + { + "epoch": 0.38748872180451127, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2148, + "step": 11442 + }, + { + "epoch": 0.38754887218045114, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2346, + "step": 11443 + }, + { + "epoch": 0.38760902255639096, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2704, + "step": 11444 + }, + { + "epoch": 0.38766917293233083, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 11445 + }, + { + "epoch": 0.3877293233082707, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1732, + "step": 11446 + }, + { + "epoch": 0.3877894736842105, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 11447 + }, + { + "epoch": 0.3878496240601504, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 11448 + }, + { + "epoch": 0.3879097744360902, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 11449 + }, + { + "epoch": 0.3879699248120301, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2982, + "step": 11450 + }, + { + "epoch": 0.38803007518796995, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2992, + "step": 11451 + }, + { + "epoch": 0.38809022556390976, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.264, + "step": 11452 + }, + { + "epoch": 0.38815037593984963, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1726, + "step": 11453 + }, + { + "epoch": 0.38821052631578945, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 11454 + }, + { + "epoch": 0.3882706766917293, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 11455 + }, + { + "epoch": 0.3883308270676692, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 11456 + }, + { + "epoch": 0.388390977443609, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 11457 + }, + { + "epoch": 0.3884511278195489, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 11458 + }, + { + "epoch": 0.3885112781954887, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.21, + "step": 11459 + }, + { + "epoch": 0.38857142857142857, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2206, + "step": 11460 + }, + { + "epoch": 0.38863157894736844, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.288, + "step": 11461 + }, + { + "epoch": 0.38869172932330825, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 11462 + }, + { + "epoch": 0.3887518796992481, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 11463 + }, + { + "epoch": 0.388812030075188, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 11464 + }, + { + "epoch": 0.3888721804511278, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2332, + "step": 11465 + }, + { + "epoch": 0.3889323308270677, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 11466 + }, + { + "epoch": 0.3889924812030075, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2305, + "step": 11467 + }, + { + "epoch": 0.38905263157894737, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 11468 + }, + { + "epoch": 0.38911278195488724, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2448, + "step": 11469 + }, + { + "epoch": 0.38917293233082706, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 11470 + }, + { + "epoch": 0.38923308270676693, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 11471 + }, + { + "epoch": 0.38929323308270675, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 11472 + }, + { + "epoch": 0.3893533834586466, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2135, + "step": 11473 + }, + { + "epoch": 0.3894135338345865, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 11474 + }, + { + "epoch": 0.3894736842105263, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 11475 + }, + { + "epoch": 0.3895338345864662, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 11476 + }, + { + "epoch": 0.389593984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 11477 + }, + { + "epoch": 0.38965413533834586, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 11478 + }, + { + "epoch": 0.38971428571428574, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 11479 + }, + { + "epoch": 0.38977443609022555, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2838, + "step": 11480 + }, + { + "epoch": 0.3898345864661654, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1966, + "step": 11481 + }, + { + "epoch": 0.38989473684210524, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 11482 + }, + { + "epoch": 0.3899548872180451, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 11483 + }, + { + "epoch": 0.390015037593985, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 11484 + }, + { + "epoch": 0.3900751879699248, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 11485 + }, + { + "epoch": 0.39013533834586467, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 11486 + }, + { + "epoch": 0.3901954887218045, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 11487 + }, + { + "epoch": 0.39025563909774436, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 11488 + }, + { + "epoch": 0.3903157894736842, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 11489 + }, + { + "epoch": 0.39037593984962404, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2182, + "step": 11490 + }, + { + "epoch": 0.3904360902255639, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3061, + "step": 11491 + }, + { + "epoch": 0.3904962406015038, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 11492 + }, + { + "epoch": 0.3905563909774436, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 11493 + }, + { + "epoch": 0.3906165413533835, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 11494 + }, + { + "epoch": 0.3906766917293233, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 11495 + }, + { + "epoch": 0.39073684210526316, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1641, + "step": 11496 + }, + { + "epoch": 0.39079699248120303, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 11497 + }, + { + "epoch": 0.39085714285714285, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3549, + "step": 11498 + }, + { + "epoch": 0.3909172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2002, + "step": 11499 + }, + { + "epoch": 0.39097744360902253, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2061, + "step": 11500 + }, + { + "epoch": 0.3910375939849624, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 11501 + }, + { + "epoch": 0.3910977443609023, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 11502 + }, + { + "epoch": 0.3911578947368421, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 11503 + }, + { + "epoch": 0.39121804511278196, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 11504 + }, + { + "epoch": 0.3912781954887218, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 11505 + }, + { + "epoch": 0.39133834586466165, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 11506 + }, + { + "epoch": 0.3913984962406015, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 11507 + }, + { + "epoch": 0.39145864661654134, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1728, + "step": 11508 + }, + { + "epoch": 0.3915187969924812, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2533, + "step": 11509 + }, + { + "epoch": 0.391578947368421, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 11510 + }, + { + "epoch": 0.3916390977443609, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.2571, + "step": 11511 + }, + { + "epoch": 0.39169924812030077, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 11512 + }, + { + "epoch": 0.3917593984962406, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.2338, + "step": 11513 + }, + { + "epoch": 0.39181954887218046, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 11514 + }, + { + "epoch": 0.39187969924812033, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1713, + "step": 11515 + }, + { + "epoch": 0.39193984962406014, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2343, + "step": 11516 + }, + { + "epoch": 0.392, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1688, + "step": 11517 + }, + { + "epoch": 0.39206015037593983, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 11518 + }, + { + "epoch": 0.3921203007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1988, + "step": 11519 + }, + { + "epoch": 0.3921804511278196, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 11520 + }, + { + "epoch": 0.3922406015037594, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 11521 + }, + { + "epoch": 0.39230075187969926, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 11522 + }, + { + "epoch": 0.3923609022556391, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2093, + "step": 11523 + }, + { + "epoch": 0.39242105263157895, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 11524 + }, + { + "epoch": 0.3924812030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 11525 + }, + { + "epoch": 0.39254135338345864, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 11526 + }, + { + "epoch": 0.3926015037593985, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1696, + "step": 11527 + }, + { + "epoch": 0.3926616541353383, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2014, + "step": 11528 + }, + { + "epoch": 0.3927218045112782, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 11529 + }, + { + "epoch": 0.39278195488721807, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2359, + "step": 11530 + }, + { + "epoch": 0.3928421052631579, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1626, + "step": 11531 + }, + { + "epoch": 0.39290225563909775, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 11532 + }, + { + "epoch": 0.39296240601503757, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 11533 + }, + { + "epoch": 0.39302255639097744, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 11534 + }, + { + "epoch": 0.3930827067669173, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 11535 + }, + { + "epoch": 0.3931428571428571, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 11536 + }, + { + "epoch": 0.393203007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 11537 + }, + { + "epoch": 0.3932631578947368, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 11538 + }, + { + "epoch": 0.3933233082706767, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 11539 + }, + { + "epoch": 0.39338345864661656, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2188, + "step": 11540 + }, + { + "epoch": 0.3934436090225564, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 11541 + }, + { + "epoch": 0.39350375939849624, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 11542 + }, + { + "epoch": 0.3935639097744361, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2291, + "step": 11543 + }, + { + "epoch": 0.39362406015037593, + "grad_norm": 2.1875, + "learning_rate": 3e-05, + "loss": 1.2208, + "step": 11544 + }, + { + "epoch": 0.3936842105263158, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 11545 + }, + { + "epoch": 0.3937443609022556, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1494, + "step": 11546 + }, + { + "epoch": 0.3938045112781955, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 11547 + }, + { + "epoch": 0.39386466165413536, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2209, + "step": 11548 + }, + { + "epoch": 0.3939248120300752, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.302, + "step": 11549 + }, + { + "epoch": 0.39398496240601505, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2813, + "step": 11550 + }, + { + "epoch": 0.39404511278195486, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2328, + "step": 11551 + }, + { + "epoch": 0.39410526315789474, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2143, + "step": 11552 + }, + { + "epoch": 0.3941654135338346, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1956, + "step": 11553 + }, + { + "epoch": 0.3942255639097744, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 11554 + }, + { + "epoch": 0.3942857142857143, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 11555 + }, + { + "epoch": 0.3943458646616541, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 11556 + }, + { + "epoch": 0.394406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 11557 + }, + { + "epoch": 0.39446616541353385, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 11558 + }, + { + "epoch": 0.39452631578947367, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2465, + "step": 11559 + }, + { + "epoch": 0.39458646616541354, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 11560 + }, + { + "epoch": 0.39464661654135336, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2836, + "step": 11561 + }, + { + "epoch": 0.39470676691729323, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2189, + "step": 11562 + }, + { + "epoch": 0.3947669172932331, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 11563 + }, + { + "epoch": 0.3948270676691729, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1937, + "step": 11564 + }, + { + "epoch": 0.3948872180451128, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 11565 + }, + { + "epoch": 0.39494736842105266, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3882, + "step": 11566 + }, + { + "epoch": 0.3950075187969925, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 11567 + }, + { + "epoch": 0.39506766917293235, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 11568 + }, + { + "epoch": 0.39512781954887216, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1929, + "step": 11569 + }, + { + "epoch": 0.39518796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1334, + "step": 11570 + }, + { + "epoch": 0.3952481203007519, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 11571 + }, + { + "epoch": 0.3953082706766917, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 11572 + }, + { + "epoch": 0.3953684210526316, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 11573 + }, + { + "epoch": 0.3954285714285714, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 11574 + }, + { + "epoch": 0.3954887218045113, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 11575 + }, + { + "epoch": 0.39554887218045115, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 11576 + }, + { + "epoch": 0.39560902255639097, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 11577 + }, + { + "epoch": 0.39566917293233084, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 11578 + }, + { + "epoch": 0.39572932330827065, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 11579 + }, + { + "epoch": 0.3957894736842105, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 11580 + }, + { + "epoch": 0.3958496240601504, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 11581 + }, + { + "epoch": 0.3959097744360902, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 11582 + }, + { + "epoch": 0.3959699248120301, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1804, + "step": 11583 + }, + { + "epoch": 0.3960300751879699, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.2441, + "step": 11584 + }, + { + "epoch": 0.39609022556390977, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2287, + "step": 11585 + }, + { + "epoch": 0.39615037593984964, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2287, + "step": 11586 + }, + { + "epoch": 0.39621052631578946, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2543, + "step": 11587 + }, + { + "epoch": 0.39627067669172933, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 11588 + }, + { + "epoch": 0.3963308270676692, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 11589 + }, + { + "epoch": 0.396390977443609, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.264, + "step": 11590 + }, + { + "epoch": 0.3964511278195489, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2128, + "step": 11591 + }, + { + "epoch": 0.3965112781954887, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2089, + "step": 11592 + }, + { + "epoch": 0.3965714285714286, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 11593 + }, + { + "epoch": 0.39663157894736845, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 11594 + }, + { + "epoch": 0.39669172932330826, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 11595 + }, + { + "epoch": 0.39675187969924813, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 11596 + }, + { + "epoch": 0.39681203007518795, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 11597 + }, + { + "epoch": 0.3968721804511278, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2324, + "step": 11598 + }, + { + "epoch": 0.3969323308270677, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2271, + "step": 11599 + }, + { + "epoch": 0.3969924812030075, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 11600 + }, + { + "epoch": 0.3970526315789474, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 11601 + }, + { + "epoch": 0.3971127819548872, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 11602 + }, + { + "epoch": 0.39717293233082707, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 11603 + }, + { + "epoch": 0.39723308270676694, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1989, + "step": 11604 + }, + { + "epoch": 0.39729323308270675, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1496, + "step": 11605 + }, + { + "epoch": 0.3973533834586466, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.485, + "step": 11606 + }, + { + "epoch": 0.39741353383458644, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 11607 + }, + { + "epoch": 0.3974736842105263, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 11608 + }, + { + "epoch": 0.3975338345864662, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 11609 + }, + { + "epoch": 0.397593984962406, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 11610 + }, + { + "epoch": 0.39765413533834587, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 11611 + }, + { + "epoch": 0.3977142857142857, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 11612 + }, + { + "epoch": 0.39777443609022556, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2255, + "step": 11613 + }, + { + "epoch": 0.39783458646616543, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 11614 + }, + { + "epoch": 0.39789473684210525, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 11615 + }, + { + "epoch": 0.3979548872180451, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3007, + "step": 11616 + }, + { + "epoch": 0.398015037593985, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 11617 + }, + { + "epoch": 0.3980751879699248, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 11618 + }, + { + "epoch": 0.3981353383458647, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 11619 + }, + { + "epoch": 0.3981954887218045, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2053, + "step": 11620 + }, + { + "epoch": 0.39825563909774436, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 11621 + }, + { + "epoch": 0.39831578947368423, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 11622 + }, + { + "epoch": 0.39837593984962405, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2134, + "step": 11623 + }, + { + "epoch": 0.3984360902255639, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 11624 + }, + { + "epoch": 0.39849624060150374, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 11625 + }, + { + "epoch": 0.3985563909774436, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 11626 + }, + { + "epoch": 0.3986165413533835, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.252, + "step": 11627 + }, + { + "epoch": 0.3986766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2183, + "step": 11628 + }, + { + "epoch": 0.39873684210526317, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 11629 + }, + { + "epoch": 0.398796992481203, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 11630 + }, + { + "epoch": 0.39885714285714285, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 11631 + }, + { + "epoch": 0.3989172932330827, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 11632 + }, + { + "epoch": 0.39897744360902254, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 11633 + }, + { + "epoch": 0.3990375939849624, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1979, + "step": 11634 + }, + { + "epoch": 0.39909774436090223, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.187, + "step": 11635 + }, + { + "epoch": 0.3991578947368421, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 11636 + }, + { + "epoch": 0.39921804511278197, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 11637 + }, + { + "epoch": 0.3992781954887218, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 11638 + }, + { + "epoch": 0.39933834586466166, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2263, + "step": 11639 + }, + { + "epoch": 0.39939849624060153, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 11640 + }, + { + "epoch": 0.39945864661654135, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 11641 + }, + { + "epoch": 0.3995187969924812, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 11642 + }, + { + "epoch": 0.39957894736842103, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.2206, + "step": 11643 + }, + { + "epoch": 0.3996390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 11644 + }, + { + "epoch": 0.3996992481203008, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 11645 + }, + { + "epoch": 0.3997593984962406, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1446, + "step": 11646 + }, + { + "epoch": 0.39981954887218046, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 11647 + }, + { + "epoch": 0.3998796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2303, + "step": 11648 + }, + { + "epoch": 0.39993984962406015, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.224, + "step": 11649 + }, + { + "epoch": 0.4, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1817, + "step": 11650 + }, + { + "epoch": 0.40006015037593984, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 11651 + }, + { + "epoch": 0.4001203007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 11652 + }, + { + "epoch": 0.4001804511278195, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 11653 + }, + { + "epoch": 0.4002406015037594, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 11654 + }, + { + "epoch": 0.40030075187969927, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 11655 + }, + { + "epoch": 0.4003609022556391, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 11656 + }, + { + "epoch": 0.40042105263157896, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2001, + "step": 11657 + }, + { + "epoch": 0.40048120300751877, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 11658 + }, + { + "epoch": 0.40054135338345864, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 11659 + }, + { + "epoch": 0.4006015037593985, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 11660 + }, + { + "epoch": 0.40066165413533833, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 11661 + }, + { + "epoch": 0.4007218045112782, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 11662 + }, + { + "epoch": 0.4007819548872181, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1914, + "step": 11663 + }, + { + "epoch": 0.4008421052631579, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 11664 + }, + { + "epoch": 0.40090225563909776, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 11665 + }, + { + "epoch": 0.4009624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 11666 + }, + { + "epoch": 0.40102255639097745, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.402, + "step": 11667 + }, + { + "epoch": 0.4010827067669173, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2277, + "step": 11668 + }, + { + "epoch": 0.40114285714285713, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.3152, + "step": 11669 + }, + { + "epoch": 0.401203007518797, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 11670 + }, + { + "epoch": 0.4012631578947368, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 11671 + }, + { + "epoch": 0.4013233082706767, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2301, + "step": 11672 + }, + { + "epoch": 0.40138345864661656, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 11673 + }, + { + "epoch": 0.4014436090225564, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 11674 + }, + { + "epoch": 0.40150375939849625, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 11675 + }, + { + "epoch": 0.40156390977443607, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1949, + "step": 11676 + }, + { + "epoch": 0.40162406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 11677 + }, + { + "epoch": 0.4016842105263158, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 11678 + }, + { + "epoch": 0.4017443609022556, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 11679 + }, + { + "epoch": 0.4018045112781955, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2507, + "step": 11680 + }, + { + "epoch": 0.4018646616541353, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 11681 + }, + { + "epoch": 0.4019248120300752, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 11682 + }, + { + "epoch": 0.40198496240601506, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1986, + "step": 11683 + }, + { + "epoch": 0.40204511278195487, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 11684 + }, + { + "epoch": 0.40210526315789474, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 11685 + }, + { + "epoch": 0.40216541353383456, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 11686 + }, + { + "epoch": 0.40222556390977443, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 11687 + }, + { + "epoch": 0.4022857142857143, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1894, + "step": 11688 + }, + { + "epoch": 0.4023458646616541, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 11689 + }, + { + "epoch": 0.402406015037594, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.1951, + "step": 11690 + }, + { + "epoch": 0.40246616541353386, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2069, + "step": 11691 + }, + { + "epoch": 0.4025263157894737, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2262, + "step": 11692 + }, + { + "epoch": 0.40258646616541355, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 11693 + }, + { + "epoch": 0.40264661654135336, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 11694 + }, + { + "epoch": 0.40270676691729324, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 11695 + }, + { + "epoch": 0.4027669172932331, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 11696 + }, + { + "epoch": 0.4028270676691729, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 11697 + }, + { + "epoch": 0.4028872180451128, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2344, + "step": 11698 + }, + { + "epoch": 0.4029473684210526, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 11699 + }, + { + "epoch": 0.4030075187969925, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 11700 + }, + { + "epoch": 0.40306766917293235, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 11701 + }, + { + "epoch": 0.40312781954887217, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 11702 + }, + { + "epoch": 0.40318796992481204, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 11703 + }, + { + "epoch": 0.40324812030075186, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 11704 + }, + { + "epoch": 0.4033082706766917, + "grad_norm": 1.6640625, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 11705 + }, + { + "epoch": 0.4033684210526316, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1728, + "step": 11706 + }, + { + "epoch": 0.4034285714285714, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 11707 + }, + { + "epoch": 0.4034887218045113, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 11708 + }, + { + "epoch": 0.4035488721804511, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2074, + "step": 11709 + }, + { + "epoch": 0.403609022556391, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 11710 + }, + { + "epoch": 0.40366917293233084, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1961, + "step": 11711 + }, + { + "epoch": 0.40372932330827066, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2114, + "step": 11712 + }, + { + "epoch": 0.40378947368421053, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2178, + "step": 11713 + }, + { + "epoch": 0.4038496240601504, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 11714 + }, + { + "epoch": 0.4039097744360902, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 11715 + }, + { + "epoch": 0.4039699248120301, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1979, + "step": 11716 + }, + { + "epoch": 0.4040300751879699, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 11717 + }, + { + "epoch": 0.4040902255639098, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 11718 + }, + { + "epoch": 0.40415037593984965, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2388, + "step": 11719 + }, + { + "epoch": 0.40421052631578946, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2376, + "step": 11720 + }, + { + "epoch": 0.40427067669172934, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 11721 + }, + { + "epoch": 0.40433082706766915, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1776, + "step": 11722 + }, + { + "epoch": 0.404390977443609, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1831, + "step": 11723 + }, + { + "epoch": 0.4044511278195489, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 11724 + }, + { + "epoch": 0.4045112781954887, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1545, + "step": 11725 + }, + { + "epoch": 0.4045714285714286, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 11726 + }, + { + "epoch": 0.4046315789473684, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3216, + "step": 11727 + }, + { + "epoch": 0.40469172932330827, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 11728 + }, + { + "epoch": 0.40475187969924814, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 11729 + }, + { + "epoch": 0.40481203007518796, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2182, + "step": 11730 + }, + { + "epoch": 0.4048721804511278, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 11731 + }, + { + "epoch": 0.40493233082706764, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2251, + "step": 11732 + }, + { + "epoch": 0.4049924812030075, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1688, + "step": 11733 + }, + { + "epoch": 0.4050526315789474, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 11734 + }, + { + "epoch": 0.4051127819548872, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2209, + "step": 11735 + }, + { + "epoch": 0.4051729323308271, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.244, + "step": 11736 + }, + { + "epoch": 0.40523308270676694, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2775, + "step": 11737 + }, + { + "epoch": 0.40529323308270676, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 11738 + }, + { + "epoch": 0.40535338345864663, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 11739 + }, + { + "epoch": 0.40541353383458645, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 11740 + }, + { + "epoch": 0.4054736842105263, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 11741 + }, + { + "epoch": 0.4055338345864662, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 11742 + }, + { + "epoch": 0.405593984962406, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.1927, + "step": 11743 + }, + { + "epoch": 0.4056541353383459, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 11744 + }, + { + "epoch": 0.4057142857142857, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2259, + "step": 11745 + }, + { + "epoch": 0.40577443609022557, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 11746 + }, + { + "epoch": 0.40583458646616544, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2143, + "step": 11747 + }, + { + "epoch": 0.40589473684210525, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 11748 + }, + { + "epoch": 0.4059548872180451, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 11749 + }, + { + "epoch": 0.40601503759398494, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 11750 + }, + { + "epoch": 0.4060751879699248, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2719, + "step": 11751 + }, + { + "epoch": 0.4061353383458647, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 11752 + }, + { + "epoch": 0.4061954887218045, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2278, + "step": 11753 + }, + { + "epoch": 0.40625563909774437, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 11754 + }, + { + "epoch": 0.4063157894736842, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.2862, + "step": 11755 + }, + { + "epoch": 0.40637593984962406, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 11756 + }, + { + "epoch": 0.40643609022556393, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2318, + "step": 11757 + }, + { + "epoch": 0.40649624060150374, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 11758 + }, + { + "epoch": 0.4065563909774436, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2264, + "step": 11759 + }, + { + "epoch": 0.40661654135338343, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.4006, + "step": 11760 + }, + { + "epoch": 0.4066766917293233, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 11761 + }, + { + "epoch": 0.4067368421052632, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 11762 + }, + { + "epoch": 0.406796992481203, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 11763 + }, + { + "epoch": 0.40685714285714286, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 11764 + }, + { + "epoch": 0.40691729323308273, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 11765 + }, + { + "epoch": 0.40697744360902255, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 11766 + }, + { + "epoch": 0.4070375939849624, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2019, + "step": 11767 + }, + { + "epoch": 0.40709774436090224, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 11768 + }, + { + "epoch": 0.4071578947368421, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 11769 + }, + { + "epoch": 0.407218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 11770 + }, + { + "epoch": 0.4072781954887218, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 11771 + }, + { + "epoch": 0.40733834586466167, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2262, + "step": 11772 + }, + { + "epoch": 0.4073984962406015, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1314, + "step": 11773 + }, + { + "epoch": 0.40745864661654135, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 11774 + }, + { + "epoch": 0.4075187969924812, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1612, + "step": 11775 + }, + { + "epoch": 0.40757894736842104, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2472, + "step": 11776 + }, + { + "epoch": 0.4076390977443609, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.187, + "step": 11777 + }, + { + "epoch": 0.40769924812030073, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2194, + "step": 11778 + }, + { + "epoch": 0.4077593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1915, + "step": 11779 + }, + { + "epoch": 0.40781954887218047, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 11780 + }, + { + "epoch": 0.4078796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 11781 + }, + { + "epoch": 0.40793984962406016, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 11782 + }, + { + "epoch": 0.408, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 11783 + }, + { + "epoch": 0.40806015037593985, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 11784 + }, + { + "epoch": 0.4081203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1718, + "step": 11785 + }, + { + "epoch": 0.40818045112781953, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3069, + "step": 11786 + }, + { + "epoch": 0.4082406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1708, + "step": 11787 + }, + { + "epoch": 0.4083007518796993, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 11788 + }, + { + "epoch": 0.4083609022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2063, + "step": 11789 + }, + { + "epoch": 0.40842105263157896, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2301, + "step": 11790 + }, + { + "epoch": 0.4084812030075188, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.19, + "step": 11791 + }, + { + "epoch": 0.40854135338345865, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2119, + "step": 11792 + }, + { + "epoch": 0.4086015037593985, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 11793 + }, + { + "epoch": 0.40866165413533834, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1617, + "step": 11794 + }, + { + "epoch": 0.4087218045112782, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 11795 + }, + { + "epoch": 0.408781954887218, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2072, + "step": 11796 + }, + { + "epoch": 0.4088421052631579, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 11797 + }, + { + "epoch": 0.40890225563909777, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 11798 + }, + { + "epoch": 0.4089624060150376, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.3017, + "step": 11799 + }, + { + "epoch": 0.40902255639097745, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 11800 + }, + { + "epoch": 0.40908270676691727, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.187, + "step": 11801 + }, + { + "epoch": 0.40914285714285714, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 11802 + }, + { + "epoch": 0.409203007518797, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 11803 + }, + { + "epoch": 0.40926315789473683, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 11804 + }, + { + "epoch": 0.4093233082706767, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1884, + "step": 11805 + }, + { + "epoch": 0.4093834586466165, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 11806 + }, + { + "epoch": 0.4094436090225564, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 11807 + }, + { + "epoch": 0.40950375939849626, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2467, + "step": 11808 + }, + { + "epoch": 0.4095639097744361, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 11809 + }, + { + "epoch": 0.40962406015037595, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 11810 + }, + { + "epoch": 0.40968421052631576, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 11811 + }, + { + "epoch": 0.40974436090225563, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 11812 + }, + { + "epoch": 0.4098045112781955, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 11813 + }, + { + "epoch": 0.4098646616541353, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 11814 + }, + { + "epoch": 0.4099248120300752, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1662, + "step": 11815 + }, + { + "epoch": 0.40998496240601506, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 11816 + }, + { + "epoch": 0.4100451127819549, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 11817 + }, + { + "epoch": 0.41010526315789475, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 11818 + }, + { + "epoch": 0.41016541353383457, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 11819 + }, + { + "epoch": 0.41022556390977444, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 11820 + }, + { + "epoch": 0.4102857142857143, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 11821 + }, + { + "epoch": 0.4103458646616541, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.227, + "step": 11822 + }, + { + "epoch": 0.410406015037594, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 11823 + }, + { + "epoch": 0.4104661654135338, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 11824 + }, + { + "epoch": 0.4105263157894737, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 11825 + }, + { + "epoch": 0.41058646616541356, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 11826 + }, + { + "epoch": 0.41064661654135337, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1965, + "step": 11827 + }, + { + "epoch": 0.41070676691729324, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2992, + "step": 11828 + }, + { + "epoch": 0.41076691729323306, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 11829 + }, + { + "epoch": 0.41082706766917293, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2652, + "step": 11830 + }, + { + "epoch": 0.4108872180451128, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 11831 + }, + { + "epoch": 0.4109473684210526, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 11832 + }, + { + "epoch": 0.4110075187969925, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 11833 + }, + { + "epoch": 0.4110676691729323, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.3299, + "step": 11834 + }, + { + "epoch": 0.4111278195488722, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 11835 + }, + { + "epoch": 0.41118796992481205, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 11836 + }, + { + "epoch": 0.41124812030075186, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 11837 + }, + { + "epoch": 0.41130827067669173, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1801, + "step": 11838 + }, + { + "epoch": 0.4113684210526316, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1991, + "step": 11839 + }, + { + "epoch": 0.4114285714285714, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 11840 + }, + { + "epoch": 0.4114887218045113, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 11841 + }, + { + "epoch": 0.4115488721804511, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 11842 + }, + { + "epoch": 0.411609022556391, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 11843 + }, + { + "epoch": 0.41166917293233085, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2279, + "step": 11844 + }, + { + "epoch": 0.41172932330827067, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 11845 + }, + { + "epoch": 0.41178947368421054, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 11846 + }, + { + "epoch": 0.41184962406015035, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 11847 + }, + { + "epoch": 0.4119097744360902, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2548, + "step": 11848 + }, + { + "epoch": 0.4119699248120301, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2324, + "step": 11849 + }, + { + "epoch": 0.4120300751879699, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 11850 + }, + { + "epoch": 0.4120902255639098, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 11851 + }, + { + "epoch": 0.4121503759398496, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1911, + "step": 11852 + }, + { + "epoch": 0.41221052631578947, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2036, + "step": 11853 + }, + { + "epoch": 0.41227067669172934, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1988, + "step": 11854 + }, + { + "epoch": 0.41233082706766916, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 11855 + }, + { + "epoch": 0.41239097744360903, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2437, + "step": 11856 + }, + { + "epoch": 0.41245112781954885, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2432, + "step": 11857 + }, + { + "epoch": 0.4125112781954887, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 11858 + }, + { + "epoch": 0.4125714285714286, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.224, + "step": 11859 + }, + { + "epoch": 0.4126315789473684, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 11860 + }, + { + "epoch": 0.4126917293233083, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1671, + "step": 11861 + }, + { + "epoch": 0.41275187969924815, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 11862 + }, + { + "epoch": 0.41281203007518796, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 11863 + }, + { + "epoch": 0.41287218045112783, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 11864 + }, + { + "epoch": 0.41293233082706765, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 11865 + }, + { + "epoch": 0.4129924812030075, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 11866 + }, + { + "epoch": 0.4130526315789474, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 11867 + }, + { + "epoch": 0.4131127819548872, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 11868 + }, + { + "epoch": 0.4131729323308271, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 11869 + }, + { + "epoch": 0.4132330827067669, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2663, + "step": 11870 + }, + { + "epoch": 0.41329323308270677, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 11871 + }, + { + "epoch": 0.41335338345864664, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 11872 + }, + { + "epoch": 0.41341353383458646, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 11873 + }, + { + "epoch": 0.4134736842105263, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 11874 + }, + { + "epoch": 0.41353383458646614, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1429, + "step": 11875 + }, + { + "epoch": 0.413593984962406, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2617, + "step": 11876 + }, + { + "epoch": 0.4136541353383459, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 11877 + }, + { + "epoch": 0.4137142857142857, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 11878 + }, + { + "epoch": 0.4137744360902256, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 11879 + }, + { + "epoch": 0.4138345864661654, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 11880 + }, + { + "epoch": 0.41389473684210526, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2155, + "step": 11881 + }, + { + "epoch": 0.41395488721804513, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 11882 + }, + { + "epoch": 0.41401503759398495, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 11883 + }, + { + "epoch": 0.4140751879699248, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2553, + "step": 11884 + }, + { + "epoch": 0.41413533834586463, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 11885 + }, + { + "epoch": 0.4141954887218045, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1925, + "step": 11886 + }, + { + "epoch": 0.4142556390977444, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 11887 + }, + { + "epoch": 0.4143157894736842, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 11888 + }, + { + "epoch": 0.41437593984962406, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 11889 + }, + { + "epoch": 0.41443609022556394, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1612, + "step": 11890 + }, + { + "epoch": 0.41449624060150375, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.3191, + "step": 11891 + }, + { + "epoch": 0.4145563909774436, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 11892 + }, + { + "epoch": 0.41461654135338344, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 11893 + }, + { + "epoch": 0.4146766917293233, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2107, + "step": 11894 + }, + { + "epoch": 0.4147368421052632, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 11895 + }, + { + "epoch": 0.414796992481203, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.155, + "step": 11896 + }, + { + "epoch": 0.41485714285714287, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.3595, + "step": 11897 + }, + { + "epoch": 0.4149172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 11898 + }, + { + "epoch": 0.41497744360902256, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 11899 + }, + { + "epoch": 0.4150375939849624, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 11900 + }, + { + "epoch": 0.41509774436090224, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.2839, + "step": 11901 + }, + { + "epoch": 0.4151578947368421, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 11902 + }, + { + "epoch": 0.41521804511278193, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 11903 + }, + { + "epoch": 0.4152781954887218, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 11904 + }, + { + "epoch": 0.4153383458646617, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 11905 + }, + { + "epoch": 0.4153984962406015, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2093, + "step": 11906 + }, + { + "epoch": 0.41545864661654136, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 11907 + }, + { + "epoch": 0.4155187969924812, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 11908 + }, + { + "epoch": 0.41557894736842105, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 11909 + }, + { + "epoch": 0.4156390977443609, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.3747, + "step": 11910 + }, + { + "epoch": 0.41569924812030074, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 11911 + }, + { + "epoch": 0.4157593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 11912 + }, + { + "epoch": 0.4158195488721805, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 11913 + }, + { + "epoch": 0.4158796992481203, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 11914 + }, + { + "epoch": 0.41593984962406017, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2694, + "step": 11915 + }, + { + "epoch": 0.416, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 11916 + }, + { + "epoch": 0.41606015037593985, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 11917 + }, + { + "epoch": 0.4161203007518797, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 11918 + }, + { + "epoch": 0.41618045112781954, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 11919 + }, + { + "epoch": 0.4162406015037594, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2488, + "step": 11920 + }, + { + "epoch": 0.4163007518796992, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1971, + "step": 11921 + }, + { + "epoch": 0.4163609022556391, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2433, + "step": 11922 + }, + { + "epoch": 0.41642105263157897, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2495, + "step": 11923 + }, + { + "epoch": 0.4164812030075188, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1879, + "step": 11924 + }, + { + "epoch": 0.41654135338345866, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 11925 + }, + { + "epoch": 0.4166015037593985, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 11926 + }, + { + "epoch": 0.41666165413533834, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 11927 + }, + { + "epoch": 0.4167218045112782, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 11928 + }, + { + "epoch": 0.41678195488721803, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2684, + "step": 11929 + }, + { + "epoch": 0.4168421052631579, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 11930 + }, + { + "epoch": 0.4169022556390977, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 11931 + }, + { + "epoch": 0.4169624060150376, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 11932 + }, + { + "epoch": 0.41702255639097746, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 11933 + }, + { + "epoch": 0.4170827067669173, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 11934 + }, + { + "epoch": 0.41714285714285715, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2289, + "step": 11935 + }, + { + "epoch": 0.417203007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 11936 + }, + { + "epoch": 0.41726315789473684, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 11937 + }, + { + "epoch": 0.4173233082706767, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 11938 + }, + { + "epoch": 0.4173834586466165, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1545, + "step": 11939 + }, + { + "epoch": 0.4174436090225564, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1767, + "step": 11940 + }, + { + "epoch": 0.41750375939849627, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2144, + "step": 11941 + }, + { + "epoch": 0.4175639097744361, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 11942 + }, + { + "epoch": 0.41762406015037595, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2028, + "step": 11943 + }, + { + "epoch": 0.41768421052631577, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2364, + "step": 11944 + }, + { + "epoch": 0.41774436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 11945 + }, + { + "epoch": 0.4178045112781955, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2431, + "step": 11946 + }, + { + "epoch": 0.4178646616541353, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1465, + "step": 11947 + }, + { + "epoch": 0.4179248120300752, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1751, + "step": 11948 + }, + { + "epoch": 0.417984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 11949 + }, + { + "epoch": 0.4180451127819549, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.204, + "step": 11950 + }, + { + "epoch": 0.41810526315789476, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 11951 + }, + { + "epoch": 0.4181654135338346, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1834, + "step": 11952 + }, + { + "epoch": 0.41822556390977444, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 11953 + }, + { + "epoch": 0.41828571428571426, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3475, + "step": 11954 + }, + { + "epoch": 0.41834586466165413, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 11955 + }, + { + "epoch": 0.418406015037594, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 11956 + }, + { + "epoch": 0.4184661654135338, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2452, + "step": 11957 + }, + { + "epoch": 0.4185263157894737, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 11958 + }, + { + "epoch": 0.4185864661654135, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2242, + "step": 11959 + }, + { + "epoch": 0.4186466165413534, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.271, + "step": 11960 + }, + { + "epoch": 0.41870676691729325, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 11961 + }, + { + "epoch": 0.41876691729323307, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 11962 + }, + { + "epoch": 0.41882706766917294, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 11963 + }, + { + "epoch": 0.4188872180451128, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 11964 + }, + { + "epoch": 0.4189473684210526, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 11965 + }, + { + "epoch": 0.4190075187969925, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 11966 + }, + { + "epoch": 0.4190676691729323, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 11967 + }, + { + "epoch": 0.4191278195488722, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 11968 + }, + { + "epoch": 0.41918796992481205, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 11969 + }, + { + "epoch": 0.41924812030075187, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2036, + "step": 11970 + }, + { + "epoch": 0.41930827067669174, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 11971 + }, + { + "epoch": 0.41936842105263156, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 11972 + }, + { + "epoch": 0.41942857142857143, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1983, + "step": 11973 + }, + { + "epoch": 0.4194887218045113, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 11974 + }, + { + "epoch": 0.4195488721804511, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 11975 + }, + { + "epoch": 0.419609022556391, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1597, + "step": 11976 + }, + { + "epoch": 0.4196691729323308, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 11977 + }, + { + "epoch": 0.4197293233082707, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.1823, + "step": 11978 + }, + { + "epoch": 0.41978947368421055, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 11979 + }, + { + "epoch": 0.41984962406015036, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1823, + "step": 11980 + }, + { + "epoch": 0.41990977443609023, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 11981 + }, + { + "epoch": 0.41996992481203005, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2682, + "step": 11982 + }, + { + "epoch": 0.4200300751879699, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 11983 + }, + { + "epoch": 0.4200902255639098, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1879, + "step": 11984 + }, + { + "epoch": 0.4201503759398496, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2697, + "step": 11985 + }, + { + "epoch": 0.4202105263157895, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 11986 + }, + { + "epoch": 0.42027067669172935, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 11987 + }, + { + "epoch": 0.42033082706766917, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 11988 + }, + { + "epoch": 0.42039097744360904, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3438, + "step": 11989 + }, + { + "epoch": 0.42045112781954885, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 11990 + }, + { + "epoch": 0.4205112781954887, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 11991 + }, + { + "epoch": 0.4205714285714286, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2348, + "step": 11992 + }, + { + "epoch": 0.4206315789473684, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 11993 + }, + { + "epoch": 0.4206917293233083, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 11994 + }, + { + "epoch": 0.4207518796992481, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 11995 + }, + { + "epoch": 0.42081203007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 11996 + }, + { + "epoch": 0.42087218045112784, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 11997 + }, + { + "epoch": 0.42093233082706766, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 11998 + }, + { + "epoch": 0.42099248120300753, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.2109, + "step": 11999 + }, + { + "epoch": 0.42105263157894735, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 12000 + }, + { + "epoch": 0.4211127819548872, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 12001 + }, + { + "epoch": 0.4211729323308271, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 12002 + }, + { + "epoch": 0.4212330827067669, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 12003 + }, + { + "epoch": 0.4212932330827068, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2226, + "step": 12004 + }, + { + "epoch": 0.4213533834586466, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 12005 + }, + { + "epoch": 0.42141353383458646, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.2021, + "step": 12006 + }, + { + "epoch": 0.42147368421052633, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.23, + "step": 12007 + }, + { + "epoch": 0.42153383458646615, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 12008 + }, + { + "epoch": 0.421593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 12009 + }, + { + "epoch": 0.4216541353383459, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 12010 + }, + { + "epoch": 0.4217142857142857, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.211, + "step": 12011 + }, + { + "epoch": 0.4217744360902256, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 12012 + }, + { + "epoch": 0.4218345864661654, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 12013 + }, + { + "epoch": 0.42189473684210527, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2271, + "step": 12014 + }, + { + "epoch": 0.42195488721804514, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2315, + "step": 12015 + }, + { + "epoch": 0.42201503759398495, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 12016 + }, + { + "epoch": 0.4220751879699248, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3034, + "step": 12017 + }, + { + "epoch": 0.42213533834586464, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2703, + "step": 12018 + }, + { + "epoch": 0.4221954887218045, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 12019 + }, + { + "epoch": 0.4222556390977444, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.3264, + "step": 12020 + }, + { + "epoch": 0.4223157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 12021 + }, + { + "epoch": 0.42237593984962407, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 12022 + }, + { + "epoch": 0.4224360902255639, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 12023 + }, + { + "epoch": 0.42249624060150376, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 12024 + }, + { + "epoch": 0.42255639097744363, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2374, + "step": 12025 + }, + { + "epoch": 0.42261654135338345, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2448, + "step": 12026 + }, + { + "epoch": 0.4226766917293233, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 12027 + }, + { + "epoch": 0.42273684210526313, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 12028 + }, + { + "epoch": 0.422796992481203, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.2085, + "step": 12029 + }, + { + "epoch": 0.4228571428571429, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2942, + "step": 12030 + }, + { + "epoch": 0.4229172932330827, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 12031 + }, + { + "epoch": 0.42297744360902256, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3335, + "step": 12032 + }, + { + "epoch": 0.4230375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 12033 + }, + { + "epoch": 0.42309774436090225, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1638, + "step": 12034 + }, + { + "epoch": 0.4231578947368421, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2788, + "step": 12035 + }, + { + "epoch": 0.42321804511278194, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 12036 + }, + { + "epoch": 0.4232781954887218, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.147, + "step": 12037 + }, + { + "epoch": 0.4233383458646617, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1711, + "step": 12038 + }, + { + "epoch": 0.4233984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2563, + "step": 12039 + }, + { + "epoch": 0.42345864661654137, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 12040 + }, + { + "epoch": 0.4235187969924812, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 12041 + }, + { + "epoch": 0.42357894736842105, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 12042 + }, + { + "epoch": 0.4236390977443609, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 12043 + }, + { + "epoch": 0.42369924812030074, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 12044 + }, + { + "epoch": 0.4237593984962406, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 12045 + }, + { + "epoch": 0.42381954887218043, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.3608, + "step": 12046 + }, + { + "epoch": 0.4238796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2688, + "step": 12047 + }, + { + "epoch": 0.42393984962406017, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 12048 + }, + { + "epoch": 0.424, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 12049 + }, + { + "epoch": 0.42406015037593986, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 12050 + }, + { + "epoch": 0.4241203007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 12051 + }, + { + "epoch": 0.42418045112781955, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 12052 + }, + { + "epoch": 0.4242406015037594, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 12053 + }, + { + "epoch": 0.42430075187969923, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 12054 + }, + { + "epoch": 0.4243609022556391, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 12055 + }, + { + "epoch": 0.4244210526315789, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.2538, + "step": 12056 + }, + { + "epoch": 0.4244812030075188, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 12057 + }, + { + "epoch": 0.42454135338345866, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 12058 + }, + { + "epoch": 0.4246015037593985, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2093, + "step": 12059 + }, + { + "epoch": 0.42466165413533835, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.24, + "step": 12060 + }, + { + "epoch": 0.4247218045112782, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1994, + "step": 12061 + }, + { + "epoch": 0.42478195488721804, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2601, + "step": 12062 + }, + { + "epoch": 0.4248421052631579, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1735, + "step": 12063 + }, + { + "epoch": 0.4249022556390977, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2034, + "step": 12064 + }, + { + "epoch": 0.4249624060150376, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 12065 + }, + { + "epoch": 0.42502255639097747, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2355, + "step": 12066 + }, + { + "epoch": 0.4250827067669173, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 12067 + }, + { + "epoch": 0.42514285714285716, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2256, + "step": 12068 + }, + { + "epoch": 0.42520300751879697, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 12069 + }, + { + "epoch": 0.42526315789473684, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 12070 + }, + { + "epoch": 0.4253233082706767, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 12071 + }, + { + "epoch": 0.42538345864661653, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 12072 + }, + { + "epoch": 0.4254436090225564, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 12073 + }, + { + "epoch": 0.4255037593984962, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 12074 + }, + { + "epoch": 0.4255639097744361, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 12075 + }, + { + "epoch": 0.42562406015037596, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 12076 + }, + { + "epoch": 0.4256842105263158, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 12077 + }, + { + "epoch": 0.42574436090225565, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 12078 + }, + { + "epoch": 0.42580451127819546, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2483, + "step": 12079 + }, + { + "epoch": 0.42586466165413533, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 12080 + }, + { + "epoch": 0.4259248120300752, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2232, + "step": 12081 + }, + { + "epoch": 0.425984962406015, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 12082 + }, + { + "epoch": 0.4260451127819549, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2006, + "step": 12083 + }, + { + "epoch": 0.4261052631578947, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2425, + "step": 12084 + }, + { + "epoch": 0.4261654135338346, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2268, + "step": 12085 + }, + { + "epoch": 0.42622556390977445, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 12086 + }, + { + "epoch": 0.42628571428571427, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 12087 + }, + { + "epoch": 0.42634586466165414, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 12088 + }, + { + "epoch": 0.426406015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 12089 + }, + { + "epoch": 0.4264661654135338, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 12090 + }, + { + "epoch": 0.4265263157894737, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 12091 + }, + { + "epoch": 0.4265864661654135, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 12092 + }, + { + "epoch": 0.4266466165413534, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 12093 + }, + { + "epoch": 0.42670676691729326, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 12094 + }, + { + "epoch": 0.4267669172932331, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 12095 + }, + { + "epoch": 0.42682706766917294, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 12096 + }, + { + "epoch": 0.42688721804511276, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2633, + "step": 12097 + }, + { + "epoch": 0.42694736842105263, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1998, + "step": 12098 + }, + { + "epoch": 0.4270075187969925, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2159, + "step": 12099 + }, + { + "epoch": 0.4270676691729323, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1927, + "step": 12100 + }, + { + "epoch": 0.4271278195488722, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3261, + "step": 12101 + }, + { + "epoch": 0.427187969924812, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.302, + "step": 12102 + }, + { + "epoch": 0.4272481203007519, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2432, + "step": 12103 + }, + { + "epoch": 0.42730827067669175, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 12104 + }, + { + "epoch": 0.42736842105263156, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 12105 + }, + { + "epoch": 0.42742857142857144, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 12106 + }, + { + "epoch": 0.42748872180451125, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 12107 + }, + { + "epoch": 0.4275488721804511, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 12108 + }, + { + "epoch": 0.427609022556391, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 12109 + }, + { + "epoch": 0.4276691729323308, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 12110 + }, + { + "epoch": 0.4277293233082707, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 12111 + }, + { + "epoch": 0.42778947368421055, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2291, + "step": 12112 + }, + { + "epoch": 0.42784962406015037, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 12113 + }, + { + "epoch": 0.42790977443609024, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 12114 + }, + { + "epoch": 0.42796992481203006, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 12115 + }, + { + "epoch": 0.4280300751879699, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2055, + "step": 12116 + }, + { + "epoch": 0.4280902255639098, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 12117 + }, + { + "epoch": 0.4281503759398496, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 12118 + }, + { + "epoch": 0.4282105263157895, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.2079, + "step": 12119 + }, + { + "epoch": 0.4282706766917293, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 12120 + }, + { + "epoch": 0.4283308270676692, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 12121 + }, + { + "epoch": 0.42839097744360904, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.136, + "step": 12122 + }, + { + "epoch": 0.42845112781954886, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.2757, + "step": 12123 + }, + { + "epoch": 0.42851127819548873, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 12124 + }, + { + "epoch": 0.42857142857142855, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2581, + "step": 12125 + }, + { + "epoch": 0.4286315789473684, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2755, + "step": 12126 + }, + { + "epoch": 0.4286917293233083, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2102, + "step": 12127 + }, + { + "epoch": 0.4287518796992481, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 12128 + }, + { + "epoch": 0.428812030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 12129 + }, + { + "epoch": 0.4288721804511278, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1927, + "step": 12130 + }, + { + "epoch": 0.42893233082706766, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 12131 + }, + { + "epoch": 0.42899248120300754, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 12132 + }, + { + "epoch": 0.42905263157894735, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 12133 + }, + { + "epoch": 0.4291127819548872, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 12134 + }, + { + "epoch": 0.4291729323308271, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1995, + "step": 12135 + }, + { + "epoch": 0.4292330827067669, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 12136 + }, + { + "epoch": 0.4292932330827068, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2174, + "step": 12137 + }, + { + "epoch": 0.4293533834586466, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 12138 + }, + { + "epoch": 0.42941353383458647, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 12139 + }, + { + "epoch": 0.42947368421052634, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 12140 + }, + { + "epoch": 0.42953383458646616, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 12141 + }, + { + "epoch": 0.42959398496240603, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 12142 + }, + { + "epoch": 0.42965413533834584, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2626, + "step": 12143 + }, + { + "epoch": 0.4297142857142857, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.2238, + "step": 12144 + }, + { + "epoch": 0.4297744360902256, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2831, + "step": 12145 + }, + { + "epoch": 0.4298345864661654, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.1833, + "step": 12146 + }, + { + "epoch": 0.4298947368421053, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 12147 + }, + { + "epoch": 0.4299548872180451, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.23, + "step": 12148 + }, + { + "epoch": 0.43001503759398496, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1707, + "step": 12149 + }, + { + "epoch": 0.43007518796992483, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 12150 + }, + { + "epoch": 0.43013533834586465, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 12151 + }, + { + "epoch": 0.4301954887218045, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1719, + "step": 12152 + }, + { + "epoch": 0.43025563909774434, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 12153 + }, + { + "epoch": 0.4303157894736842, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.4005, + "step": 12154 + }, + { + "epoch": 0.4303759398496241, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1836, + "step": 12155 + }, + { + "epoch": 0.4304360902255639, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 12156 + }, + { + "epoch": 0.43049624060150377, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 12157 + }, + { + "epoch": 0.4305563909774436, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2868, + "step": 12158 + }, + { + "epoch": 0.43061654135338345, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.3279, + "step": 12159 + }, + { + "epoch": 0.4306766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 12160 + }, + { + "epoch": 0.43073684210526314, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2908, + "step": 12161 + }, + { + "epoch": 0.430796992481203, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 12162 + }, + { + "epoch": 0.4308571428571429, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 12163 + }, + { + "epoch": 0.4309172932330827, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 12164 + }, + { + "epoch": 0.43097744360902257, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2322, + "step": 12165 + }, + { + "epoch": 0.4310375939849624, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 12166 + }, + { + "epoch": 0.43109774436090226, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 12167 + }, + { + "epoch": 0.43115789473684213, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 12168 + }, + { + "epoch": 0.43121804511278194, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2138, + "step": 12169 + }, + { + "epoch": 0.4312781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 12170 + }, + { + "epoch": 0.43133834586466163, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 12171 + }, + { + "epoch": 0.4313984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 12172 + }, + { + "epoch": 0.4314586466165414, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 12173 + }, + { + "epoch": 0.4315187969924812, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 12174 + }, + { + "epoch": 0.43157894736842106, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 12175 + }, + { + "epoch": 0.4316390977443609, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2743, + "step": 12176 + }, + { + "epoch": 0.43169924812030075, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 12177 + }, + { + "epoch": 0.4317593984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1918, + "step": 12178 + }, + { + "epoch": 0.43181954887218044, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 12179 + }, + { + "epoch": 0.4318796992481203, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 12180 + }, + { + "epoch": 0.4319398496240601, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 12181 + }, + { + "epoch": 0.432, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 12182 + }, + { + "epoch": 0.43206015037593987, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 12183 + }, + { + "epoch": 0.4321203007518797, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2095, + "step": 12184 + }, + { + "epoch": 0.43218045112781955, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2483, + "step": 12185 + }, + { + "epoch": 0.4322406015037594, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 12186 + }, + { + "epoch": 0.43230075187969924, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 12187 + }, + { + "epoch": 0.4323609022556391, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 12188 + }, + { + "epoch": 0.43242105263157893, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 12189 + }, + { + "epoch": 0.4324812030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 12190 + }, + { + "epoch": 0.43254135338345867, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 12191 + }, + { + "epoch": 0.4326015037593985, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 12192 + }, + { + "epoch": 0.43266165413533836, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 12193 + }, + { + "epoch": 0.4327218045112782, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.3088, + "step": 12194 + }, + { + "epoch": 0.43278195488721805, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2175, + "step": 12195 + }, + { + "epoch": 0.4328421052631579, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 12196 + }, + { + "epoch": 0.43290225563909773, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 12197 + }, + { + "epoch": 0.4329624060150376, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 12198 + }, + { + "epoch": 0.4330225563909774, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 12199 + }, + { + "epoch": 0.4330827067669173, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2705, + "step": 12200 + }, + { + "epoch": 0.43314285714285716, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2549, + "step": 12201 + }, + { + "epoch": 0.433203007518797, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2586, + "step": 12202 + }, + { + "epoch": 0.43326315789473685, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 12203 + }, + { + "epoch": 0.43332330827067667, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2994, + "step": 12204 + }, + { + "epoch": 0.43338345864661654, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 12205 + }, + { + "epoch": 0.4334436090225564, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2493, + "step": 12206 + }, + { + "epoch": 0.4335037593984962, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 12207 + }, + { + "epoch": 0.4335639097744361, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 12208 + }, + { + "epoch": 0.43362406015037597, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 12209 + }, + { + "epoch": 0.4336842105263158, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 12210 + }, + { + "epoch": 0.43374436090225565, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 12211 + }, + { + "epoch": 0.43380451127819547, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 12212 + }, + { + "epoch": 0.43386466165413534, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 12213 + }, + { + "epoch": 0.4339248120300752, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3762, + "step": 12214 + }, + { + "epoch": 0.43398496240601503, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3523, + "step": 12215 + }, + { + "epoch": 0.4340451127819549, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 12216 + }, + { + "epoch": 0.4341052631578947, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 12217 + }, + { + "epoch": 0.4341654135338346, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2201, + "step": 12218 + }, + { + "epoch": 0.43422556390977446, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 12219 + }, + { + "epoch": 0.4342857142857143, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 12220 + }, + { + "epoch": 0.43434586466165415, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 12221 + }, + { + "epoch": 0.43440601503759396, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1465, + "step": 12222 + }, + { + "epoch": 0.43446616541353383, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 12223 + }, + { + "epoch": 0.4345263157894737, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 12224 + }, + { + "epoch": 0.4345864661654135, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 12225 + }, + { + "epoch": 0.4346466165413534, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 12226 + }, + { + "epoch": 0.4347067669172932, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 12227 + }, + { + "epoch": 0.4347669172932331, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2122, + "step": 12228 + }, + { + "epoch": 0.43482706766917295, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 12229 + }, + { + "epoch": 0.43488721804511277, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 12230 + }, + { + "epoch": 0.43494736842105264, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 12231 + }, + { + "epoch": 0.43500751879699245, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 12232 + }, + { + "epoch": 0.4350676691729323, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2981, + "step": 12233 + }, + { + "epoch": 0.4351278195488722, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 12234 + }, + { + "epoch": 0.435187969924812, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 12235 + }, + { + "epoch": 0.4352481203007519, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 12236 + }, + { + "epoch": 0.43530827067669176, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 12237 + }, + { + "epoch": 0.43536842105263157, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 12238 + }, + { + "epoch": 0.43542857142857144, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 12239 + }, + { + "epoch": 0.43548872180451126, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2095, + "step": 12240 + }, + { + "epoch": 0.43554887218045113, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 12241 + }, + { + "epoch": 0.435609022556391, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 12242 + }, + { + "epoch": 0.4356691729323308, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2287, + "step": 12243 + }, + { + "epoch": 0.4357293233082707, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 12244 + }, + { + "epoch": 0.4357894736842105, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 12245 + }, + { + "epoch": 0.4358496240601504, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2143, + "step": 12246 + }, + { + "epoch": 0.43590977443609025, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 12247 + }, + { + "epoch": 0.43596992481203006, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 12248 + }, + { + "epoch": 0.43603007518796993, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.156, + "step": 12249 + }, + { + "epoch": 0.43609022556390975, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3023, + "step": 12250 + }, + { + "epoch": 0.4361503759398496, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2144, + "step": 12251 + }, + { + "epoch": 0.4362105263157895, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 12252 + }, + { + "epoch": 0.4362706766917293, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.195, + "step": 12253 + }, + { + "epoch": 0.4363308270676692, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.2503, + "step": 12254 + }, + { + "epoch": 0.436390977443609, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 12255 + }, + { + "epoch": 0.43645112781954887, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 12256 + }, + { + "epoch": 0.43651127819548874, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 12257 + }, + { + "epoch": 0.43657142857142855, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1983, + "step": 12258 + }, + { + "epoch": 0.4366315789473684, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 12259 + }, + { + "epoch": 0.4366917293233083, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1518, + "step": 12260 + }, + { + "epoch": 0.4367518796992481, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 12261 + }, + { + "epoch": 0.436812030075188, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2476, + "step": 12262 + }, + { + "epoch": 0.4368721804511278, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 12263 + }, + { + "epoch": 0.43693233082706767, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 12264 + }, + { + "epoch": 0.43699248120300754, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 12265 + }, + { + "epoch": 0.43705263157894736, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2114, + "step": 12266 + }, + { + "epoch": 0.43711278195488723, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.1793, + "step": 12267 + }, + { + "epoch": 0.43717293233082705, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 12268 + }, + { + "epoch": 0.4372330827067669, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 12269 + }, + { + "epoch": 0.4372932330827068, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 12270 + }, + { + "epoch": 0.4373533834586466, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 12271 + }, + { + "epoch": 0.4374135338345865, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 12272 + }, + { + "epoch": 0.4374736842105263, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 12273 + }, + { + "epoch": 0.43753383458646616, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2036, + "step": 12274 + }, + { + "epoch": 0.43759398496240604, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1444, + "step": 12275 + }, + { + "epoch": 0.43765413533834585, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.215, + "step": 12276 + }, + { + "epoch": 0.4377142857142857, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2305, + "step": 12277 + }, + { + "epoch": 0.43777443609022554, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 12278 + }, + { + "epoch": 0.4378345864661654, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 12279 + }, + { + "epoch": 0.4378947368421053, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1763, + "step": 12280 + }, + { + "epoch": 0.4379548872180451, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1518, + "step": 12281 + }, + { + "epoch": 0.43801503759398497, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 12282 + }, + { + "epoch": 0.43807518796992484, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2264, + "step": 12283 + }, + { + "epoch": 0.43813533834586466, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2253, + "step": 12284 + }, + { + "epoch": 0.4381954887218045, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 12285 + }, + { + "epoch": 0.43825563909774434, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 12286 + }, + { + "epoch": 0.4383157894736842, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 12287 + }, + { + "epoch": 0.4383759398496241, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 12288 + }, + { + "epoch": 0.4384360902255639, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 12289 + }, + { + "epoch": 0.4384962406015038, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 12290 + }, + { + "epoch": 0.4385563909774436, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 12291 + }, + { + "epoch": 0.43861654135338346, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 12292 + }, + { + "epoch": 0.43867669172932333, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 12293 + }, + { + "epoch": 0.43873684210526315, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2346, + "step": 12294 + }, + { + "epoch": 0.438796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 12295 + }, + { + "epoch": 0.43885714285714283, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 12296 + }, + { + "epoch": 0.4389172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 12297 + }, + { + "epoch": 0.4389774436090226, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1689, + "step": 12298 + }, + { + "epoch": 0.4390375939849624, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 12299 + }, + { + "epoch": 0.43909774436090226, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 12300 + }, + { + "epoch": 0.4391578947368421, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 12301 + }, + { + "epoch": 0.43921804511278195, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1882, + "step": 12302 + }, + { + "epoch": 0.4392781954887218, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1985, + "step": 12303 + }, + { + "epoch": 0.43933834586466164, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 12304 + }, + { + "epoch": 0.4393984962406015, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 12305 + }, + { + "epoch": 0.4394586466165413, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 12306 + }, + { + "epoch": 0.4395187969924812, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 12307 + }, + { + "epoch": 0.43957894736842107, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 12308 + }, + { + "epoch": 0.4396390977443609, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 12309 + }, + { + "epoch": 0.43969924812030076, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 12310 + }, + { + "epoch": 0.4397593984962406, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1773, + "step": 12311 + }, + { + "epoch": 0.43981954887218044, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 12312 + }, + { + "epoch": 0.4398796992481203, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 12313 + }, + { + "epoch": 0.43993984962406013, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2456, + "step": 12314 + }, + { + "epoch": 0.44, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2148, + "step": 12315 + }, + { + "epoch": 0.4400601503759399, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 12316 + }, + { + "epoch": 0.4401203007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 12317 + }, + { + "epoch": 0.44018045112781956, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.3175, + "step": 12318 + }, + { + "epoch": 0.4402406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 12319 + }, + { + "epoch": 0.44030075187969925, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2002, + "step": 12320 + }, + { + "epoch": 0.4403609022556391, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 12321 + }, + { + "epoch": 0.44042105263157894, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 12322 + }, + { + "epoch": 0.4404812030075188, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2286, + "step": 12323 + }, + { + "epoch": 0.4405413533834586, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 12324 + }, + { + "epoch": 0.4406015037593985, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 12325 + }, + { + "epoch": 0.44066165413533837, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 12326 + }, + { + "epoch": 0.4407218045112782, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1662, + "step": 12327 + }, + { + "epoch": 0.44078195488721805, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 12328 + }, + { + "epoch": 0.44084210526315787, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2632, + "step": 12329 + }, + { + "epoch": 0.44090225563909774, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.237, + "step": 12330 + }, + { + "epoch": 0.4409624060150376, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1837, + "step": 12331 + }, + { + "epoch": 0.4410225563909774, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 12332 + }, + { + "epoch": 0.4410827067669173, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 12333 + }, + { + "epoch": 0.44114285714285717, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 12334 + }, + { + "epoch": 0.441203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1876, + "step": 12335 + }, + { + "epoch": 0.44126315789473686, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.468, + "step": 12336 + }, + { + "epoch": 0.4413233082706767, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 12337 + }, + { + "epoch": 0.44138345864661654, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 12338 + }, + { + "epoch": 0.4414436090225564, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 12339 + }, + { + "epoch": 0.44150375939849623, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 12340 + }, + { + "epoch": 0.4415639097744361, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2032, + "step": 12341 + }, + { + "epoch": 0.4416240601503759, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 12342 + }, + { + "epoch": 0.4416842105263158, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 12343 + }, + { + "epoch": 0.44174436090225566, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1972, + "step": 12344 + }, + { + "epoch": 0.4418045112781955, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 12345 + }, + { + "epoch": 0.44186466165413535, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2029, + "step": 12346 + }, + { + "epoch": 0.44192481203007516, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 12347 + }, + { + "epoch": 0.44198496240601504, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2046, + "step": 12348 + }, + { + "epoch": 0.4420451127819549, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 12349 + }, + { + "epoch": 0.4421052631578947, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 12350 + }, + { + "epoch": 0.4421654135338346, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 12351 + }, + { + "epoch": 0.4422255639097744, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 12352 + }, + { + "epoch": 0.4422857142857143, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2143, + "step": 12353 + }, + { + "epoch": 0.44234586466165415, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 12354 + }, + { + "epoch": 0.44240601503759397, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2676, + "step": 12355 + }, + { + "epoch": 0.44246616541353384, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 12356 + }, + { + "epoch": 0.44252631578947366, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.3307, + "step": 12357 + }, + { + "epoch": 0.44258646616541353, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 12358 + }, + { + "epoch": 0.4426466165413534, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1925, + "step": 12359 + }, + { + "epoch": 0.4427067669172932, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 12360 + }, + { + "epoch": 0.4427669172932331, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2486, + "step": 12361 + }, + { + "epoch": 0.44282706766917296, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 12362 + }, + { + "epoch": 0.4428872180451128, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 12363 + }, + { + "epoch": 0.44294736842105265, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.3065, + "step": 12364 + }, + { + "epoch": 0.44300751879699246, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 12365 + }, + { + "epoch": 0.44306766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 12366 + }, + { + "epoch": 0.4431278195488722, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 12367 + }, + { + "epoch": 0.443187969924812, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.275, + "step": 12368 + }, + { + "epoch": 0.4432481203007519, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 12369 + }, + { + "epoch": 0.4433082706766917, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 12370 + }, + { + "epoch": 0.4433684210526316, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 12371 + }, + { + "epoch": 0.44342857142857145, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 12372 + }, + { + "epoch": 0.44348872180451127, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1582, + "step": 12373 + }, + { + "epoch": 0.44354887218045114, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 12374 + }, + { + "epoch": 0.44360902255639095, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2173, + "step": 12375 + }, + { + "epoch": 0.4436691729323308, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2998, + "step": 12376 + }, + { + "epoch": 0.4437293233082707, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 12377 + }, + { + "epoch": 0.4437894736842105, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 12378 + }, + { + "epoch": 0.4438496240601504, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 12379 + }, + { + "epoch": 0.4439097744360902, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2112, + "step": 12380 + }, + { + "epoch": 0.44396992481203007, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 12381 + }, + { + "epoch": 0.44403007518796994, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2064, + "step": 12382 + }, + { + "epoch": 0.44409022556390976, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 12383 + }, + { + "epoch": 0.44415037593984963, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 12384 + }, + { + "epoch": 0.4442105263157895, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 12385 + }, + { + "epoch": 0.4442706766917293, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1746, + "step": 12386 + }, + { + "epoch": 0.4443308270676692, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 12387 + }, + { + "epoch": 0.444390977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 12388 + }, + { + "epoch": 0.4444511278195489, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 12389 + }, + { + "epoch": 0.44451127819548875, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 12390 + }, + { + "epoch": 0.44457142857142856, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2225, + "step": 12391 + }, + { + "epoch": 0.44463157894736843, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 12392 + }, + { + "epoch": 0.44469172932330825, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 12393 + }, + { + "epoch": 0.4447518796992481, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 12394 + }, + { + "epoch": 0.444812030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 12395 + }, + { + "epoch": 0.4448721804511278, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 12396 + }, + { + "epoch": 0.4449323308270677, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 12397 + }, + { + "epoch": 0.4449924812030075, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 12398 + }, + { + "epoch": 0.44505263157894737, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2186, + "step": 12399 + }, + { + "epoch": 0.44511278195488724, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 12400 + }, + { + "epoch": 0.44517293233082705, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 12401 + }, + { + "epoch": 0.4452330827067669, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1949, + "step": 12402 + }, + { + "epoch": 0.44529323308270674, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 12403 + }, + { + "epoch": 0.4453533834586466, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1917, + "step": 12404 + }, + { + "epoch": 0.4454135338345865, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 12405 + }, + { + "epoch": 0.4454736842105263, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 12406 + }, + { + "epoch": 0.44553383458646617, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 12407 + }, + { + "epoch": 0.44559398496240604, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 12408 + }, + { + "epoch": 0.44565413533834586, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2439, + "step": 12409 + }, + { + "epoch": 0.44571428571428573, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1464, + "step": 12410 + }, + { + "epoch": 0.44577443609022555, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1667, + "step": 12411 + }, + { + "epoch": 0.4458345864661654, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2369, + "step": 12412 + }, + { + "epoch": 0.4458947368421053, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.1518, + "step": 12413 + }, + { + "epoch": 0.4459548872180451, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 12414 + }, + { + "epoch": 0.446015037593985, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.258, + "step": 12415 + }, + { + "epoch": 0.4460751879699248, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2163, + "step": 12416 + }, + { + "epoch": 0.44613533834586466, + "grad_norm": 0.439453125, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 12417 + }, + { + "epoch": 0.44619548872180453, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1757, + "step": 12418 + }, + { + "epoch": 0.44625563909774435, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 12419 + }, + { + "epoch": 0.4463157894736842, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 12420 + }, + { + "epoch": 0.44637593984962404, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2062, + "step": 12421 + }, + { + "epoch": 0.4464360902255639, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2044, + "step": 12422 + }, + { + "epoch": 0.4464962406015038, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1721, + "step": 12423 + }, + { + "epoch": 0.4465563909774436, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 12424 + }, + { + "epoch": 0.44661654135338347, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 12425 + }, + { + "epoch": 0.4466766917293233, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1714, + "step": 12426 + }, + { + "epoch": 0.44673684210526315, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 12427 + }, + { + "epoch": 0.446796992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 12428 + }, + { + "epoch": 0.44685714285714284, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 12429 + }, + { + "epoch": 0.4469172932330827, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 12430 + }, + { + "epoch": 0.44697744360902253, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 12431 + }, + { + "epoch": 0.4470375939849624, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 12432 + }, + { + "epoch": 0.44709774436090227, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 12433 + }, + { + "epoch": 0.4471578947368421, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 12434 + }, + { + "epoch": 0.44721804511278196, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 12435 + }, + { + "epoch": 0.44727819548872183, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 12436 + }, + { + "epoch": 0.44733834586466165, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2892, + "step": 12437 + }, + { + "epoch": 0.4473984962406015, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 12438 + }, + { + "epoch": 0.44745864661654133, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 12439 + }, + { + "epoch": 0.4475187969924812, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1526, + "step": 12440 + }, + { + "epoch": 0.4475789473684211, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2575, + "step": 12441 + }, + { + "epoch": 0.4476390977443609, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2355, + "step": 12442 + }, + { + "epoch": 0.44769924812030076, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 12443 + }, + { + "epoch": 0.4477593984962406, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 12444 + }, + { + "epoch": 0.44781954887218045, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 12445 + }, + { + "epoch": 0.4478796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 12446 + }, + { + "epoch": 0.44793984962406014, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 12447 + }, + { + "epoch": 0.448, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2465, + "step": 12448 + }, + { + "epoch": 0.4480601503759398, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 12449 + }, + { + "epoch": 0.4481203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 12450 + }, + { + "epoch": 0.44818045112781957, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 12451 + }, + { + "epoch": 0.4482406015037594, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1923, + "step": 12452 + }, + { + "epoch": 0.44830075187969926, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 12453 + }, + { + "epoch": 0.44836090225563907, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 12454 + }, + { + "epoch": 0.44842105263157894, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2414, + "step": 12455 + }, + { + "epoch": 0.4484812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 12456 + }, + { + "epoch": 0.44854135338345863, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 12457 + }, + { + "epoch": 0.4486015037593985, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2018, + "step": 12458 + }, + { + "epoch": 0.4486616541353384, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 12459 + }, + { + "epoch": 0.4487218045112782, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 12460 + }, + { + "epoch": 0.44878195488721806, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 12461 + }, + { + "epoch": 0.4488421052631579, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 12462 + }, + { + "epoch": 0.44890225563909775, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1989, + "step": 12463 + }, + { + "epoch": 0.4489624060150376, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 12464 + }, + { + "epoch": 0.44902255639097743, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 12465 + }, + { + "epoch": 0.4490827067669173, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 12466 + }, + { + "epoch": 0.4491428571428571, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 12467 + }, + { + "epoch": 0.449203007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 12468 + }, + { + "epoch": 0.44926315789473686, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 12469 + }, + { + "epoch": 0.4493233082706767, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2351, + "step": 12470 + }, + { + "epoch": 0.44938345864661655, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 12471 + }, + { + "epoch": 0.44944360902255637, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2242, + "step": 12472 + }, + { + "epoch": 0.44950375939849624, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 12473 + }, + { + "epoch": 0.4495639097744361, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 12474 + }, + { + "epoch": 0.4496240601503759, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 12475 + }, + { + "epoch": 0.4496842105263158, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 12476 + }, + { + "epoch": 0.4497443609022556, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 12477 + }, + { + "epoch": 0.4498045112781955, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 12478 + }, + { + "epoch": 0.44986466165413536, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 12479 + }, + { + "epoch": 0.44992481203007517, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 12480 + }, + { + "epoch": 0.44998496240601504, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 12481 + }, + { + "epoch": 0.4500451127819549, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 12482 + }, + { + "epoch": 0.45010526315789473, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.228, + "step": 12483 + }, + { + "epoch": 0.4501654135338346, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 12484 + }, + { + "epoch": 0.4502255639097744, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 12485 + }, + { + "epoch": 0.4502857142857143, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 12486 + }, + { + "epoch": 0.45034586466165416, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 12487 + }, + { + "epoch": 0.450406015037594, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 12488 + }, + { + "epoch": 0.45046616541353385, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1726, + "step": 12489 + }, + { + "epoch": 0.45052631578947366, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.233, + "step": 12490 + }, + { + "epoch": 0.45058646616541354, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2993, + "step": 12491 + }, + { + "epoch": 0.4506466165413534, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 12492 + }, + { + "epoch": 0.4507067669172932, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2101, + "step": 12493 + }, + { + "epoch": 0.4507669172932331, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 12494 + }, + { + "epoch": 0.4508270676691729, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 12495 + }, + { + "epoch": 0.4508872180451128, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 12496 + }, + { + "epoch": 0.45094736842105265, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 12497 + }, + { + "epoch": 0.45100751879699247, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 12498 + }, + { + "epoch": 0.45106766917293234, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 12499 + }, + { + "epoch": 0.45112781954887216, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.373, + "step": 12500 + }, + { + "epoch": 0.451187969924812, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 12501 + }, + { + "epoch": 0.4512481203007519, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 12502 + }, + { + "epoch": 0.4513082706766917, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 12503 + }, + { + "epoch": 0.4513684210526316, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 12504 + }, + { + "epoch": 0.4514285714285714, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 12505 + }, + { + "epoch": 0.4514887218045113, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 12506 + }, + { + "epoch": 0.45154887218045114, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2564, + "step": 12507 + }, + { + "epoch": 0.45160902255639096, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2079, + "step": 12508 + }, + { + "epoch": 0.45166917293233083, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 12509 + }, + { + "epoch": 0.4517293233082707, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1691, + "step": 12510 + }, + { + "epoch": 0.4517894736842105, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 12511 + }, + { + "epoch": 0.4518496240601504, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 12512 + }, + { + "epoch": 0.4519097744360902, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 12513 + }, + { + "epoch": 0.4519699248120301, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 12514 + }, + { + "epoch": 0.45203007518796995, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 12515 + }, + { + "epoch": 0.45209022556390976, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 12516 + }, + { + "epoch": 0.45215037593984964, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 12517 + }, + { + "epoch": 0.45221052631578945, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 12518 + }, + { + "epoch": 0.4522706766917293, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 12519 + }, + { + "epoch": 0.4523308270676692, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1998, + "step": 12520 + }, + { + "epoch": 0.452390977443609, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2264, + "step": 12521 + }, + { + "epoch": 0.4524511278195489, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 12522 + }, + { + "epoch": 0.4525112781954887, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1333, + "step": 12523 + }, + { + "epoch": 0.45257142857142857, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1896, + "step": 12524 + }, + { + "epoch": 0.45263157894736844, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 12525 + }, + { + "epoch": 0.45269172932330826, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 12526 + }, + { + "epoch": 0.4527518796992481, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 12527 + }, + { + "epoch": 0.45281203007518794, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 12528 + }, + { + "epoch": 0.4528721804511278, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 12529 + }, + { + "epoch": 0.4529323308270677, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 12530 + }, + { + "epoch": 0.4529924812030075, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 12531 + }, + { + "epoch": 0.4530526315789474, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 12532 + }, + { + "epoch": 0.45311278195488724, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 12533 + }, + { + "epoch": 0.45317293233082706, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 12534 + }, + { + "epoch": 0.45323308270676693, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2144, + "step": 12535 + }, + { + "epoch": 0.45329323308270675, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2452, + "step": 12536 + }, + { + "epoch": 0.4533533834586466, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 12537 + }, + { + "epoch": 0.4534135338345865, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.1264, + "step": 12538 + }, + { + "epoch": 0.4534736842105263, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2552, + "step": 12539 + }, + { + "epoch": 0.4535338345864662, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 12540 + }, + { + "epoch": 0.453593984962406, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 12541 + }, + { + "epoch": 0.45365413533834587, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 12542 + }, + { + "epoch": 0.45371428571428574, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 12543 + }, + { + "epoch": 0.45377443609022555, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 12544 + }, + { + "epoch": 0.4538345864661654, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 12545 + }, + { + "epoch": 0.45389473684210524, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 12546 + }, + { + "epoch": 0.4539548872180451, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2379, + "step": 12547 + }, + { + "epoch": 0.454015037593985, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 12548 + }, + { + "epoch": 0.4540751879699248, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2603, + "step": 12549 + }, + { + "epoch": 0.45413533834586467, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.271, + "step": 12550 + }, + { + "epoch": 0.4541954887218045, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 12551 + }, + { + "epoch": 0.45425563909774436, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 12552 + }, + { + "epoch": 0.45431578947368423, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 12553 + }, + { + "epoch": 0.45437593984962404, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 12554 + }, + { + "epoch": 0.4544360902255639, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 12555 + }, + { + "epoch": 0.4544962406015038, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 12556 + }, + { + "epoch": 0.4545563909774436, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 12557 + }, + { + "epoch": 0.4546165413533835, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 12558 + }, + { + "epoch": 0.4546766917293233, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 12559 + }, + { + "epoch": 0.45473684210526316, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 12560 + }, + { + "epoch": 0.45479699248120303, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 12561 + }, + { + "epoch": 0.45485714285714285, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 12562 + }, + { + "epoch": 0.4549172932330827, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 12563 + }, + { + "epoch": 0.45497744360902254, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 12564 + }, + { + "epoch": 0.4550375939849624, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 12565 + }, + { + "epoch": 0.4550977443609023, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 12566 + }, + { + "epoch": 0.4551578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2005, + "step": 12567 + }, + { + "epoch": 0.45521804511278197, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 12568 + }, + { + "epoch": 0.4552781954887218, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 12569 + }, + { + "epoch": 0.45533834586466165, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 12570 + }, + { + "epoch": 0.4553984962406015, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 12571 + }, + { + "epoch": 0.45545864661654134, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 12572 + }, + { + "epoch": 0.4555187969924812, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 12573 + }, + { + "epoch": 0.45557894736842103, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 12574 + }, + { + "epoch": 0.4556390977443609, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2726, + "step": 12575 + }, + { + "epoch": 0.45569924812030077, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1914, + "step": 12576 + }, + { + "epoch": 0.4557593984962406, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 12577 + }, + { + "epoch": 0.45581954887218046, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 12578 + }, + { + "epoch": 0.4558796992481203, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 12579 + }, + { + "epoch": 0.45593984962406015, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 12580 + }, + { + "epoch": 0.456, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 12581 + }, + { + "epoch": 0.45606015037593983, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 12582 + }, + { + "epoch": 0.4561203007518797, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 12583 + }, + { + "epoch": 0.4561804511278196, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2166, + "step": 12584 + }, + { + "epoch": 0.4562406015037594, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2578, + "step": 12585 + }, + { + "epoch": 0.45630075187969926, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 12586 + }, + { + "epoch": 0.4563609022556391, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2522, + "step": 12587 + }, + { + "epoch": 0.45642105263157895, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.3081, + "step": 12588 + }, + { + "epoch": 0.4564812030075188, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 12589 + }, + { + "epoch": 0.45654135338345864, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 12590 + }, + { + "epoch": 0.4566015037593985, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 12591 + }, + { + "epoch": 0.4566616541353383, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 12592 + }, + { + "epoch": 0.4567218045112782, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1582, + "step": 12593 + }, + { + "epoch": 0.45678195488721807, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 12594 + }, + { + "epoch": 0.4568421052631579, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 12595 + }, + { + "epoch": 0.45690225563909775, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 12596 + }, + { + "epoch": 0.45696240601503757, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 12597 + }, + { + "epoch": 0.45702255639097744, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 12598 + }, + { + "epoch": 0.4570827067669173, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 12599 + }, + { + "epoch": 0.45714285714285713, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 12600 + }, + { + "epoch": 0.457203007518797, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2591, + "step": 12601 + }, + { + "epoch": 0.4572631578947368, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 12602 + }, + { + "epoch": 0.4573233082706767, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 12603 + }, + { + "epoch": 0.45738345864661656, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 12604 + }, + { + "epoch": 0.4574436090225564, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 12605 + }, + { + "epoch": 0.45750375939849625, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 12606 + }, + { + "epoch": 0.4575639097744361, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 12607 + }, + { + "epoch": 0.45762406015037593, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2184, + "step": 12608 + }, + { + "epoch": 0.4576842105263158, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 12609 + }, + { + "epoch": 0.4577443609022556, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 12610 + }, + { + "epoch": 0.4578045112781955, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 12611 + }, + { + "epoch": 0.45786466165413536, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1832, + "step": 12612 + }, + { + "epoch": 0.4579248120300752, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 12613 + }, + { + "epoch": 0.45798496240601505, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 12614 + }, + { + "epoch": 0.45804511278195487, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2635, + "step": 12615 + }, + { + "epoch": 0.45810526315789474, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1567, + "step": 12616 + }, + { + "epoch": 0.4581654135338346, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 12617 + }, + { + "epoch": 0.4582255639097744, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2167, + "step": 12618 + }, + { + "epoch": 0.4582857142857143, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 12619 + }, + { + "epoch": 0.4583458646616541, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 12620 + }, + { + "epoch": 0.458406015037594, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.2823, + "step": 12621 + }, + { + "epoch": 0.45846616541353385, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 12622 + }, + { + "epoch": 0.45852631578947367, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 12623 + }, + { + "epoch": 0.45858646616541354, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 12624 + }, + { + "epoch": 0.45864661654135336, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1808, + "step": 12625 + }, + { + "epoch": 0.45870676691729323, + "grad_norm": 1.6875, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 12626 + }, + { + "epoch": 0.4587669172932331, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 12627 + }, + { + "epoch": 0.4588270676691729, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 12628 + }, + { + "epoch": 0.4588872180451128, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 12629 + }, + { + "epoch": 0.4589473684210526, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 12630 + }, + { + "epoch": 0.4590075187969925, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2792, + "step": 12631 + }, + { + "epoch": 0.45906766917293235, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 12632 + }, + { + "epoch": 0.45912781954887216, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 12633 + }, + { + "epoch": 0.45918796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 12634 + }, + { + "epoch": 0.4592481203007519, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 12635 + }, + { + "epoch": 0.4593082706766917, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2085, + "step": 12636 + }, + { + "epoch": 0.4593684210526316, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2252, + "step": 12637 + }, + { + "epoch": 0.4594285714285714, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2484, + "step": 12638 + }, + { + "epoch": 0.4594887218045113, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 12639 + }, + { + "epoch": 0.45954887218045115, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 12640 + }, + { + "epoch": 0.45960902255639097, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 12641 + }, + { + "epoch": 0.45966917293233084, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 12642 + }, + { + "epoch": 0.45972932330827065, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 12643 + }, + { + "epoch": 0.4597894736842105, + "grad_norm": 0.458984375, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 12644 + }, + { + "epoch": 0.4598496240601504, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2291, + "step": 12645 + }, + { + "epoch": 0.4599097744360902, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 12646 + }, + { + "epoch": 0.4599699248120301, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 12647 + }, + { + "epoch": 0.4600300751879699, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 12648 + }, + { + "epoch": 0.46009022556390977, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 12649 + }, + { + "epoch": 0.46015037593984964, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 12650 + }, + { + "epoch": 0.46021052631578946, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 12651 + }, + { + "epoch": 0.46027067669172933, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 12652 + }, + { + "epoch": 0.46033082706766915, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2322, + "step": 12653 + }, + { + "epoch": 0.460390977443609, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 12654 + }, + { + "epoch": 0.4604511278195489, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 12655 + }, + { + "epoch": 0.4605112781954887, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2284, + "step": 12656 + }, + { + "epoch": 0.4605714285714286, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 12657 + }, + { + "epoch": 0.46063157894736845, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1905, + "step": 12658 + }, + { + "epoch": 0.46069172932330826, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 12659 + }, + { + "epoch": 0.46075187969924813, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 12660 + }, + { + "epoch": 0.46081203007518795, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 12661 + }, + { + "epoch": 0.4608721804511278, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2371, + "step": 12662 + }, + { + "epoch": 0.4609323308270677, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 12663 + }, + { + "epoch": 0.4609924812030075, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 12664 + }, + { + "epoch": 0.4610526315789474, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2323, + "step": 12665 + }, + { + "epoch": 0.4611127819548872, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 12666 + }, + { + "epoch": 0.46117293233082707, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 12667 + }, + { + "epoch": 0.46123308270676694, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 12668 + }, + { + "epoch": 0.46129323308270676, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.221, + "step": 12669 + }, + { + "epoch": 0.4613533834586466, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 12670 + }, + { + "epoch": 0.46141353383458644, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 12671 + }, + { + "epoch": 0.4614736842105263, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 12672 + }, + { + "epoch": 0.4615338345864662, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1512, + "step": 12673 + }, + { + "epoch": 0.461593984962406, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 12674 + }, + { + "epoch": 0.4616541353383459, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2717, + "step": 12675 + }, + { + "epoch": 0.4617142857142857, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 12676 + }, + { + "epoch": 0.46177443609022556, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 12677 + }, + { + "epoch": 0.46183458646616543, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 12678 + }, + { + "epoch": 0.46189473684210525, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2276, + "step": 12679 + }, + { + "epoch": 0.4619548872180451, + "grad_norm": 1.6015625, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 12680 + }, + { + "epoch": 0.462015037593985, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.272, + "step": 12681 + }, + { + "epoch": 0.4620751879699248, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1943, + "step": 12682 + }, + { + "epoch": 0.4621353383458647, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 12683 + }, + { + "epoch": 0.4621954887218045, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 12684 + }, + { + "epoch": 0.46225563909774436, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.259, + "step": 12685 + }, + { + "epoch": 0.46231578947368424, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 12686 + }, + { + "epoch": 0.46237593984962405, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 12687 + }, + { + "epoch": 0.4624360902255639, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 12688 + }, + { + "epoch": 0.46249624060150374, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 12689 + }, + { + "epoch": 0.4625563909774436, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 12690 + }, + { + "epoch": 0.4626165413533835, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1604, + "step": 12691 + }, + { + "epoch": 0.4626766917293233, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 12692 + }, + { + "epoch": 0.46273684210526317, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2118, + "step": 12693 + }, + { + "epoch": 0.462796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3101, + "step": 12694 + }, + { + "epoch": 0.46285714285714286, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 12695 + }, + { + "epoch": 0.4629172932330827, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2038, + "step": 12696 + }, + { + "epoch": 0.46297744360902254, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 12697 + }, + { + "epoch": 0.4630375939849624, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 12698 + }, + { + "epoch": 0.46309774436090223, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 12699 + }, + { + "epoch": 0.4631578947368421, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2149, + "step": 12700 + }, + { + "epoch": 0.463218045112782, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2113, + "step": 12701 + }, + { + "epoch": 0.4632781954887218, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 12702 + }, + { + "epoch": 0.46333834586466166, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 12703 + }, + { + "epoch": 0.4633984962406015, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1955, + "step": 12704 + }, + { + "epoch": 0.46345864661654135, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 12705 + }, + { + "epoch": 0.4635187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2854, + "step": 12706 + }, + { + "epoch": 0.46357894736842103, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 12707 + }, + { + "epoch": 0.4636390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 12708 + }, + { + "epoch": 0.4636992481203008, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1915, + "step": 12709 + }, + { + "epoch": 0.4637593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 12710 + }, + { + "epoch": 0.46381954887218046, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 12711 + }, + { + "epoch": 0.4638796992481203, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 12712 + }, + { + "epoch": 0.46393984962406015, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 12713 + }, + { + "epoch": 0.464, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 12714 + }, + { + "epoch": 0.46406015037593984, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1823, + "step": 12715 + }, + { + "epoch": 0.4641203007518797, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1632, + "step": 12716 + }, + { + "epoch": 0.4641804511278195, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2618, + "step": 12717 + }, + { + "epoch": 0.4642406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 12718 + }, + { + "epoch": 0.46430075187969927, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1714, + "step": 12719 + }, + { + "epoch": 0.4643609022556391, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 12720 + }, + { + "epoch": 0.46442105263157896, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 12721 + }, + { + "epoch": 0.4644812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2178, + "step": 12722 + }, + { + "epoch": 0.46454135338345864, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2136, + "step": 12723 + }, + { + "epoch": 0.4646015037593985, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2173, + "step": 12724 + }, + { + "epoch": 0.46466165413533833, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2362, + "step": 12725 + }, + { + "epoch": 0.4647218045112782, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 12726 + }, + { + "epoch": 0.464781954887218, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2976, + "step": 12727 + }, + { + "epoch": 0.4648421052631579, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 12728 + }, + { + "epoch": 0.46490225563909776, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 12729 + }, + { + "epoch": 0.4649624060150376, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 12730 + }, + { + "epoch": 0.46502255639097745, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 12731 + }, + { + "epoch": 0.4650827067669173, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.301, + "step": 12732 + }, + { + "epoch": 0.46514285714285714, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2545, + "step": 12733 + }, + { + "epoch": 0.465203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2838, + "step": 12734 + }, + { + "epoch": 0.4652631578947368, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2441, + "step": 12735 + }, + { + "epoch": 0.4653233082706767, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 12736 + }, + { + "epoch": 0.46538345864661657, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 12737 + }, + { + "epoch": 0.4654436090225564, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 12738 + }, + { + "epoch": 0.46550375939849625, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 12739 + }, + { + "epoch": 0.46556390977443607, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2478, + "step": 12740 + }, + { + "epoch": 0.46562406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 12741 + }, + { + "epoch": 0.4656842105263158, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 12742 + }, + { + "epoch": 0.4657443609022556, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 12743 + }, + { + "epoch": 0.4658045112781955, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 12744 + }, + { + "epoch": 0.4658646616541353, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 12745 + }, + { + "epoch": 0.4659248120300752, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 12746 + }, + { + "epoch": 0.46598496240601506, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 12747 + }, + { + "epoch": 0.4660451127819549, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 12748 + }, + { + "epoch": 0.46610526315789474, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 12749 + }, + { + "epoch": 0.46616541353383456, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 12750 + }, + { + "epoch": 0.46622556390977443, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 12751 + }, + { + "epoch": 0.4662857142857143, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2803, + "step": 12752 + }, + { + "epoch": 0.4663458646616541, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 12753 + }, + { + "epoch": 0.466406015037594, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 12754 + }, + { + "epoch": 0.46646616541353386, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1733, + "step": 12755 + }, + { + "epoch": 0.4665263157894737, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 12756 + }, + { + "epoch": 0.46658646616541355, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 12757 + }, + { + "epoch": 0.46664661654135337, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 12758 + }, + { + "epoch": 0.46670676691729324, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2495, + "step": 12759 + }, + { + "epoch": 0.4667669172932331, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2936, + "step": 12760 + }, + { + "epoch": 0.4668270676691729, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 12761 + }, + { + "epoch": 0.4668872180451128, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 12762 + }, + { + "epoch": 0.4669473684210526, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 12763 + }, + { + "epoch": 0.4670075187969925, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 12764 + }, + { + "epoch": 0.46706766917293235, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 12765 + }, + { + "epoch": 0.46712781954887217, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 12766 + }, + { + "epoch": 0.46718796992481204, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 12767 + }, + { + "epoch": 0.46724812030075186, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 12768 + }, + { + "epoch": 0.46730827067669173, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1837, + "step": 12769 + }, + { + "epoch": 0.4673684210526316, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1996, + "step": 12770 + }, + { + "epoch": 0.4674285714285714, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 12771 + }, + { + "epoch": 0.4674887218045113, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 12772 + }, + { + "epoch": 0.4675488721804511, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 12773 + }, + { + "epoch": 0.467609022556391, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 12774 + }, + { + "epoch": 0.46766917293233085, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 12775 + }, + { + "epoch": 0.46772932330827066, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1334, + "step": 12776 + }, + { + "epoch": 0.46778947368421053, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 12777 + }, + { + "epoch": 0.46784962406015035, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 12778 + }, + { + "epoch": 0.4679097744360902, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 12779 + }, + { + "epoch": 0.4679699248120301, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 12780 + }, + { + "epoch": 0.4680300751879699, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 12781 + }, + { + "epoch": 0.4680902255639098, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2643, + "step": 12782 + }, + { + "epoch": 0.46815037593984965, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2694, + "step": 12783 + }, + { + "epoch": 0.46821052631578947, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 12784 + }, + { + "epoch": 0.46827067669172934, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2546, + "step": 12785 + }, + { + "epoch": 0.46833082706766915, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1938, + "step": 12786 + }, + { + "epoch": 0.468390977443609, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 12787 + }, + { + "epoch": 0.4684511278195489, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2291, + "step": 12788 + }, + { + "epoch": 0.4685112781954887, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1928, + "step": 12789 + }, + { + "epoch": 0.4685714285714286, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1954, + "step": 12790 + }, + { + "epoch": 0.4686315789473684, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.282, + "step": 12791 + }, + { + "epoch": 0.46869172932330827, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 12792 + }, + { + "epoch": 0.46875187969924814, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 12793 + }, + { + "epoch": 0.46881203007518796, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 12794 + }, + { + "epoch": 0.46887218045112783, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 12795 + }, + { + "epoch": 0.46893233082706764, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 12796 + }, + { + "epoch": 0.4689924812030075, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 12797 + }, + { + "epoch": 0.4690526315789474, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 12798 + }, + { + "epoch": 0.4691127819548872, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.4358, + "step": 12799 + }, + { + "epoch": 0.4691729323308271, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 12800 + }, + { + "epoch": 0.4692330827067669, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 12801 + }, + { + "epoch": 0.46929323308270676, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 12802 + }, + { + "epoch": 0.46935338345864663, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 12803 + }, + { + "epoch": 0.46941353383458645, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 12804 + }, + { + "epoch": 0.4694736842105263, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2116, + "step": 12805 + }, + { + "epoch": 0.4695338345864662, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 12806 + }, + { + "epoch": 0.469593984962406, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 12807 + }, + { + "epoch": 0.4696541353383459, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 12808 + }, + { + "epoch": 0.4697142857142857, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2578, + "step": 12809 + }, + { + "epoch": 0.46977443609022557, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 12810 + }, + { + "epoch": 0.46983458646616544, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 12811 + }, + { + "epoch": 0.46989473684210525, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 12812 + }, + { + "epoch": 0.4699548872180451, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 12813 + }, + { + "epoch": 0.47001503759398494, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2027, + "step": 12814 + }, + { + "epoch": 0.4700751879699248, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2032, + "step": 12815 + }, + { + "epoch": 0.4701353383458647, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 12816 + }, + { + "epoch": 0.4701954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 12817 + }, + { + "epoch": 0.47025563909774437, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2209, + "step": 12818 + }, + { + "epoch": 0.4703157894736842, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 12819 + }, + { + "epoch": 0.47037593984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 12820 + }, + { + "epoch": 0.47043609022556393, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2444, + "step": 12821 + }, + { + "epoch": 0.47049624060150375, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 12822 + }, + { + "epoch": 0.4705563909774436, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 12823 + }, + { + "epoch": 0.47061654135338343, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 12824 + }, + { + "epoch": 0.4706766917293233, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1972, + "step": 12825 + }, + { + "epoch": 0.4707368421052632, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 12826 + }, + { + "epoch": 0.470796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2402, + "step": 12827 + }, + { + "epoch": 0.47085714285714286, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 12828 + }, + { + "epoch": 0.47091729323308273, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1728, + "step": 12829 + }, + { + "epoch": 0.47097744360902255, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2992, + "step": 12830 + }, + { + "epoch": 0.4710375939849624, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3338, + "step": 12831 + }, + { + "epoch": 0.47109774436090224, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 12832 + }, + { + "epoch": 0.4711578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2202, + "step": 12833 + }, + { + "epoch": 0.471218045112782, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 12834 + }, + { + "epoch": 0.4712781954887218, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.3891, + "step": 12835 + }, + { + "epoch": 0.47133834586466167, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 12836 + }, + { + "epoch": 0.4713984962406015, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2604, + "step": 12837 + }, + { + "epoch": 0.47145864661654135, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 12838 + }, + { + "epoch": 0.4715187969924812, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 12839 + }, + { + "epoch": 0.47157894736842104, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 12840 + }, + { + "epoch": 0.4716390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 12841 + }, + { + "epoch": 0.47169924812030073, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 12842 + }, + { + "epoch": 0.4717593984962406, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1746, + "step": 12843 + }, + { + "epoch": 0.47181954887218047, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 12844 + }, + { + "epoch": 0.4718796992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2238, + "step": 12845 + }, + { + "epoch": 0.47193984962406016, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 12846 + }, + { + "epoch": 0.472, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 12847 + }, + { + "epoch": 0.47206015037593985, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 12848 + }, + { + "epoch": 0.4721203007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 12849 + }, + { + "epoch": 0.47218045112781953, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2151, + "step": 12850 + }, + { + "epoch": 0.4722406015037594, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.2505, + "step": 12851 + }, + { + "epoch": 0.4723007518796992, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 12852 + }, + { + "epoch": 0.4723609022556391, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 12853 + }, + { + "epoch": 0.47242105263157896, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 12854 + }, + { + "epoch": 0.4724812030075188, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 12855 + }, + { + "epoch": 0.47254135338345865, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1612, + "step": 12856 + }, + { + "epoch": 0.4726015037593985, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 12857 + }, + { + "epoch": 0.47266165413533834, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 12858 + }, + { + "epoch": 0.4727218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 12859 + }, + { + "epoch": 0.472781954887218, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 12860 + }, + { + "epoch": 0.4728421052631579, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3041, + "step": 12861 + }, + { + "epoch": 0.47290225563909777, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 12862 + }, + { + "epoch": 0.4729624060150376, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.155, + "step": 12863 + }, + { + "epoch": 0.47302255639097746, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 12864 + }, + { + "epoch": 0.47308270676691727, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 12865 + }, + { + "epoch": 0.47314285714285714, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 12866 + }, + { + "epoch": 0.473203007518797, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 12867 + }, + { + "epoch": 0.47326315789473683, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 12868 + }, + { + "epoch": 0.4733233082706767, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 12869 + }, + { + "epoch": 0.4733834586466165, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 12870 + }, + { + "epoch": 0.4734436090225564, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 12871 + }, + { + "epoch": 0.47350375939849626, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 12872 + }, + { + "epoch": 0.4735639097744361, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 12873 + }, + { + "epoch": 0.47362406015037595, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 12874 + }, + { + "epoch": 0.47368421052631576, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 12875 + }, + { + "epoch": 0.47374436090225563, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 12876 + }, + { + "epoch": 0.4738045112781955, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 12877 + }, + { + "epoch": 0.4738646616541353, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 12878 + }, + { + "epoch": 0.4739248120300752, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 12879 + }, + { + "epoch": 0.47398496240601506, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 12880 + }, + { + "epoch": 0.4740451127819549, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2039, + "step": 12881 + }, + { + "epoch": 0.47410526315789475, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 12882 + }, + { + "epoch": 0.47416541353383457, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 12883 + }, + { + "epoch": 0.47422556390977444, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 12884 + }, + { + "epoch": 0.4742857142857143, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 12885 + }, + { + "epoch": 0.4743458646616541, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1617, + "step": 12886 + }, + { + "epoch": 0.474406015037594, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 12887 + }, + { + "epoch": 0.4744661654135338, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 12888 + }, + { + "epoch": 0.4745263157894737, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 12889 + }, + { + "epoch": 0.47458646616541356, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1512, + "step": 12890 + }, + { + "epoch": 0.47464661654135337, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 12891 + }, + { + "epoch": 0.47470676691729324, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 12892 + }, + { + "epoch": 0.47476691729323306, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 12893 + }, + { + "epoch": 0.47482706766917293, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 12894 + }, + { + "epoch": 0.4748872180451128, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 12895 + }, + { + "epoch": 0.4749473684210526, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 12896 + }, + { + "epoch": 0.4750075187969925, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 12897 + }, + { + "epoch": 0.4750676691729323, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 12898 + }, + { + "epoch": 0.4751278195488722, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1811, + "step": 12899 + }, + { + "epoch": 0.47518796992481205, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 12900 + }, + { + "epoch": 0.47524812030075186, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2843, + "step": 12901 + }, + { + "epoch": 0.47530827067669174, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 12902 + }, + { + "epoch": 0.47536842105263155, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 12903 + }, + { + "epoch": 0.4754285714285714, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1726, + "step": 12904 + }, + { + "epoch": 0.4754887218045113, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 12905 + }, + { + "epoch": 0.4755488721804511, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2066, + "step": 12906 + }, + { + "epoch": 0.475609022556391, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 12907 + }, + { + "epoch": 0.47566917293233085, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 12908 + }, + { + "epoch": 0.47572932330827067, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 12909 + }, + { + "epoch": 0.47578947368421054, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1933, + "step": 12910 + }, + { + "epoch": 0.47584962406015036, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1882, + "step": 12911 + }, + { + "epoch": 0.4759097744360902, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 12912 + }, + { + "epoch": 0.4759699248120301, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 12913 + }, + { + "epoch": 0.4760300751879699, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 12914 + }, + { + "epoch": 0.4760902255639098, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2064, + "step": 12915 + }, + { + "epoch": 0.4761503759398496, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2052, + "step": 12916 + }, + { + "epoch": 0.4762105263157895, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 12917 + }, + { + "epoch": 0.47627067669172934, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1965, + "step": 12918 + }, + { + "epoch": 0.47633082706766916, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1804, + "step": 12919 + }, + { + "epoch": 0.47639097744360903, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2197, + "step": 12920 + }, + { + "epoch": 0.47645112781954885, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 12921 + }, + { + "epoch": 0.4765112781954887, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 12922 + }, + { + "epoch": 0.4765714285714286, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 12923 + }, + { + "epoch": 0.4766315789473684, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 12924 + }, + { + "epoch": 0.4766917293233083, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 12925 + }, + { + "epoch": 0.4767518796992481, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 12926 + }, + { + "epoch": 0.47681203007518796, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 12927 + }, + { + "epoch": 0.47687218045112784, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 12928 + }, + { + "epoch": 0.47693233082706765, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1934, + "step": 12929 + }, + { + "epoch": 0.4769924812030075, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 12930 + }, + { + "epoch": 0.4770526315789474, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2301, + "step": 12931 + }, + { + "epoch": 0.4771127819548872, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1541, + "step": 12932 + }, + { + "epoch": 0.4771729323308271, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 12933 + }, + { + "epoch": 0.4772330827067669, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 12934 + }, + { + "epoch": 0.47729323308270677, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 12935 + }, + { + "epoch": 0.47735338345864664, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3164, + "step": 12936 + }, + { + "epoch": 0.47741353383458646, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 12937 + }, + { + "epoch": 0.47747368421052633, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 12938 + }, + { + "epoch": 0.47753383458646614, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2295, + "step": 12939 + }, + { + "epoch": 0.477593984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 12940 + }, + { + "epoch": 0.4776541353383459, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 12941 + }, + { + "epoch": 0.4777142857142857, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 12942 + }, + { + "epoch": 0.4777744360902256, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 12943 + }, + { + "epoch": 0.4778345864661654, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2126, + "step": 12944 + }, + { + "epoch": 0.47789473684210526, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 12945 + }, + { + "epoch": 0.47795488721804513, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 12946 + }, + { + "epoch": 0.47801503759398495, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1651, + "step": 12947 + }, + { + "epoch": 0.4780751879699248, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 12948 + }, + { + "epoch": 0.47813533834586464, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1876, + "step": 12949 + }, + { + "epoch": 0.4781954887218045, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 12950 + }, + { + "epoch": 0.4782556390977444, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 12951 + }, + { + "epoch": 0.4783157894736842, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 12952 + }, + { + "epoch": 0.47837593984962407, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2035, + "step": 12953 + }, + { + "epoch": 0.47843609022556394, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.2318, + "step": 12954 + }, + { + "epoch": 0.47849624060150375, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 12955 + }, + { + "epoch": 0.4785563909774436, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 12956 + }, + { + "epoch": 0.47861654135338344, + "grad_norm": 0.45703125, + "learning_rate": 3e-05, + "loss": 1.1861, + "step": 12957 + }, + { + "epoch": 0.4786766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 12958 + }, + { + "epoch": 0.4787368421052632, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 12959 + }, + { + "epoch": 0.478796992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 12960 + }, + { + "epoch": 0.47885714285714287, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 12961 + }, + { + "epoch": 0.4789172932330827, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 12962 + }, + { + "epoch": 0.47897744360902256, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.3244, + "step": 12963 + }, + { + "epoch": 0.47903759398496243, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 12964 + }, + { + "epoch": 0.47909774436090224, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 12965 + }, + { + "epoch": 0.4791578947368421, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1957, + "step": 12966 + }, + { + "epoch": 0.47921804511278193, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 12967 + }, + { + "epoch": 0.4792781954887218, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 12968 + }, + { + "epoch": 0.4793383458646617, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 12969 + }, + { + "epoch": 0.4793984962406015, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 12970 + }, + { + "epoch": 0.47945864661654136, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 12971 + }, + { + "epoch": 0.4795187969924812, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 12972 + }, + { + "epoch": 0.47957894736842105, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2145, + "step": 12973 + }, + { + "epoch": 0.4796390977443609, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 12974 + }, + { + "epoch": 0.47969924812030074, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2091, + "step": 12975 + }, + { + "epoch": 0.4797593984962406, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1805, + "step": 12976 + }, + { + "epoch": 0.4798195488721804, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 12977 + }, + { + "epoch": 0.4798796992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 12978 + }, + { + "epoch": 0.47993984962406017, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 12979 + }, + { + "epoch": 0.48, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 12980 + }, + { + "epoch": 0.48006015037593985, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1617, + "step": 12981 + }, + { + "epoch": 0.4801203007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 12982 + }, + { + "epoch": 0.48018045112781954, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 12983 + }, + { + "epoch": 0.4802406015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1177, + "step": 12984 + }, + { + "epoch": 0.48030075187969923, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 12985 + }, + { + "epoch": 0.4803609022556391, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 12986 + }, + { + "epoch": 0.48042105263157897, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 12987 + }, + { + "epoch": 0.4804812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2413, + "step": 12988 + }, + { + "epoch": 0.48054135338345866, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 12989 + }, + { + "epoch": 0.4806015037593985, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2259, + "step": 12990 + }, + { + "epoch": 0.48066165413533835, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 12991 + }, + { + "epoch": 0.4807218045112782, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 12992 + }, + { + "epoch": 0.48078195488721803, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 12993 + }, + { + "epoch": 0.4808421052631579, + "grad_norm": 1.390625, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 12994 + }, + { + "epoch": 0.4809022556390977, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 12995 + }, + { + "epoch": 0.4809624060150376, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 12996 + }, + { + "epoch": 0.48102255639097746, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 12997 + }, + { + "epoch": 0.4810827067669173, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2108, + "step": 12998 + }, + { + "epoch": 0.48114285714285715, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 12999 + }, + { + "epoch": 0.48120300751879697, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 13000 + }, + { + "epoch": 0.48126315789473684, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 13001 + }, + { + "epoch": 0.4813233082706767, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 13002 + }, + { + "epoch": 0.4813834586466165, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1881, + "step": 13003 + }, + { + "epoch": 0.4814436090225564, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 13004 + }, + { + "epoch": 0.48150375939849627, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 13005 + }, + { + "epoch": 0.4815639097744361, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 13006 + }, + { + "epoch": 0.48162406015037595, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.291, + "step": 13007 + }, + { + "epoch": 0.48168421052631577, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3358, + "step": 13008 + }, + { + "epoch": 0.48174436090225564, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1582, + "step": 13009 + }, + { + "epoch": 0.4818045112781955, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 13010 + }, + { + "epoch": 0.48186466165413533, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 13011 + }, + { + "epoch": 0.4819248120300752, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 13012 + }, + { + "epoch": 0.481984962406015, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 13013 + }, + { + "epoch": 0.4820451127819549, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 13014 + }, + { + "epoch": 0.48210526315789476, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 13015 + }, + { + "epoch": 0.4821654135338346, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 13016 + }, + { + "epoch": 0.48222556390977445, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 13017 + }, + { + "epoch": 0.48228571428571426, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 13018 + }, + { + "epoch": 0.48234586466165413, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2097, + "step": 13019 + }, + { + "epoch": 0.482406015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 13020 + }, + { + "epoch": 0.4824661654135338, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 13021 + }, + { + "epoch": 0.4825263157894737, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 13022 + }, + { + "epoch": 0.4825864661654135, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 13023 + }, + { + "epoch": 0.4826466165413534, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 13024 + }, + { + "epoch": 0.48270676691729325, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 13025 + }, + { + "epoch": 0.48276691729323307, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 13026 + }, + { + "epoch": 0.48282706766917294, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 13027 + }, + { + "epoch": 0.4828872180451128, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 13028 + }, + { + "epoch": 0.4829473684210526, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.3916, + "step": 13029 + }, + { + "epoch": 0.4830075187969925, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 13030 + }, + { + "epoch": 0.4830676691729323, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 13031 + }, + { + "epoch": 0.4831278195488722, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 13032 + }, + { + "epoch": 0.48318796992481206, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 13033 + }, + { + "epoch": 0.48324812030075187, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 13034 + }, + { + "epoch": 0.48330827067669174, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 13035 + }, + { + "epoch": 0.48336842105263156, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1894, + "step": 13036 + }, + { + "epoch": 0.48342857142857143, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 13037 + }, + { + "epoch": 0.4834887218045113, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2861, + "step": 13038 + }, + { + "epoch": 0.4835488721804511, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 13039 + }, + { + "epoch": 0.483609022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 13040 + }, + { + "epoch": 0.4836691729323308, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 13041 + }, + { + "epoch": 0.4837293233082707, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 13042 + }, + { + "epoch": 0.48378947368421055, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 13043 + }, + { + "epoch": 0.48384962406015036, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 13044 + }, + { + "epoch": 0.48390977443609023, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 13045 + }, + { + "epoch": 0.48396992481203005, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1832, + "step": 13046 + }, + { + "epoch": 0.4840300751879699, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 13047 + }, + { + "epoch": 0.4840902255639098, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 13048 + }, + { + "epoch": 0.4841503759398496, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 13049 + }, + { + "epoch": 0.4842105263157895, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2257, + "step": 13050 + }, + { + "epoch": 0.4842706766917293, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 13051 + }, + { + "epoch": 0.48433082706766917, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 13052 + }, + { + "epoch": 0.48439097744360904, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 13053 + }, + { + "epoch": 0.48445112781954885, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2444, + "step": 13054 + }, + { + "epoch": 0.4845112781954887, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 13055 + }, + { + "epoch": 0.4845714285714286, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 13056 + }, + { + "epoch": 0.4846315789473684, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2044, + "step": 13057 + }, + { + "epoch": 0.4846917293233083, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 13058 + }, + { + "epoch": 0.4847518796992481, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 13059 + }, + { + "epoch": 0.48481203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1905, + "step": 13060 + }, + { + "epoch": 0.48487218045112784, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2536, + "step": 13061 + }, + { + "epoch": 0.48493233082706766, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 13062 + }, + { + "epoch": 0.48499248120300753, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.3882, + "step": 13063 + }, + { + "epoch": 0.48505263157894735, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2947, + "step": 13064 + }, + { + "epoch": 0.4851127819548872, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2291, + "step": 13065 + }, + { + "epoch": 0.4851729323308271, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.3557, + "step": 13066 + }, + { + "epoch": 0.4852330827067669, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 13067 + }, + { + "epoch": 0.4852932330827068, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 13068 + }, + { + "epoch": 0.4853533834586466, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2122, + "step": 13069 + }, + { + "epoch": 0.48541353383458646, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 13070 + }, + { + "epoch": 0.48547368421052634, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 13071 + }, + { + "epoch": 0.48553383458646615, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1708, + "step": 13072 + }, + { + "epoch": 0.485593984962406, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 13073 + }, + { + "epoch": 0.48565413533834584, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 13074 + }, + { + "epoch": 0.4857142857142857, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2242, + "step": 13075 + }, + { + "epoch": 0.4857744360902256, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 13076 + }, + { + "epoch": 0.4858345864661654, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 13077 + }, + { + "epoch": 0.48589473684210527, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 13078 + }, + { + "epoch": 0.48595488721804514, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2846, + "step": 13079 + }, + { + "epoch": 0.48601503759398496, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.2841, + "step": 13080 + }, + { + "epoch": 0.4860751879699248, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 13081 + }, + { + "epoch": 0.48613533834586464, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 13082 + }, + { + "epoch": 0.4861954887218045, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2688, + "step": 13083 + }, + { + "epoch": 0.4862556390977444, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 13084 + }, + { + "epoch": 0.4863157894736842, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 13085 + }, + { + "epoch": 0.4863759398496241, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 13086 + }, + { + "epoch": 0.4864360902255639, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1758, + "step": 13087 + }, + { + "epoch": 0.48649624060150376, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 13088 + }, + { + "epoch": 0.48655639097744363, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 13089 + }, + { + "epoch": 0.48661654135338345, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 13090 + }, + { + "epoch": 0.4866766917293233, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 13091 + }, + { + "epoch": 0.48673684210526313, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 13092 + }, + { + "epoch": 0.486796992481203, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 13093 + }, + { + "epoch": 0.4868571428571429, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 13094 + }, + { + "epoch": 0.4869172932330827, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.2425, + "step": 13095 + }, + { + "epoch": 0.48697744360902256, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 13096 + }, + { + "epoch": 0.4870375939849624, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 13097 + }, + { + "epoch": 0.48709774436090225, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 13098 + }, + { + "epoch": 0.4871578947368421, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 13099 + }, + { + "epoch": 0.48721804511278194, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1494, + "step": 13100 + }, + { + "epoch": 0.4872781954887218, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 13101 + }, + { + "epoch": 0.4873383458646617, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 13102 + }, + { + "epoch": 0.4873984962406015, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1727, + "step": 13103 + }, + { + "epoch": 0.48745864661654137, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 13104 + }, + { + "epoch": 0.4875187969924812, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 13105 + }, + { + "epoch": 0.48757894736842106, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2463, + "step": 13106 + }, + { + "epoch": 0.4876390977443609, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 13107 + }, + { + "epoch": 0.48769924812030074, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 13108 + }, + { + "epoch": 0.4877593984962406, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 13109 + }, + { + "epoch": 0.48781954887218043, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 13110 + }, + { + "epoch": 0.4878796992481203, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 13111 + }, + { + "epoch": 0.4879398496240602, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 13112 + }, + { + "epoch": 0.488, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2783, + "step": 13113 + }, + { + "epoch": 0.48806015037593986, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 13114 + }, + { + "epoch": 0.4881203007518797, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 13115 + }, + { + "epoch": 0.48818045112781955, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2188, + "step": 13116 + }, + { + "epoch": 0.4882406015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2221, + "step": 13117 + }, + { + "epoch": 0.48830075187969924, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 13118 + }, + { + "epoch": 0.4883609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 13119 + }, + { + "epoch": 0.4884210526315789, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.216, + "step": 13120 + }, + { + "epoch": 0.4884812030075188, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2927, + "step": 13121 + }, + { + "epoch": 0.48854135338345867, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.466, + "step": 13122 + }, + { + "epoch": 0.4886015037593985, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 13123 + }, + { + "epoch": 0.48866165413533835, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 13124 + }, + { + "epoch": 0.48872180451127817, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 13125 + }, + { + "epoch": 0.48878195488721804, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2122, + "step": 13126 + }, + { + "epoch": 0.4888421052631579, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2495, + "step": 13127 + }, + { + "epoch": 0.4889022556390977, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.228, + "step": 13128 + }, + { + "epoch": 0.4889624060150376, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2142, + "step": 13129 + }, + { + "epoch": 0.48902255639097747, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 13130 + }, + { + "epoch": 0.4890827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 13131 + }, + { + "epoch": 0.48914285714285716, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 13132 + }, + { + "epoch": 0.489203007518797, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 13133 + }, + { + "epoch": 0.48926315789473684, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 13134 + }, + { + "epoch": 0.4893233082706767, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1833, + "step": 13135 + }, + { + "epoch": 0.48938345864661653, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1988, + "step": 13136 + }, + { + "epoch": 0.4894436090225564, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2564, + "step": 13137 + }, + { + "epoch": 0.4895037593984962, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 13138 + }, + { + "epoch": 0.4895639097744361, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2264, + "step": 13139 + }, + { + "epoch": 0.48962406015037596, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 13140 + }, + { + "epoch": 0.4896842105263158, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2233, + "step": 13141 + }, + { + "epoch": 0.48974436090225565, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 13142 + }, + { + "epoch": 0.48980451127819546, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 13143 + }, + { + "epoch": 0.48986466165413534, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 13144 + }, + { + "epoch": 0.4899248120300752, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 13145 + }, + { + "epoch": 0.489984962406015, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 13146 + }, + { + "epoch": 0.4900451127819549, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 13147 + }, + { + "epoch": 0.4901052631578947, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2129, + "step": 13148 + }, + { + "epoch": 0.4901654135338346, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 13149 + }, + { + "epoch": 0.49022556390977445, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 13150 + }, + { + "epoch": 0.49028571428571427, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 13151 + }, + { + "epoch": 0.49034586466165414, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1418, + "step": 13152 + }, + { + "epoch": 0.490406015037594, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 13153 + }, + { + "epoch": 0.49046616541353383, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 13154 + }, + { + "epoch": 0.4905263157894737, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 13155 + }, + { + "epoch": 0.4905864661654135, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 13156 + }, + { + "epoch": 0.4906466165413534, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 13157 + }, + { + "epoch": 0.49070676691729326, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2356, + "step": 13158 + }, + { + "epoch": 0.4907669172932331, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2102, + "step": 13159 + }, + { + "epoch": 0.49082706766917295, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 13160 + }, + { + "epoch": 0.49088721804511276, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 13161 + }, + { + "epoch": 0.49094736842105263, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 13162 + }, + { + "epoch": 0.4910075187969925, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 13163 + }, + { + "epoch": 0.4910676691729323, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 13164 + }, + { + "epoch": 0.4911278195488722, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 13165 + }, + { + "epoch": 0.491187969924812, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 13166 + }, + { + "epoch": 0.4912481203007519, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 13167 + }, + { + "epoch": 0.49130827067669175, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 13168 + }, + { + "epoch": 0.49136842105263157, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 13169 + }, + { + "epoch": 0.49142857142857144, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 13170 + }, + { + "epoch": 0.49148872180451125, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 13171 + }, + { + "epoch": 0.4915488721804511, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 13172 + }, + { + "epoch": 0.491609022556391, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 13173 + }, + { + "epoch": 0.4916691729323308, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 13174 + }, + { + "epoch": 0.4917293233082707, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 13175 + }, + { + "epoch": 0.4917894736842105, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 13176 + }, + { + "epoch": 0.49184962406015037, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 13177 + }, + { + "epoch": 0.49190977443609024, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 13178 + }, + { + "epoch": 0.49196992481203006, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.3233, + "step": 13179 + }, + { + "epoch": 0.49203007518796993, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 13180 + }, + { + "epoch": 0.4920902255639098, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 13181 + }, + { + "epoch": 0.4921503759398496, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2161, + "step": 13182 + }, + { + "epoch": 0.4922105263157895, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 13183 + }, + { + "epoch": 0.4922706766917293, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 13184 + }, + { + "epoch": 0.4923308270676692, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1623, + "step": 13185 + }, + { + "epoch": 0.49239097744360905, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 13186 + }, + { + "epoch": 0.49245112781954886, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 13187 + }, + { + "epoch": 0.49251127819548873, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2455, + "step": 13188 + }, + { + "epoch": 0.49257142857142855, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 13189 + }, + { + "epoch": 0.4926315789473684, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 13190 + }, + { + "epoch": 0.4926917293233083, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 13191 + }, + { + "epoch": 0.4927518796992481, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 13192 + }, + { + "epoch": 0.492812030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1633, + "step": 13193 + }, + { + "epoch": 0.4928721804511278, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 13194 + }, + { + "epoch": 0.49293233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 13195 + }, + { + "epoch": 0.49299248120300754, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 13196 + }, + { + "epoch": 0.49305263157894735, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 13197 + }, + { + "epoch": 0.4931127819548872, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 13198 + }, + { + "epoch": 0.49317293233082704, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 13199 + }, + { + "epoch": 0.4932330827067669, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 13200 + }, + { + "epoch": 0.4932932330827068, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 13201 + }, + { + "epoch": 0.4933533834586466, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1646, + "step": 13202 + }, + { + "epoch": 0.49341353383458647, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 13203 + }, + { + "epoch": 0.49347368421052634, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 13204 + }, + { + "epoch": 0.49353383458646616, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1804, + "step": 13205 + }, + { + "epoch": 0.49359398496240603, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1314, + "step": 13206 + }, + { + "epoch": 0.49365413533834585, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 13207 + }, + { + "epoch": 0.4937142857142857, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1682, + "step": 13208 + }, + { + "epoch": 0.4937744360902256, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 13209 + }, + { + "epoch": 0.4938345864661654, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 13210 + }, + { + "epoch": 0.4938947368421053, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.2286, + "step": 13211 + }, + { + "epoch": 0.4939548872180451, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1776, + "step": 13212 + }, + { + "epoch": 0.49401503759398496, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 13213 + }, + { + "epoch": 0.49407518796992483, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 13214 + }, + { + "epoch": 0.49413533834586465, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2529, + "step": 13215 + }, + { + "epoch": 0.4941954887218045, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 13216 + }, + { + "epoch": 0.49425563909774434, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 13217 + }, + { + "epoch": 0.4943157894736842, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2169, + "step": 13218 + }, + { + "epoch": 0.4943759398496241, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2232, + "step": 13219 + }, + { + "epoch": 0.4944360902255639, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 13220 + }, + { + "epoch": 0.49449624060150377, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2302, + "step": 13221 + }, + { + "epoch": 0.4945563909774436, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2533, + "step": 13222 + }, + { + "epoch": 0.49461654135338345, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 13223 + }, + { + "epoch": 0.4946766917293233, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 13224 + }, + { + "epoch": 0.49473684210526314, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 13225 + }, + { + "epoch": 0.494796992481203, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 13226 + }, + { + "epoch": 0.4948571428571429, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 13227 + }, + { + "epoch": 0.4949172932330827, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2394, + "step": 13228 + }, + { + "epoch": 0.49497744360902257, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 13229 + }, + { + "epoch": 0.4950375939849624, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 13230 + }, + { + "epoch": 0.49509774436090226, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2471, + "step": 13231 + }, + { + "epoch": 0.49515789473684213, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 13232 + }, + { + "epoch": 0.49521804511278195, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.3048, + "step": 13233 + }, + { + "epoch": 0.4952781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2331, + "step": 13234 + }, + { + "epoch": 0.49533834586466163, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 13235 + }, + { + "epoch": 0.4953984962406015, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 13236 + }, + { + "epoch": 0.4954586466165414, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2383, + "step": 13237 + }, + { + "epoch": 0.4955187969924812, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 13238 + }, + { + "epoch": 0.49557894736842106, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 13239 + }, + { + "epoch": 0.4956390977443609, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 13240 + }, + { + "epoch": 0.49569924812030075, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 13241 + }, + { + "epoch": 0.4957593984962406, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 13242 + }, + { + "epoch": 0.49581954887218044, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 13243 + }, + { + "epoch": 0.4958796992481203, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 13244 + }, + { + "epoch": 0.4959398496240601, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1837, + "step": 13245 + }, + { + "epoch": 0.496, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 13246 + }, + { + "epoch": 0.49606015037593987, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2722, + "step": 13247 + }, + { + "epoch": 0.4961203007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 13248 + }, + { + "epoch": 0.49618045112781956, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 13249 + }, + { + "epoch": 0.49624060150375937, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 13250 + }, + { + "epoch": 0.49630075187969924, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 13251 + }, + { + "epoch": 0.4963609022556391, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1786, + "step": 13252 + }, + { + "epoch": 0.49642105263157893, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 13253 + }, + { + "epoch": 0.4964812030075188, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.2664, + "step": 13254 + }, + { + "epoch": 0.4965413533834587, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 13255 + }, + { + "epoch": 0.4966015037593985, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 13256 + }, + { + "epoch": 0.49666165413533836, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.2159, + "step": 13257 + }, + { + "epoch": 0.4967218045112782, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1711, + "step": 13258 + }, + { + "epoch": 0.49678195488721805, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 13259 + }, + { + "epoch": 0.4968421052631579, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 13260 + }, + { + "epoch": 0.49690225563909773, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 13261 + }, + { + "epoch": 0.4969624060150376, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 13262 + }, + { + "epoch": 0.4970225563909774, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 13263 + }, + { + "epoch": 0.4970827067669173, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 13264 + }, + { + "epoch": 0.49714285714285716, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 13265 + }, + { + "epoch": 0.497203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2193, + "step": 13266 + }, + { + "epoch": 0.49726315789473685, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 13267 + }, + { + "epoch": 0.49732330827067667, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 13268 + }, + { + "epoch": 0.49738345864661654, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 13269 + }, + { + "epoch": 0.4974436090225564, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2739, + "step": 13270 + }, + { + "epoch": 0.4975037593984962, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 13271 + }, + { + "epoch": 0.4975639097744361, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1793, + "step": 13272 + }, + { + "epoch": 0.4976240601503759, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 13273 + }, + { + "epoch": 0.4976842105263158, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.199, + "step": 13274 + }, + { + "epoch": 0.49774436090225566, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 13275 + }, + { + "epoch": 0.49780451127819547, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 13276 + }, + { + "epoch": 0.49786466165413534, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1896, + "step": 13277 + }, + { + "epoch": 0.4979248120300752, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 13278 + }, + { + "epoch": 0.49798496240601503, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 13279 + }, + { + "epoch": 0.4980451127819549, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 13280 + }, + { + "epoch": 0.4981052631578947, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 13281 + }, + { + "epoch": 0.4981654135338346, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 13282 + }, + { + "epoch": 0.49822556390977446, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 13283 + }, + { + "epoch": 0.4982857142857143, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 13284 + }, + { + "epoch": 0.49834586466165415, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 13285 + }, + { + "epoch": 0.49840601503759396, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 13286 + }, + { + "epoch": 0.49846616541353383, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 13287 + }, + { + "epoch": 0.4985263157894737, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.3249, + "step": 13288 + }, + { + "epoch": 0.4985864661654135, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 13289 + }, + { + "epoch": 0.4986466165413534, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 13290 + }, + { + "epoch": 0.4987067669172932, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 13291 + }, + { + "epoch": 0.4987669172932331, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.3112, + "step": 13292 + }, + { + "epoch": 0.49882706766917295, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2769, + "step": 13293 + }, + { + "epoch": 0.49888721804511277, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 13294 + }, + { + "epoch": 0.49894736842105264, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 13295 + }, + { + "epoch": 0.49900751879699246, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 13296 + }, + { + "epoch": 0.4990676691729323, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 13297 + }, + { + "epoch": 0.4991278195488722, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 13298 + }, + { + "epoch": 0.499187969924812, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 13299 + }, + { + "epoch": 0.4992481203007519, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2573, + "step": 13300 + }, + { + "epoch": 0.49930827067669176, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 13301 + }, + { + "epoch": 0.4993684210526316, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2497, + "step": 13302 + }, + { + "epoch": 0.49942857142857144, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1773, + "step": 13303 + }, + { + "epoch": 0.49948872180451126, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1914, + "step": 13304 + }, + { + "epoch": 0.49954887218045113, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 13305 + }, + { + "epoch": 0.499609022556391, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 13306 + }, + { + "epoch": 0.4996691729323308, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 13307 + }, + { + "epoch": 0.4997293233082707, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 13308 + }, + { + "epoch": 0.4997894736842105, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1971, + "step": 13309 + }, + { + "epoch": 0.4998496240601504, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2028, + "step": 13310 + }, + { + "epoch": 0.49990977443609025, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 13311 + }, + { + "epoch": 0.49996992481203006, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 13312 + }, + { + "epoch": 0.5000300751879699, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.2574, + "step": 13313 + }, + { + "epoch": 0.5000902255639098, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 13314 + }, + { + "epoch": 0.5001503759398497, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2058, + "step": 13315 + }, + { + "epoch": 0.5002105263157894, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 13316 + }, + { + "epoch": 0.5002706766917293, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 13317 + }, + { + "epoch": 0.5003308270676692, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 13318 + }, + { + "epoch": 0.500390977443609, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 13319 + }, + { + "epoch": 0.5004511278195489, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 13320 + }, + { + "epoch": 0.5005112781954887, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 13321 + }, + { + "epoch": 0.5005714285714286, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1651, + "step": 13322 + }, + { + "epoch": 0.5006315789473684, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2227, + "step": 13323 + }, + { + "epoch": 0.5006917293233083, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 13324 + }, + { + "epoch": 0.5007518796992482, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 13325 + }, + { + "epoch": 0.5008120300751879, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1552, + "step": 13326 + }, + { + "epoch": 0.5008721804511278, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 13327 + }, + { + "epoch": 0.5009323308270677, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2604, + "step": 13328 + }, + { + "epoch": 0.5009924812030075, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 13329 + }, + { + "epoch": 0.5010526315789474, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 13330 + }, + { + "epoch": 0.5011127819548872, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 13331 + }, + { + "epoch": 0.501172932330827, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1953, + "step": 13332 + }, + { + "epoch": 0.5012330827067669, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 13333 + }, + { + "epoch": 0.5012932330827068, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1444, + "step": 13334 + }, + { + "epoch": 0.5013533834586467, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 13335 + }, + { + "epoch": 0.5014135338345864, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 13336 + }, + { + "epoch": 0.5014736842105263, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 13337 + }, + { + "epoch": 0.5015338345864662, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 13338 + }, + { + "epoch": 0.501593984962406, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 13339 + }, + { + "epoch": 0.5016541353383459, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.2218, + "step": 13340 + }, + { + "epoch": 0.5017142857142857, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 13341 + }, + { + "epoch": 0.5017744360902255, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 13342 + }, + { + "epoch": 0.5018345864661654, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 13343 + }, + { + "epoch": 0.5018947368421053, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 13344 + }, + { + "epoch": 0.5019548872180452, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 13345 + }, + { + "epoch": 0.5020150375939849, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 13346 + }, + { + "epoch": 0.5020751879699248, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 13347 + }, + { + "epoch": 0.5021353383458647, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 13348 + }, + { + "epoch": 0.5021954887218045, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.4, + "step": 13349 + }, + { + "epoch": 0.5022556390977444, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1641, + "step": 13350 + }, + { + "epoch": 0.5023157894736842, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 13351 + }, + { + "epoch": 0.502375939849624, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 13352 + }, + { + "epoch": 0.5024360902255639, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2256, + "step": 13353 + }, + { + "epoch": 0.5024962406015038, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 13354 + }, + { + "epoch": 0.5025563909774436, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 13355 + }, + { + "epoch": 0.5026165413533834, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2485, + "step": 13356 + }, + { + "epoch": 0.5026766917293233, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1865, + "step": 13357 + }, + { + "epoch": 0.5027368421052631, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1773, + "step": 13358 + }, + { + "epoch": 0.502796992481203, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.2758, + "step": 13359 + }, + { + "epoch": 0.5028571428571429, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2094, + "step": 13360 + }, + { + "epoch": 0.5029172932330827, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 13361 + }, + { + "epoch": 0.5029774436090225, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 13362 + }, + { + "epoch": 0.5030375939849624, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2185, + "step": 13363 + }, + { + "epoch": 0.5030977443609023, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 13364 + }, + { + "epoch": 0.5031578947368421, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.231, + "step": 13365 + }, + { + "epoch": 0.503218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 13366 + }, + { + "epoch": 0.5032781954887218, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1494, + "step": 13367 + }, + { + "epoch": 0.5033383458646616, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.4226, + "step": 13368 + }, + { + "epoch": 0.5033984962406015, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 13369 + }, + { + "epoch": 0.5034586466165414, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 13370 + }, + { + "epoch": 0.5035187969924813, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 13371 + }, + { + "epoch": 0.503578947368421, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2052, + "step": 13372 + }, + { + "epoch": 0.5036390977443609, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1979, + "step": 13373 + }, + { + "epoch": 0.5036992481203008, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 13374 + }, + { + "epoch": 0.5037593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 13375 + }, + { + "epoch": 0.5038195488721805, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 13376 + }, + { + "epoch": 0.5038796992481203, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 13377 + }, + { + "epoch": 0.5039398496240601, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 13378 + }, + { + "epoch": 0.504, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 13379 + }, + { + "epoch": 0.5040601503759399, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 13380 + }, + { + "epoch": 0.5041203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 13381 + }, + { + "epoch": 0.5041804511278195, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 13382 + }, + { + "epoch": 0.5042406015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 13383 + }, + { + "epoch": 0.5043007518796992, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 13384 + }, + { + "epoch": 0.5043609022556391, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 13385 + }, + { + "epoch": 0.504421052631579, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 13386 + }, + { + "epoch": 0.5044812030075188, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 13387 + }, + { + "epoch": 0.5045413533834586, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 13388 + }, + { + "epoch": 0.5046015037593985, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 13389 + }, + { + "epoch": 0.5046616541353384, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.2199, + "step": 13390 + }, + { + "epoch": 0.5047218045112782, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 13391 + }, + { + "epoch": 0.504781954887218, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 13392 + }, + { + "epoch": 0.5048421052631579, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1264, + "step": 13393 + }, + { + "epoch": 0.5049022556390977, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 13394 + }, + { + "epoch": 0.5049624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 13395 + }, + { + "epoch": 0.5050225563909775, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 13396 + }, + { + "epoch": 0.5050827067669172, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 13397 + }, + { + "epoch": 0.5051428571428571, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2233, + "step": 13398 + }, + { + "epoch": 0.505203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 13399 + }, + { + "epoch": 0.5052631578947369, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 13400 + }, + { + "epoch": 0.5053233082706767, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2963, + "step": 13401 + }, + { + "epoch": 0.5053834586466165, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 13402 + }, + { + "epoch": 0.5054436090225564, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.2342, + "step": 13403 + }, + { + "epoch": 0.5055037593984962, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 13404 + }, + { + "epoch": 0.5055639097744361, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.257, + "step": 13405 + }, + { + "epoch": 0.505624060150376, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 13406 + }, + { + "epoch": 0.5056842105263157, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 13407 + }, + { + "epoch": 0.5057443609022556, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.3063, + "step": 13408 + }, + { + "epoch": 0.5058045112781955, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 13409 + }, + { + "epoch": 0.5058646616541354, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 13410 + }, + { + "epoch": 0.5059248120300752, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2133, + "step": 13411 + }, + { + "epoch": 0.5059849624060151, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 13412 + }, + { + "epoch": 0.5060451127819549, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2244, + "step": 13413 + }, + { + "epoch": 0.5061052631578947, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 13414 + }, + { + "epoch": 0.5061654135338346, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 13415 + }, + { + "epoch": 0.5062255639097745, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 13416 + }, + { + "epoch": 0.5062857142857143, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 13417 + }, + { + "epoch": 0.5063458646616541, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 13418 + }, + { + "epoch": 0.506406015037594, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 13419 + }, + { + "epoch": 0.5064661654135338, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 13420 + }, + { + "epoch": 0.5065263157894737, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2046, + "step": 13421 + }, + { + "epoch": 0.5065864661654136, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2511, + "step": 13422 + }, + { + "epoch": 0.5066466165413533, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 13423 + }, + { + "epoch": 0.5067067669172932, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.188, + "step": 13424 + }, + { + "epoch": 0.5067669172932331, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 13425 + }, + { + "epoch": 0.506827067669173, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2448, + "step": 13426 + }, + { + "epoch": 0.5068872180451128, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 13427 + }, + { + "epoch": 0.5069473684210526, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 13428 + }, + { + "epoch": 0.5070075187969925, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 13429 + }, + { + "epoch": 0.5070676691729323, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1714, + "step": 13430 + }, + { + "epoch": 0.5071278195488722, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1985, + "step": 13431 + }, + { + "epoch": 0.5071879699248121, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 13432 + }, + { + "epoch": 0.5072481203007518, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 13433 + }, + { + "epoch": 0.5073082706766917, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2087, + "step": 13434 + }, + { + "epoch": 0.5073684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 13435 + }, + { + "epoch": 0.5074285714285715, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 13436 + }, + { + "epoch": 0.5074887218045113, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 13437 + }, + { + "epoch": 0.5075488721804511, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.256, + "step": 13438 + }, + { + "epoch": 0.507609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 13439 + }, + { + "epoch": 0.5076691729323308, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 13440 + }, + { + "epoch": 0.5077293233082707, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 13441 + }, + { + "epoch": 0.5077894736842106, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1713, + "step": 13442 + }, + { + "epoch": 0.5078496240601503, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 13443 + }, + { + "epoch": 0.5079097744360902, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2342, + "step": 13444 + }, + { + "epoch": 0.5079699248120301, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 13445 + }, + { + "epoch": 0.5080300751879699, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 13446 + }, + { + "epoch": 0.5080902255639098, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 13447 + }, + { + "epoch": 0.5081503759398496, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2147, + "step": 13448 + }, + { + "epoch": 0.5082105263157894, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 13449 + }, + { + "epoch": 0.5082706766917293, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 13450 + }, + { + "epoch": 0.5083308270676692, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2154, + "step": 13451 + }, + { + "epoch": 0.5083909774436091, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 13452 + }, + { + "epoch": 0.5084511278195488, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3052, + "step": 13453 + }, + { + "epoch": 0.5085112781954887, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 13454 + }, + { + "epoch": 0.5085714285714286, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2472, + "step": 13455 + }, + { + "epoch": 0.5086315789473684, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 13456 + }, + { + "epoch": 0.5086917293233083, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 13457 + }, + { + "epoch": 0.5087518796992481, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 13458 + }, + { + "epoch": 0.5088120300751879, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 13459 + }, + { + "epoch": 0.5088721804511278, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1814, + "step": 13460 + }, + { + "epoch": 0.5089323308270677, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 13461 + }, + { + "epoch": 0.5089924812030076, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 13462 + }, + { + "epoch": 0.5090526315789474, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1482, + "step": 13463 + }, + { + "epoch": 0.5091127819548872, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2553, + "step": 13464 + }, + { + "epoch": 0.509172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 13465 + }, + { + "epoch": 0.5092330827067669, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 13466 + }, + { + "epoch": 0.5092932330827068, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 13467 + }, + { + "epoch": 0.5093533834586467, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 13468 + }, + { + "epoch": 0.5094135338345864, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 13469 + }, + { + "epoch": 0.5094736842105263, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1868, + "step": 13470 + }, + { + "epoch": 0.5095338345864662, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 13471 + }, + { + "epoch": 0.509593984962406, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 13472 + }, + { + "epoch": 0.5096541353383459, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 13473 + }, + { + "epoch": 0.5097142857142857, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 13474 + }, + { + "epoch": 0.5097744360902255, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 13475 + }, + { + "epoch": 0.5098345864661654, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1731, + "step": 13476 + }, + { + "epoch": 0.5098947368421053, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2412, + "step": 13477 + }, + { + "epoch": 0.5099548872180452, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2879, + "step": 13478 + }, + { + "epoch": 0.5100150375939849, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.195, + "step": 13479 + }, + { + "epoch": 0.5100751879699248, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2473, + "step": 13480 + }, + { + "epoch": 0.5101353383458647, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 13481 + }, + { + "epoch": 0.5101954887218045, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1585, + "step": 13482 + }, + { + "epoch": 0.5102556390977444, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 13483 + }, + { + "epoch": 0.5103157894736842, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 13484 + }, + { + "epoch": 0.510375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 13485 + }, + { + "epoch": 0.5104360902255639, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 13486 + }, + { + "epoch": 0.5104962406015038, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.23, + "step": 13487 + }, + { + "epoch": 0.5105563909774437, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 13488 + }, + { + "epoch": 0.5106165413533834, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.3619, + "step": 13489 + }, + { + "epoch": 0.5106766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 13490 + }, + { + "epoch": 0.5107368421052632, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 13491 + }, + { + "epoch": 0.510796992481203, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 13492 + }, + { + "epoch": 0.5108571428571429, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 13493 + }, + { + "epoch": 0.5109172932330827, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 13494 + }, + { + "epoch": 0.5109774436090225, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 13495 + }, + { + "epoch": 0.5110375939849624, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 13496 + }, + { + "epoch": 0.5110977443609023, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 13497 + }, + { + "epoch": 0.5111578947368421, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1936, + "step": 13498 + }, + { + "epoch": 0.5112180451127819, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 13499 + }, + { + "epoch": 0.5112781954887218, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 13500 + }, + { + "epoch": 0.5113383458646616, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2104, + "step": 13501 + }, + { + "epoch": 0.5113984962406015, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1966, + "step": 13502 + }, + { + "epoch": 0.5114586466165414, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 13503 + }, + { + "epoch": 0.5115187969924812, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 13504 + }, + { + "epoch": 0.511578947368421, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 13505 + }, + { + "epoch": 0.5116390977443609, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.3549, + "step": 13506 + }, + { + "epoch": 0.5116992481203008, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 13507 + }, + { + "epoch": 0.5117593984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.253, + "step": 13508 + }, + { + "epoch": 0.5118195488721804, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 13509 + }, + { + "epoch": 0.5118796992481203, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 13510 + }, + { + "epoch": 0.5119398496240601, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 13511 + }, + { + "epoch": 0.512, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 13512 + }, + { + "epoch": 0.5120601503759399, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 13513 + }, + { + "epoch": 0.5121203007518798, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2114, + "step": 13514 + }, + { + "epoch": 0.5121804511278195, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 13515 + }, + { + "epoch": 0.5122406015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2879, + "step": 13516 + }, + { + "epoch": 0.5123007518796993, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1464, + "step": 13517 + }, + { + "epoch": 0.5123609022556391, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 13518 + }, + { + "epoch": 0.512421052631579, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 13519 + }, + { + "epoch": 0.5124812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2092, + "step": 13520 + }, + { + "epoch": 0.5125413533834586, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 13521 + }, + { + "epoch": 0.5126015037593985, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 13522 + }, + { + "epoch": 0.5126616541353384, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 13523 + }, + { + "epoch": 0.5127218045112782, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.4135, + "step": 13524 + }, + { + "epoch": 0.512781954887218, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1913, + "step": 13525 + }, + { + "epoch": 0.5128421052631579, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 13526 + }, + { + "epoch": 0.5129022556390977, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 13527 + }, + { + "epoch": 0.5129624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1543, + "step": 13528 + }, + { + "epoch": 0.5130225563909775, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 13529 + }, + { + "epoch": 0.5130827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 13530 + }, + { + "epoch": 0.5131428571428571, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 13531 + }, + { + "epoch": 0.513203007518797, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2147, + "step": 13532 + }, + { + "epoch": 0.5132631578947369, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 13533 + }, + { + "epoch": 0.5133233082706767, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 13534 + }, + { + "epoch": 0.5133834586466165, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 13535 + }, + { + "epoch": 0.5134436090225564, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 13536 + }, + { + "epoch": 0.5135037593984962, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2579, + "step": 13537 + }, + { + "epoch": 0.5135639097744361, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2162, + "step": 13538 + }, + { + "epoch": 0.513624060150376, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 13539 + }, + { + "epoch": 0.5136842105263157, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 13540 + }, + { + "epoch": 0.5137443609022556, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 13541 + }, + { + "epoch": 0.5138045112781955, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 13542 + }, + { + "epoch": 0.5138646616541354, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 13543 + }, + { + "epoch": 0.5139248120300752, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2118, + "step": 13544 + }, + { + "epoch": 0.513984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 13545 + }, + { + "epoch": 0.5140451127819549, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 13546 + }, + { + "epoch": 0.5141052631578947, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 13547 + }, + { + "epoch": 0.5141654135338346, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 13548 + }, + { + "epoch": 0.5142255639097745, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 13549 + }, + { + "epoch": 0.5142857142857142, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 13550 + }, + { + "epoch": 0.5143458646616541, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 13551 + }, + { + "epoch": 0.514406015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 13552 + }, + { + "epoch": 0.5144661654135338, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2044, + "step": 13553 + }, + { + "epoch": 0.5145263157894737, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2019, + "step": 13554 + }, + { + "epoch": 0.5145864661654135, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 13555 + }, + { + "epoch": 0.5146466165413534, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 13556 + }, + { + "epoch": 0.5147067669172932, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 13557 + }, + { + "epoch": 0.5147669172932331, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 13558 + }, + { + "epoch": 0.514827067669173, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 13559 + }, + { + "epoch": 0.5148872180451127, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 13560 + }, + { + "epoch": 0.5149473684210526, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 13561 + }, + { + "epoch": 0.5150075187969925, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 13562 + }, + { + "epoch": 0.5150676691729323, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 13563 + }, + { + "epoch": 0.5151278195488722, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 13564 + }, + { + "epoch": 0.5151879699248121, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1177, + "step": 13565 + }, + { + "epoch": 0.5152481203007518, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 13566 + }, + { + "epoch": 0.5153082706766917, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 13567 + }, + { + "epoch": 0.5153684210526316, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 13568 + }, + { + "epoch": 0.5154285714285715, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1882, + "step": 13569 + }, + { + "epoch": 0.5154887218045113, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 13570 + }, + { + "epoch": 0.5155488721804511, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 13571 + }, + { + "epoch": 0.515609022556391, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2055, + "step": 13572 + }, + { + "epoch": 0.5156691729323308, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2309, + "step": 13573 + }, + { + "epoch": 0.5157293233082707, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 13574 + }, + { + "epoch": 0.5157894736842106, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 13575 + }, + { + "epoch": 0.5158496240601503, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2769, + "step": 13576 + }, + { + "epoch": 0.5159097744360902, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 13577 + }, + { + "epoch": 0.5159699248120301, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2513, + "step": 13578 + }, + { + "epoch": 0.51603007518797, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 13579 + }, + { + "epoch": 0.5160902255639098, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2899, + "step": 13580 + }, + { + "epoch": 0.5161503759398496, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 13581 + }, + { + "epoch": 0.5162105263157895, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 13582 + }, + { + "epoch": 0.5162706766917293, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 13583 + }, + { + "epoch": 0.5163308270676692, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 13584 + }, + { + "epoch": 0.5163909774436091, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 13585 + }, + { + "epoch": 0.5164511278195488, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.186, + "step": 13586 + }, + { + "epoch": 0.5165112781954887, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 13587 + }, + { + "epoch": 0.5165714285714286, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 13588 + }, + { + "epoch": 0.5166315789473684, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 13589 + }, + { + "epoch": 0.5166917293233083, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 13590 + }, + { + "epoch": 0.5167518796992481, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 13591 + }, + { + "epoch": 0.516812030075188, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.216, + "step": 13592 + }, + { + "epoch": 0.5168721804511278, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 13593 + }, + { + "epoch": 0.5169323308270677, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 13594 + }, + { + "epoch": 0.5169924812030076, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 13595 + }, + { + "epoch": 0.5170526315789473, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2114, + "step": 13596 + }, + { + "epoch": 0.5171127819548872, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 13597 + }, + { + "epoch": 0.5171729323308271, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 13598 + }, + { + "epoch": 0.5172330827067669, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 13599 + }, + { + "epoch": 0.5172932330827068, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 13600 + }, + { + "epoch": 0.5173533834586466, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 13601 + }, + { + "epoch": 0.5174135338345864, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 13602 + }, + { + "epoch": 0.5174736842105263, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 13603 + }, + { + "epoch": 0.5175338345864662, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 13604 + }, + { + "epoch": 0.517593984962406, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 13605 + }, + { + "epoch": 0.5176541353383458, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 13606 + }, + { + "epoch": 0.5177142857142857, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 13607 + }, + { + "epoch": 0.5177744360902256, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 13608 + }, + { + "epoch": 0.5178345864661654, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 13609 + }, + { + "epoch": 0.5178947368421053, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1728, + "step": 13610 + }, + { + "epoch": 0.5179548872180452, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2678, + "step": 13611 + }, + { + "epoch": 0.5180150375939849, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 13612 + }, + { + "epoch": 0.5180751879699248, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1919, + "step": 13613 + }, + { + "epoch": 0.5181353383458647, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 13614 + }, + { + "epoch": 0.5181954887218045, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 13615 + }, + { + "epoch": 0.5182556390977444, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 13616 + }, + { + "epoch": 0.5183157894736842, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 13617 + }, + { + "epoch": 0.518375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 13618 + }, + { + "epoch": 0.5184360902255639, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 13619 + }, + { + "epoch": 0.5184962406015038, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 13620 + }, + { + "epoch": 0.5185563909774437, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2178, + "step": 13621 + }, + { + "epoch": 0.5186165413533834, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3278, + "step": 13622 + }, + { + "epoch": 0.5186766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 13623 + }, + { + "epoch": 0.5187368421052632, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 13624 + }, + { + "epoch": 0.518796992481203, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.2037, + "step": 13625 + }, + { + "epoch": 0.5188571428571429, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 13626 + }, + { + "epoch": 0.5189172932330827, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 13627 + }, + { + "epoch": 0.5189774436090225, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 13628 + }, + { + "epoch": 0.5190375939849624, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 13629 + }, + { + "epoch": 0.5190977443609023, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 13630 + }, + { + "epoch": 0.5191578947368422, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2287, + "step": 13631 + }, + { + "epoch": 0.5192180451127819, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 13632 + }, + { + "epoch": 0.5192781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 13633 + }, + { + "epoch": 0.5193383458646617, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 13634 + }, + { + "epoch": 0.5193984962406015, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 13635 + }, + { + "epoch": 0.5194586466165414, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 13636 + }, + { + "epoch": 0.5195187969924812, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 13637 + }, + { + "epoch": 0.519578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2447, + "step": 13638 + }, + { + "epoch": 0.5196390977443609, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 13639 + }, + { + "epoch": 0.5196992481203008, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1641, + "step": 13640 + }, + { + "epoch": 0.5197593984962406, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 13641 + }, + { + "epoch": 0.5198195488721804, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 13642 + }, + { + "epoch": 0.5198796992481203, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1845, + "step": 13643 + }, + { + "epoch": 0.5199398496240601, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 13644 + }, + { + "epoch": 0.52, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 13645 + }, + { + "epoch": 0.5200601503759399, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 13646 + }, + { + "epoch": 0.5201203007518796, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 13647 + }, + { + "epoch": 0.5201804511278195, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 13648 + }, + { + "epoch": 0.5202406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 13649 + }, + { + "epoch": 0.5203007518796993, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 13650 + }, + { + "epoch": 0.5203609022556391, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 13651 + }, + { + "epoch": 0.5204210526315789, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 13652 + }, + { + "epoch": 0.5204812030075188, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 13653 + }, + { + "epoch": 0.5205413533834586, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 13654 + }, + { + "epoch": 0.5206015037593985, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2918, + "step": 13655 + }, + { + "epoch": 0.5206616541353384, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 13656 + }, + { + "epoch": 0.5207218045112781, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 13657 + }, + { + "epoch": 0.520781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 13658 + }, + { + "epoch": 0.5208421052631579, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2312, + "step": 13659 + }, + { + "epoch": 0.5209022556390978, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 13660 + }, + { + "epoch": 0.5209624060150376, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 13661 + }, + { + "epoch": 0.5210225563909775, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1859, + "step": 13662 + }, + { + "epoch": 0.5210827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 13663 + }, + { + "epoch": 0.5211428571428571, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 13664 + }, + { + "epoch": 0.521203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 13665 + }, + { + "epoch": 0.5212631578947369, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 13666 + }, + { + "epoch": 0.5213233082706767, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2417, + "step": 13667 + }, + { + "epoch": 0.5213834586466165, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 13668 + }, + { + "epoch": 0.5214436090225564, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 13669 + }, + { + "epoch": 0.5215037593984962, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 13670 + }, + { + "epoch": 0.5215639097744361, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 13671 + }, + { + "epoch": 0.521624060150376, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1552, + "step": 13672 + }, + { + "epoch": 0.5216842105263158, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 13673 + }, + { + "epoch": 0.5217443609022556, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1966, + "step": 13674 + }, + { + "epoch": 0.5218045112781955, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 13675 + }, + { + "epoch": 0.5218646616541354, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2376, + "step": 13676 + }, + { + "epoch": 0.5219248120300752, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 13677 + }, + { + "epoch": 0.521984962406015, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2236, + "step": 13678 + }, + { + "epoch": 0.5220451127819549, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1671, + "step": 13679 + }, + { + "epoch": 0.5221052631578947, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 13680 + }, + { + "epoch": 0.5221654135338346, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3422, + "step": 13681 + }, + { + "epoch": 0.5222255639097745, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1868, + "step": 13682 + }, + { + "epoch": 0.5222857142857142, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 13683 + }, + { + "epoch": 0.5223458646616541, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 13684 + }, + { + "epoch": 0.522406015037594, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2187, + "step": 13685 + }, + { + "epoch": 0.5224661654135339, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 13686 + }, + { + "epoch": 0.5225263157894737, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 13687 + }, + { + "epoch": 0.5225864661654135, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2262, + "step": 13688 + }, + { + "epoch": 0.5226466165413534, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 13689 + }, + { + "epoch": 0.5227067669172932, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2011, + "step": 13690 + }, + { + "epoch": 0.5227669172932331, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 13691 + }, + { + "epoch": 0.522827067669173, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.22, + "step": 13692 + }, + { + "epoch": 0.5228872180451127, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 13693 + }, + { + "epoch": 0.5229473684210526, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2104, + "step": 13694 + }, + { + "epoch": 0.5230075187969925, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 13695 + }, + { + "epoch": 0.5230676691729323, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 13696 + }, + { + "epoch": 0.5231278195488722, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 13697 + }, + { + "epoch": 0.523187969924812, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 13698 + }, + { + "epoch": 0.5232481203007519, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1879, + "step": 13699 + }, + { + "epoch": 0.5233082706766917, + "grad_norm": 1.484375, + "learning_rate": 3e-05, + "loss": 1.2575, + "step": 13700 + }, + { + "epoch": 0.5233684210526316, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 13701 + }, + { + "epoch": 0.5234285714285715, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 13702 + }, + { + "epoch": 0.5234887218045112, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 13703 + }, + { + "epoch": 0.5235488721804511, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 13704 + }, + { + "epoch": 0.523609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 13705 + }, + { + "epoch": 0.5236691729323308, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 13706 + }, + { + "epoch": 0.5237293233082707, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 13707 + }, + { + "epoch": 0.5237894736842105, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.197, + "step": 13708 + }, + { + "epoch": 0.5238496240601503, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 13709 + }, + { + "epoch": 0.5239097744360902, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 13710 + }, + { + "epoch": 0.5239699248120301, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 13711 + }, + { + "epoch": 0.52403007518797, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2532, + "step": 13712 + }, + { + "epoch": 0.5240902255639098, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 13713 + }, + { + "epoch": 0.5241503759398496, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 13714 + }, + { + "epoch": 0.5242105263157895, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 13715 + }, + { + "epoch": 0.5242706766917293, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1961, + "step": 13716 + }, + { + "epoch": 0.5243308270676692, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 13717 + }, + { + "epoch": 0.5243909774436091, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 13718 + }, + { + "epoch": 0.5244511278195488, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 13719 + }, + { + "epoch": 0.5245112781954887, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 13720 + }, + { + "epoch": 0.5245714285714286, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1954, + "step": 13721 + }, + { + "epoch": 0.5246315789473684, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1467, + "step": 13722 + }, + { + "epoch": 0.5246917293233083, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 13723 + }, + { + "epoch": 0.5247518796992481, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1635, + "step": 13724 + }, + { + "epoch": 0.524812030075188, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 13725 + }, + { + "epoch": 0.5248721804511278, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1988, + "step": 13726 + }, + { + "epoch": 0.5249323308270677, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 13727 + }, + { + "epoch": 0.5249924812030076, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 13728 + }, + { + "epoch": 0.5250526315789473, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 13729 + }, + { + "epoch": 0.5251127819548872, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 13730 + }, + { + "epoch": 0.5251729323308271, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2601, + "step": 13731 + }, + { + "epoch": 0.5252330827067669, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2857, + "step": 13732 + }, + { + "epoch": 0.5252932330827068, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2913, + "step": 13733 + }, + { + "epoch": 0.5253533834586466, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 13734 + }, + { + "epoch": 0.5254135338345864, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 13735 + }, + { + "epoch": 0.5254736842105263, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 13736 + }, + { + "epoch": 0.5255338345864662, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 13737 + }, + { + "epoch": 0.5255939849624061, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 13738 + }, + { + "epoch": 0.5256541353383458, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2544, + "step": 13739 + }, + { + "epoch": 0.5257142857142857, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 13740 + }, + { + "epoch": 0.5257744360902256, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 13741 + }, + { + "epoch": 0.5258345864661654, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 13742 + }, + { + "epoch": 0.5258947368421053, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2441, + "step": 13743 + }, + { + "epoch": 0.5259548872180451, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1912, + "step": 13744 + }, + { + "epoch": 0.5260150375939849, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 13745 + }, + { + "epoch": 0.5260751879699248, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2661, + "step": 13746 + }, + { + "epoch": 0.5261353383458647, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 13747 + }, + { + "epoch": 0.5261954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 13748 + }, + { + "epoch": 0.5262556390977443, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2204, + "step": 13749 + }, + { + "epoch": 0.5263157894736842, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 13750 + }, + { + "epoch": 0.526375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 13751 + }, + { + "epoch": 0.5264360902255639, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1997, + "step": 13752 + }, + { + "epoch": 0.5264962406015038, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 13753 + }, + { + "epoch": 0.5265563909774436, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 13754 + }, + { + "epoch": 0.5266165413533834, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 13755 + }, + { + "epoch": 0.5266766917293233, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 13756 + }, + { + "epoch": 0.5267368421052632, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 13757 + }, + { + "epoch": 0.526796992481203, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 13758 + }, + { + "epoch": 0.5268571428571428, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2739, + "step": 13759 + }, + { + "epoch": 0.5269172932330827, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 13760 + }, + { + "epoch": 0.5269774436090225, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 13761 + }, + { + "epoch": 0.5270375939849624, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 13762 + }, + { + "epoch": 0.5270977443609023, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 13763 + }, + { + "epoch": 0.5271578947368422, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 13764 + }, + { + "epoch": 0.5272180451127819, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2013, + "step": 13765 + }, + { + "epoch": 0.5272781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 13766 + }, + { + "epoch": 0.5273383458646617, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 13767 + }, + { + "epoch": 0.5273984962406015, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 13768 + }, + { + "epoch": 0.5274586466165414, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 13769 + }, + { + "epoch": 0.5275187969924812, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 13770 + }, + { + "epoch": 0.527578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 13771 + }, + { + "epoch": 0.5276390977443609, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 13772 + }, + { + "epoch": 0.5276992481203008, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2277, + "step": 13773 + }, + { + "epoch": 0.5277593984962407, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2532, + "step": 13774 + }, + { + "epoch": 0.5278195488721804, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 13775 + }, + { + "epoch": 0.5278796992481203, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 13776 + }, + { + "epoch": 0.5279398496240602, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 13777 + }, + { + "epoch": 0.528, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 13778 + }, + { + "epoch": 0.5280601503759399, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 13779 + }, + { + "epoch": 0.5281203007518797, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 13780 + }, + { + "epoch": 0.5281804511278195, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2116, + "step": 13781 + }, + { + "epoch": 0.5282406015037594, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 13782 + }, + { + "epoch": 0.5283007518796993, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1868, + "step": 13783 + }, + { + "epoch": 0.5283609022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 13784 + }, + { + "epoch": 0.5284210526315789, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.2243, + "step": 13785 + }, + { + "epoch": 0.5284812030075188, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2142, + "step": 13786 + }, + { + "epoch": 0.5285413533834586, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 13787 + }, + { + "epoch": 0.5286015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 13788 + }, + { + "epoch": 0.5286616541353384, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 13789 + }, + { + "epoch": 0.5287218045112781, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.344, + "step": 13790 + }, + { + "epoch": 0.528781954887218, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 13791 + }, + { + "epoch": 0.5288421052631579, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 13792 + }, + { + "epoch": 0.5289022556390978, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2265, + "step": 13793 + }, + { + "epoch": 0.5289624060150376, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 13794 + }, + { + "epoch": 0.5290225563909774, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 13795 + }, + { + "epoch": 0.5290827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2761, + "step": 13796 + }, + { + "epoch": 0.5291428571428571, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 13797 + }, + { + "epoch": 0.529203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 13798 + }, + { + "epoch": 0.5292631578947369, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2567, + "step": 13799 + }, + { + "epoch": 0.5293233082706766, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 13800 + }, + { + "epoch": 0.5293834586466165, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 13801 + }, + { + "epoch": 0.5294436090225564, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1943, + "step": 13802 + }, + { + "epoch": 0.5295037593984963, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1995, + "step": 13803 + }, + { + "epoch": 0.5295639097744361, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1872, + "step": 13804 + }, + { + "epoch": 0.5296240601503759, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 13805 + }, + { + "epoch": 0.5296842105263158, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3133, + "step": 13806 + }, + { + "epoch": 0.5297443609022556, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 13807 + }, + { + "epoch": 0.5298045112781955, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 13808 + }, + { + "epoch": 0.5298646616541354, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1705, + "step": 13809 + }, + { + "epoch": 0.5299248120300752, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 13810 + }, + { + "epoch": 0.529984962406015, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 13811 + }, + { + "epoch": 0.5300451127819549, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 13812 + }, + { + "epoch": 0.5301052631578947, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1716, + "step": 13813 + }, + { + "epoch": 0.5301654135338346, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2427, + "step": 13814 + }, + { + "epoch": 0.5302255639097745, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2546, + "step": 13815 + }, + { + "epoch": 0.5302857142857142, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1718, + "step": 13816 + }, + { + "epoch": 0.5303458646616541, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1776, + "step": 13817 + }, + { + "epoch": 0.530406015037594, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 13818 + }, + { + "epoch": 0.5304661654135339, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 13819 + }, + { + "epoch": 0.5305263157894737, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 13820 + }, + { + "epoch": 0.5305864661654135, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 13821 + }, + { + "epoch": 0.5306466165413534, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2053, + "step": 13822 + }, + { + "epoch": 0.5307067669172932, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 13823 + }, + { + "epoch": 0.5307669172932331, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1725, + "step": 13824 + }, + { + "epoch": 0.530827067669173, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1839, + "step": 13825 + }, + { + "epoch": 0.5308872180451127, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 13826 + }, + { + "epoch": 0.5309473684210526, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 13827 + }, + { + "epoch": 0.5310075187969925, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2572, + "step": 13828 + }, + { + "epoch": 0.5310676691729324, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 13829 + }, + { + "epoch": 0.5311278195488722, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2088, + "step": 13830 + }, + { + "epoch": 0.531187969924812, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2673, + "step": 13831 + }, + { + "epoch": 0.5312481203007519, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 13832 + }, + { + "epoch": 0.5313082706766917, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 13833 + }, + { + "epoch": 0.5313684210526316, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 13834 + }, + { + "epoch": 0.5314285714285715, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 13835 + }, + { + "epoch": 0.5314887218045112, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 13836 + }, + { + "epoch": 0.5315488721804511, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 13837 + }, + { + "epoch": 0.531609022556391, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 13838 + }, + { + "epoch": 0.5316691729323308, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1955, + "step": 13839 + }, + { + "epoch": 0.5317293233082707, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 13840 + }, + { + "epoch": 0.5317894736842105, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2695, + "step": 13841 + }, + { + "epoch": 0.5318496240601503, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 13842 + }, + { + "epoch": 0.5319097744360902, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2647, + "step": 13843 + }, + { + "epoch": 0.5319699248120301, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 13844 + }, + { + "epoch": 0.53203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 13845 + }, + { + "epoch": 0.5320902255639097, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 13846 + }, + { + "epoch": 0.5321503759398496, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1918, + "step": 13847 + }, + { + "epoch": 0.5322105263157895, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 13848 + }, + { + "epoch": 0.5322706766917293, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1357, + "step": 13849 + }, + { + "epoch": 0.5323308270676692, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 13850 + }, + { + "epoch": 0.532390977443609, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 13851 + }, + { + "epoch": 0.5324511278195488, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1908, + "step": 13852 + }, + { + "epoch": 0.5325112781954887, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 13853 + }, + { + "epoch": 0.5325714285714286, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.26, + "step": 13854 + }, + { + "epoch": 0.5326315789473685, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2086, + "step": 13855 + }, + { + "epoch": 0.5326917293233082, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1326, + "step": 13856 + }, + { + "epoch": 0.5327518796992481, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 13857 + }, + { + "epoch": 0.532812030075188, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 13858 + }, + { + "epoch": 0.5328721804511278, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 13859 + }, + { + "epoch": 0.5329323308270677, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 13860 + }, + { + "epoch": 0.5329924812030076, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 13861 + }, + { + "epoch": 0.5330526315789473, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1768, + "step": 13862 + }, + { + "epoch": 0.5331127819548872, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 13863 + }, + { + "epoch": 0.5331729323308271, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 13864 + }, + { + "epoch": 0.533233082706767, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1823, + "step": 13865 + }, + { + "epoch": 0.5332932330827068, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 13866 + }, + { + "epoch": 0.5333533834586466, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 13867 + }, + { + "epoch": 0.5334135338345865, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 13868 + }, + { + "epoch": 0.5334736842105263, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1653, + "step": 13869 + }, + { + "epoch": 0.5335338345864662, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 13870 + }, + { + "epoch": 0.5335939849624061, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 13871 + }, + { + "epoch": 0.5336541353383458, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.197, + "step": 13872 + }, + { + "epoch": 0.5337142857142857, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.163, + "step": 13873 + }, + { + "epoch": 0.5337744360902256, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1722, + "step": 13874 + }, + { + "epoch": 0.5338345864661654, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 13875 + }, + { + "epoch": 0.5338947368421053, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 13876 + }, + { + "epoch": 0.5339548872180451, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1692, + "step": 13877 + }, + { + "epoch": 0.5340150375939849, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 13878 + }, + { + "epoch": 0.5340751879699248, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 13879 + }, + { + "epoch": 0.5341353383458647, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 13880 + }, + { + "epoch": 0.5341954887218046, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.2348, + "step": 13881 + }, + { + "epoch": 0.5342556390977443, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 13882 + }, + { + "epoch": 0.5343157894736842, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 13883 + }, + { + "epoch": 0.5343759398496241, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 13884 + }, + { + "epoch": 0.5344360902255639, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2339, + "step": 13885 + }, + { + "epoch": 0.5344962406015038, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 13886 + }, + { + "epoch": 0.5345563909774436, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2235, + "step": 13887 + }, + { + "epoch": 0.5346165413533834, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2614, + "step": 13888 + }, + { + "epoch": 0.5346766917293233, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 13889 + }, + { + "epoch": 0.5347368421052632, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 13890 + }, + { + "epoch": 0.534796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 13891 + }, + { + "epoch": 0.5348571428571428, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 13892 + }, + { + "epoch": 0.5349172932330827, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 13893 + }, + { + "epoch": 0.5349774436090226, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 13894 + }, + { + "epoch": 0.5350375939849624, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 13895 + }, + { + "epoch": 0.5350977443609023, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 13896 + }, + { + "epoch": 0.535157894736842, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 13897 + }, + { + "epoch": 0.5352180451127819, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 13898 + }, + { + "epoch": 0.5352781954887218, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 13899 + }, + { + "epoch": 0.5353383458646617, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1814, + "step": 13900 + }, + { + "epoch": 0.5353984962406015, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2278, + "step": 13901 + }, + { + "epoch": 0.5354586466165413, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 13902 + }, + { + "epoch": 0.5355187969924812, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 13903 + }, + { + "epoch": 0.535578947368421, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2142, + "step": 13904 + }, + { + "epoch": 0.5356390977443609, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 13905 + }, + { + "epoch": 0.5356992481203008, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 13906 + }, + { + "epoch": 0.5357593984962405, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 13907 + }, + { + "epoch": 0.5358195488721804, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 13908 + }, + { + "epoch": 0.5358796992481203, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2287, + "step": 13909 + }, + { + "epoch": 0.5359398496240602, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2145, + "step": 13910 + }, + { + "epoch": 0.536, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 13911 + }, + { + "epoch": 0.5360601503759399, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 13912 + }, + { + "epoch": 0.5361203007518797, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 13913 + }, + { + "epoch": 0.5361804511278195, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1786, + "step": 13914 + }, + { + "epoch": 0.5362406015037594, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1638, + "step": 13915 + }, + { + "epoch": 0.5363007518796993, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1751, + "step": 13916 + }, + { + "epoch": 0.5363609022556391, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 13917 + }, + { + "epoch": 0.5364210526315789, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1991, + "step": 13918 + }, + { + "epoch": 0.5364812030075188, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 13919 + }, + { + "epoch": 0.5365413533834587, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2127, + "step": 13920 + }, + { + "epoch": 0.5366015037593985, + "grad_norm": 1.6328125, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 13921 + }, + { + "epoch": 0.5366616541353384, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 13922 + }, + { + "epoch": 0.5367218045112782, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 13923 + }, + { + "epoch": 0.536781954887218, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 13924 + }, + { + "epoch": 0.5368421052631579, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9511, + "step": 13925 + }, + { + "epoch": 0.5369022556390978, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 13926 + }, + { + "epoch": 0.5369624060150376, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2664, + "step": 13927 + }, + { + "epoch": 0.5370225563909774, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 13928 + }, + { + "epoch": 0.5370827067669173, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 13929 + }, + { + "epoch": 0.5371428571428571, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1537, + "step": 13930 + }, + { + "epoch": 0.537203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 13931 + }, + { + "epoch": 0.5372631578947369, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.238, + "step": 13932 + }, + { + "epoch": 0.5373233082706766, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 13933 + }, + { + "epoch": 0.5373834586466165, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2124, + "step": 13934 + }, + { + "epoch": 0.5374436090225564, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 13935 + }, + { + "epoch": 0.5375037593984963, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 13936 + }, + { + "epoch": 0.5375639097744361, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 13937 + }, + { + "epoch": 0.5376240601503759, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 13938 + }, + { + "epoch": 0.5376842105263158, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 13939 + }, + { + "epoch": 0.5377443609022556, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2197, + "step": 13940 + }, + { + "epoch": 0.5378045112781955, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 13941 + }, + { + "epoch": 0.5378646616541354, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 13942 + }, + { + "epoch": 0.5379248120300751, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1865, + "step": 13943 + }, + { + "epoch": 0.537984962406015, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 13944 + }, + { + "epoch": 0.5380451127819549, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 13945 + }, + { + "epoch": 0.5381052631578948, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 13946 + }, + { + "epoch": 0.5381654135338346, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2779, + "step": 13947 + }, + { + "epoch": 0.5382255639097744, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.197, + "step": 13948 + }, + { + "epoch": 0.5382857142857143, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 13949 + }, + { + "epoch": 0.5383458646616541, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 13950 + }, + { + "epoch": 0.538406015037594, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 13951 + }, + { + "epoch": 0.5384661654135339, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2139, + "step": 13952 + }, + { + "epoch": 0.5385263157894736, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 13953 + }, + { + "epoch": 0.5385864661654135, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1727, + "step": 13954 + }, + { + "epoch": 0.5386466165413534, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 13955 + }, + { + "epoch": 0.5387067669172932, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 13956 + }, + { + "epoch": 0.5387669172932331, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 13957 + }, + { + "epoch": 0.538827067669173, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 13958 + }, + { + "epoch": 0.5388872180451127, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 13959 + }, + { + "epoch": 0.5389473684210526, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 13960 + }, + { + "epoch": 0.5390075187969925, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 13961 + }, + { + "epoch": 0.5390676691729324, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1832, + "step": 13962 + }, + { + "epoch": 0.5391278195488722, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 13963 + }, + { + "epoch": 0.539187969924812, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 13964 + }, + { + "epoch": 0.5392481203007519, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 13965 + }, + { + "epoch": 0.5393082706766917, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2651, + "step": 13966 + }, + { + "epoch": 0.5393684210526316, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 13967 + }, + { + "epoch": 0.5394285714285715, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2093, + "step": 13968 + }, + { + "epoch": 0.5394887218045112, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 13969 + }, + { + "epoch": 0.5395488721804511, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 13970 + }, + { + "epoch": 0.539609022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 13971 + }, + { + "epoch": 0.5396691729323309, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 13972 + }, + { + "epoch": 0.5397293233082707, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 13973 + }, + { + "epoch": 0.5397894736842105, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 13974 + }, + { + "epoch": 0.5398496240601504, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 13975 + }, + { + "epoch": 0.5399097744360902, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 13976 + }, + { + "epoch": 0.5399699248120301, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 13977 + }, + { + "epoch": 0.54003007518797, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 13978 + }, + { + "epoch": 0.5400902255639097, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 13979 + }, + { + "epoch": 0.5401503759398496, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 13980 + }, + { + "epoch": 0.5402105263157895, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 13981 + }, + { + "epoch": 0.5402706766917293, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 13982 + }, + { + "epoch": 0.5403308270676692, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 13983 + }, + { + "epoch": 0.540390977443609, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 13984 + }, + { + "epoch": 0.5404511278195488, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 13985 + }, + { + "epoch": 0.5405112781954887, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1677, + "step": 13986 + }, + { + "epoch": 0.5405714285714286, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 13987 + }, + { + "epoch": 0.5406315789473685, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 13988 + }, + { + "epoch": 0.5406917293233082, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 13989 + }, + { + "epoch": 0.5407518796992481, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 13990 + }, + { + "epoch": 0.540812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 13991 + }, + { + "epoch": 0.5408721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 13992 + }, + { + "epoch": 0.5409323308270677, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2074, + "step": 13993 + }, + { + "epoch": 0.5409924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 13994 + }, + { + "epoch": 0.5410526315789473, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1859, + "step": 13995 + }, + { + "epoch": 0.5411127819548872, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 13996 + }, + { + "epoch": 0.5411729323308271, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 13997 + }, + { + "epoch": 0.541233082706767, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1692, + "step": 13998 + }, + { + "epoch": 0.5412932330827067, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 13999 + }, + { + "epoch": 0.5413533834586466, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 14000 + }, + { + "epoch": 0.5414135338345865, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 14001 + }, + { + "epoch": 0.5414736842105263, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 14002 + }, + { + "epoch": 0.5415338345864662, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2225, + "step": 14003 + }, + { + "epoch": 0.541593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 14004 + }, + { + "epoch": 0.5416541353383458, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 14005 + }, + { + "epoch": 0.5417142857142857, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 14006 + }, + { + "epoch": 0.5417744360902256, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1884, + "step": 14007 + }, + { + "epoch": 0.5418345864661654, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 14008 + }, + { + "epoch": 0.5418947368421053, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.3873, + "step": 14009 + }, + { + "epoch": 0.5419548872180451, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2185, + "step": 14010 + }, + { + "epoch": 0.542015037593985, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 14011 + }, + { + "epoch": 0.5420751879699248, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2303, + "step": 14012 + }, + { + "epoch": 0.5421353383458647, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2002, + "step": 14013 + }, + { + "epoch": 0.5421954887218046, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 14014 + }, + { + "epoch": 0.5422556390977443, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 14015 + }, + { + "epoch": 0.5423157894736842, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 14016 + }, + { + "epoch": 0.5423759398496241, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 14017 + }, + { + "epoch": 0.5424360902255639, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 14018 + }, + { + "epoch": 0.5424962406015038, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 14019 + }, + { + "epoch": 0.5425563909774436, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 14020 + }, + { + "epoch": 0.5426165413533834, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 14021 + }, + { + "epoch": 0.5426766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 14022 + }, + { + "epoch": 0.5427368421052632, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 14023 + }, + { + "epoch": 0.5427969924812031, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.148, + "step": 14024 + }, + { + "epoch": 0.5428571428571428, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 14025 + }, + { + "epoch": 0.5429172932330827, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 14026 + }, + { + "epoch": 0.5429774436090226, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 14027 + }, + { + "epoch": 0.5430375939849624, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 14028 + }, + { + "epoch": 0.5430977443609023, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2221, + "step": 14029 + }, + { + "epoch": 0.5431578947368421, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2136, + "step": 14030 + }, + { + "epoch": 0.5432180451127819, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 14031 + }, + { + "epoch": 0.5432781954887218, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 14032 + }, + { + "epoch": 0.5433383458646617, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 14033 + }, + { + "epoch": 0.5433984962406015, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 14034 + }, + { + "epoch": 0.5434586466165413, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.186, + "step": 14035 + }, + { + "epoch": 0.5435187969924812, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 14036 + }, + { + "epoch": 0.543578947368421, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 14037 + }, + { + "epoch": 0.5436390977443609, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 14038 + }, + { + "epoch": 0.5436992481203008, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1811, + "step": 14039 + }, + { + "epoch": 0.5437593984962406, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 14040 + }, + { + "epoch": 0.5438195488721804, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 14041 + }, + { + "epoch": 0.5438796992481203, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 14042 + }, + { + "epoch": 0.5439398496240602, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 14043 + }, + { + "epoch": 0.544, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 14044 + }, + { + "epoch": 0.5440601503759398, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 14045 + }, + { + "epoch": 0.5441203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 14046 + }, + { + "epoch": 0.5441804511278195, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.4851, + "step": 14047 + }, + { + "epoch": 0.5442406015037594, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 14048 + }, + { + "epoch": 0.5443007518796993, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 14049 + }, + { + "epoch": 0.544360902255639, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 14050 + }, + { + "epoch": 0.5444210526315789, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 14051 + }, + { + "epoch": 0.5444812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2598, + "step": 14052 + }, + { + "epoch": 0.5445413533834587, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 14053 + }, + { + "epoch": 0.5446015037593985, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 14054 + }, + { + "epoch": 0.5446616541353383, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 14055 + }, + { + "epoch": 0.5447218045112782, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 14056 + }, + { + "epoch": 0.544781954887218, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2548, + "step": 14057 + }, + { + "epoch": 0.5448421052631579, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2127, + "step": 14058 + }, + { + "epoch": 0.5449022556390978, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 14059 + }, + { + "epoch": 0.5449624060150376, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 14060 + }, + { + "epoch": 0.5450225563909774, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1994, + "step": 14061 + }, + { + "epoch": 0.5450827067669173, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 14062 + }, + { + "epoch": 0.5451428571428572, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3153, + "step": 14063 + }, + { + "epoch": 0.545203007518797, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 14064 + }, + { + "epoch": 0.5452631578947369, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 14065 + }, + { + "epoch": 0.5453233082706767, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 14066 + }, + { + "epoch": 0.5453834586466165, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 14067 + }, + { + "epoch": 0.5454436090225564, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 14068 + }, + { + "epoch": 0.5455037593984963, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 14069 + }, + { + "epoch": 0.5455639097744361, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2383, + "step": 14070 + }, + { + "epoch": 0.5456240601503759, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 14071 + }, + { + "epoch": 0.5456842105263158, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 14072 + }, + { + "epoch": 0.5457443609022556, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 14073 + }, + { + "epoch": 0.5458045112781955, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2413, + "step": 14074 + }, + { + "epoch": 0.5458646616541354, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 14075 + }, + { + "epoch": 0.5459248120300751, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2039, + "step": 14076 + }, + { + "epoch": 0.545984962406015, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 14077 + }, + { + "epoch": 0.5460451127819549, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.147, + "step": 14078 + }, + { + "epoch": 0.5461052631578948, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2092, + "step": 14079 + }, + { + "epoch": 0.5461654135338346, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 14080 + }, + { + "epoch": 0.5462255639097744, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 14081 + }, + { + "epoch": 0.5462857142857143, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 14082 + }, + { + "epoch": 0.5463458646616541, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2047, + "step": 14083 + }, + { + "epoch": 0.546406015037594, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 14084 + }, + { + "epoch": 0.5464661654135339, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 14085 + }, + { + "epoch": 0.5465263157894736, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 14086 + }, + { + "epoch": 0.5465864661654135, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 14087 + }, + { + "epoch": 0.5466466165413534, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1541, + "step": 14088 + }, + { + "epoch": 0.5467067669172933, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 14089 + }, + { + "epoch": 0.5467669172932331, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 14090 + }, + { + "epoch": 0.5468270676691729, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2447, + "step": 14091 + }, + { + "epoch": 0.5468872180451128, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 14092 + }, + { + "epoch": 0.5469473684210526, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 14093 + }, + { + "epoch": 0.5470075187969925, + "grad_norm": 1.390625, + "learning_rate": 3e-05, + "loss": 1.2396, + "step": 14094 + }, + { + "epoch": 0.5470676691729324, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 14095 + }, + { + "epoch": 0.5471278195488721, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 14096 + }, + { + "epoch": 0.547187969924812, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1518, + "step": 14097 + }, + { + "epoch": 0.5472481203007519, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1953, + "step": 14098 + }, + { + "epoch": 0.5473082706766917, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1845, + "step": 14099 + }, + { + "epoch": 0.5473684210526316, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 14100 + }, + { + "epoch": 0.5474285714285714, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 14101 + }, + { + "epoch": 0.5474887218045112, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 14102 + }, + { + "epoch": 0.5475488721804511, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 14103 + }, + { + "epoch": 0.547609022556391, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1564, + "step": 14104 + }, + { + "epoch": 0.5476691729323309, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1604, + "step": 14105 + }, + { + "epoch": 0.5477293233082706, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 14106 + }, + { + "epoch": 0.5477894736842105, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.188, + "step": 14107 + }, + { + "epoch": 0.5478496240601504, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 14108 + }, + { + "epoch": 0.5479097744360902, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 14109 + }, + { + "epoch": 0.5479699248120301, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 14110 + }, + { + "epoch": 0.54803007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1959, + "step": 14111 + }, + { + "epoch": 0.5480902255639097, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 14112 + }, + { + "epoch": 0.5481503759398496, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 14113 + }, + { + "epoch": 0.5482105263157895, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 14114 + }, + { + "epoch": 0.5482706766917294, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 14115 + }, + { + "epoch": 0.5483308270676692, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1778, + "step": 14116 + }, + { + "epoch": 0.548390977443609, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 14117 + }, + { + "epoch": 0.5484511278195489, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 14118 + }, + { + "epoch": 0.5485112781954887, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 14119 + }, + { + "epoch": 0.5485714285714286, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 14120 + }, + { + "epoch": 0.5486315789473685, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2173, + "step": 14121 + }, + { + "epoch": 0.5486917293233082, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 14122 + }, + { + "epoch": 0.5487518796992481, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 14123 + }, + { + "epoch": 0.548812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 14124 + }, + { + "epoch": 0.5488721804511278, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 14125 + }, + { + "epoch": 0.5489323308270677, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 14126 + }, + { + "epoch": 0.5489924812030075, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 14127 + }, + { + "epoch": 0.5490526315789473, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 14128 + }, + { + "epoch": 0.5491127819548872, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 14129 + }, + { + "epoch": 0.5491729323308271, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 14130 + }, + { + "epoch": 0.549233082706767, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 14131 + }, + { + "epoch": 0.5492932330827067, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1881, + "step": 14132 + }, + { + "epoch": 0.5493533834586466, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 14133 + }, + { + "epoch": 0.5494135338345865, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 14134 + }, + { + "epoch": 0.5494736842105263, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2341, + "step": 14135 + }, + { + "epoch": 0.5495338345864662, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2377, + "step": 14136 + }, + { + "epoch": 0.549593984962406, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 14137 + }, + { + "epoch": 0.5496541353383458, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 14138 + }, + { + "epoch": 0.5497142857142857, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 14139 + }, + { + "epoch": 0.5497744360902256, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 14140 + }, + { + "epoch": 0.5498345864661655, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2801, + "step": 14141 + }, + { + "epoch": 0.5498947368421052, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 14142 + }, + { + "epoch": 0.5499548872180451, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1482, + "step": 14143 + }, + { + "epoch": 0.550015037593985, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 14144 + }, + { + "epoch": 0.5500751879699248, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 14145 + }, + { + "epoch": 0.5501353383458647, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 14146 + }, + { + "epoch": 0.5501954887218045, + "grad_norm": 1.4453125, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 14147 + }, + { + "epoch": 0.5502556390977443, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 14148 + }, + { + "epoch": 0.5503157894736842, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.2074, + "step": 14149 + }, + { + "epoch": 0.5503759398496241, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 14150 + }, + { + "epoch": 0.550436090225564, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 14151 + }, + { + "epoch": 0.5504962406015037, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 14152 + }, + { + "epoch": 0.5505563909774436, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 14153 + }, + { + "epoch": 0.5506165413533834, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 14154 + }, + { + "epoch": 0.5506766917293233, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2369, + "step": 14155 + }, + { + "epoch": 0.5507368421052632, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 14156 + }, + { + "epoch": 0.5507969924812031, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2047, + "step": 14157 + }, + { + "epoch": 0.5508571428571428, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 14158 + }, + { + "epoch": 0.5509172932330827, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 14159 + }, + { + "epoch": 0.5509774436090226, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 14160 + }, + { + "epoch": 0.5510375939849624, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 14161 + }, + { + "epoch": 0.5510977443609023, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 14162 + }, + { + "epoch": 0.5511578947368421, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 14163 + }, + { + "epoch": 0.5512180451127819, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 14164 + }, + { + "epoch": 0.5512781954887218, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 14165 + }, + { + "epoch": 0.5513383458646617, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 14166 + }, + { + "epoch": 0.5513984962406016, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 14167 + }, + { + "epoch": 0.5514586466165413, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 14168 + }, + { + "epoch": 0.5515187969924812, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 14169 + }, + { + "epoch": 0.5515789473684211, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 14170 + }, + { + "epoch": 0.5516390977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 14171 + }, + { + "epoch": 0.5516992481203008, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 14172 + }, + { + "epoch": 0.5517593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 14173 + }, + { + "epoch": 0.5518195488721804, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 14174 + }, + { + "epoch": 0.5518796992481203, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 14175 + }, + { + "epoch": 0.5519398496240602, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 14176 + }, + { + "epoch": 0.552, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 14177 + }, + { + "epoch": 0.5520601503759398, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 14178 + }, + { + "epoch": 0.5521203007518797, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 14179 + }, + { + "epoch": 0.5521804511278195, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 14180 + }, + { + "epoch": 0.5522406015037594, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2397, + "step": 14181 + }, + { + "epoch": 0.5523007518796993, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 14182 + }, + { + "epoch": 0.552360902255639, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 14183 + }, + { + "epoch": 0.5524210526315789, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1896, + "step": 14184 + }, + { + "epoch": 0.5524812030075188, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 14185 + }, + { + "epoch": 0.5525413533834587, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2166, + "step": 14186 + }, + { + "epoch": 0.5526015037593985, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 14187 + }, + { + "epoch": 0.5526616541353383, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 14188 + }, + { + "epoch": 0.5527218045112782, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 14189 + }, + { + "epoch": 0.552781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 14190 + }, + { + "epoch": 0.5528421052631579, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2822, + "step": 14191 + }, + { + "epoch": 0.5529022556390978, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1671, + "step": 14192 + }, + { + "epoch": 0.5529624060150375, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2259, + "step": 14193 + }, + { + "epoch": 0.5530225563909774, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 14194 + }, + { + "epoch": 0.5530827067669173, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 14195 + }, + { + "epoch": 0.5531428571428572, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 14196 + }, + { + "epoch": 0.553203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 14197 + }, + { + "epoch": 0.5532631578947368, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 14198 + }, + { + "epoch": 0.5533233082706767, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 14199 + }, + { + "epoch": 0.5533834586466165, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 14200 + }, + { + "epoch": 0.5534436090225564, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 14201 + }, + { + "epoch": 0.5535037593984963, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.189, + "step": 14202 + }, + { + "epoch": 0.553563909774436, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 14203 + }, + { + "epoch": 0.5536240601503759, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 14204 + }, + { + "epoch": 0.5536842105263158, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2061, + "step": 14205 + }, + { + "epoch": 0.5537443609022557, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 14206 + }, + { + "epoch": 0.5538045112781955, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 14207 + }, + { + "epoch": 0.5538646616541354, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 14208 + }, + { + "epoch": 0.5539248120300752, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 14209 + }, + { + "epoch": 0.553984962406015, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 14210 + }, + { + "epoch": 0.5540451127819549, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 14211 + }, + { + "epoch": 0.5541052631578948, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 14212 + }, + { + "epoch": 0.5541654135338346, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2431, + "step": 14213 + }, + { + "epoch": 0.5542255639097744, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 14214 + }, + { + "epoch": 0.5542857142857143, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 14215 + }, + { + "epoch": 0.5543458646616541, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 14216 + }, + { + "epoch": 0.554406015037594, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 14217 + }, + { + "epoch": 0.5544661654135339, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 14218 + }, + { + "epoch": 0.5545263157894736, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 14219 + }, + { + "epoch": 0.5545864661654135, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 14220 + }, + { + "epoch": 0.5546466165413534, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 14221 + }, + { + "epoch": 0.5547067669172933, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 14222 + }, + { + "epoch": 0.5547669172932331, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1817, + "step": 14223 + }, + { + "epoch": 0.5548270676691729, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 14224 + }, + { + "epoch": 0.5548872180451128, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 14225 + }, + { + "epoch": 0.5549473684210526, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 14226 + }, + { + "epoch": 0.5550075187969925, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 14227 + }, + { + "epoch": 0.5550676691729324, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 14228 + }, + { + "epoch": 0.5551278195488721, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 14229 + }, + { + "epoch": 0.555187969924812, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 14230 + }, + { + "epoch": 0.5552481203007519, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 14231 + }, + { + "epoch": 0.5553082706766918, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.156, + "step": 14232 + }, + { + "epoch": 0.5553684210526316, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 14233 + }, + { + "epoch": 0.5554285714285714, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 14234 + }, + { + "epoch": 0.5554887218045113, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 14235 + }, + { + "epoch": 0.5555488721804511, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2317, + "step": 14236 + }, + { + "epoch": 0.555609022556391, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 14237 + }, + { + "epoch": 0.5556691729323309, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 14238 + }, + { + "epoch": 0.5557293233082706, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 14239 + }, + { + "epoch": 0.5557894736842105, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 14240 + }, + { + "epoch": 0.5558496240601504, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 14241 + }, + { + "epoch": 0.5559097744360902, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1711, + "step": 14242 + }, + { + "epoch": 0.5559699248120301, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2107, + "step": 14243 + }, + { + "epoch": 0.5560300751879699, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2522, + "step": 14244 + }, + { + "epoch": 0.5560902255639097, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 14245 + }, + { + "epoch": 0.5561503759398496, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 14246 + }, + { + "epoch": 0.5562105263157895, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 14247 + }, + { + "epoch": 0.5562706766917294, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1704, + "step": 14248 + }, + { + "epoch": 0.5563308270676691, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1314, + "step": 14249 + }, + { + "epoch": 0.556390977443609, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 14250 + }, + { + "epoch": 0.5564511278195489, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1604, + "step": 14251 + }, + { + "epoch": 0.5565112781954887, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 14252 + }, + { + "epoch": 0.5565714285714286, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 14253 + }, + { + "epoch": 0.5566315789473684, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2064, + "step": 14254 + }, + { + "epoch": 0.5566917293233082, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 14255 + }, + { + "epoch": 0.5567518796992481, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 14256 + }, + { + "epoch": 0.556812030075188, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2386, + "step": 14257 + }, + { + "epoch": 0.5568721804511279, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2156, + "step": 14258 + }, + { + "epoch": 0.5569323308270677, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 14259 + }, + { + "epoch": 0.5569924812030075, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 14260 + }, + { + "epoch": 0.5570526315789474, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 14261 + }, + { + "epoch": 0.5571127819548872, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 14262 + }, + { + "epoch": 0.5571729323308271, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 14263 + }, + { + "epoch": 0.557233082706767, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 14264 + }, + { + "epoch": 0.5572932330827067, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 14265 + }, + { + "epoch": 0.5573533834586466, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 14266 + }, + { + "epoch": 0.5574135338345865, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2324, + "step": 14267 + }, + { + "epoch": 0.5574736842105263, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 14268 + }, + { + "epoch": 0.5575338345864662, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 14269 + }, + { + "epoch": 0.557593984962406, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 14270 + }, + { + "epoch": 0.5576541353383458, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 14271 + }, + { + "epoch": 0.5577142857142857, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1636, + "step": 14272 + }, + { + "epoch": 0.5577744360902256, + "grad_norm": 1.703125, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 14273 + }, + { + "epoch": 0.5578345864661655, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 14274 + }, + { + "epoch": 0.5578947368421052, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 14275 + }, + { + "epoch": 0.5579548872180451, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.3861, + "step": 14276 + }, + { + "epoch": 0.558015037593985, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 14277 + }, + { + "epoch": 0.5580751879699248, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2557, + "step": 14278 + }, + { + "epoch": 0.5581353383458647, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 14279 + }, + { + "epoch": 0.5581954887218045, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1928, + "step": 14280 + }, + { + "epoch": 0.5582556390977443, + "grad_norm": 1.484375, + "learning_rate": 3e-05, + "loss": 1.1764, + "step": 14281 + }, + { + "epoch": 0.5583157894736842, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 14282 + }, + { + "epoch": 0.5583759398496241, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1646, + "step": 14283 + }, + { + "epoch": 0.558436090225564, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 14284 + }, + { + "epoch": 0.5584962406015037, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 14285 + }, + { + "epoch": 0.5585563909774436, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 14286 + }, + { + "epoch": 0.5586165413533835, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1727, + "step": 14287 + }, + { + "epoch": 0.5586766917293233, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.396, + "step": 14288 + }, + { + "epoch": 0.5587368421052632, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 14289 + }, + { + "epoch": 0.558796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 14290 + }, + { + "epoch": 0.5588571428571428, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2564, + "step": 14291 + }, + { + "epoch": 0.5589172932330827, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2262, + "step": 14292 + }, + { + "epoch": 0.5589774436090226, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2053, + "step": 14293 + }, + { + "epoch": 0.5590375939849624, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 14294 + }, + { + "epoch": 0.5590977443609022, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 14295 + }, + { + "epoch": 0.5591578947368421, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 14296 + }, + { + "epoch": 0.559218045112782, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 14297 + }, + { + "epoch": 0.5592781954887218, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.3208, + "step": 14298 + }, + { + "epoch": 0.5593383458646617, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 14299 + }, + { + "epoch": 0.5593984962406015, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2752, + "step": 14300 + }, + { + "epoch": 0.5594586466165413, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 14301 + }, + { + "epoch": 0.5595187969924812, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2404, + "step": 14302 + }, + { + "epoch": 0.5595789473684211, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 14303 + }, + { + "epoch": 0.5596390977443609, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 14304 + }, + { + "epoch": 0.5596992481203007, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 14305 + }, + { + "epoch": 0.5597593984962406, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 14306 + }, + { + "epoch": 0.5598195488721804, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 14307 + }, + { + "epoch": 0.5598796992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 14308 + }, + { + "epoch": 0.5599398496240602, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1908, + "step": 14309 + }, + { + "epoch": 0.56, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 14310 + }, + { + "epoch": 0.5600601503759398, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 14311 + }, + { + "epoch": 0.5601203007518797, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 14312 + }, + { + "epoch": 0.5601804511278196, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1637, + "step": 14313 + }, + { + "epoch": 0.5602406015037594, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 14314 + }, + { + "epoch": 0.5603007518796993, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2793, + "step": 14315 + }, + { + "epoch": 0.5603609022556391, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1768, + "step": 14316 + }, + { + "epoch": 0.5604210526315789, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 14317 + }, + { + "epoch": 0.5604812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 14318 + }, + { + "epoch": 0.5605413533834587, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 14319 + }, + { + "epoch": 0.5606015037593985, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 14320 + }, + { + "epoch": 0.5606616541353383, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 14321 + }, + { + "epoch": 0.5607218045112782, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 14322 + }, + { + "epoch": 0.560781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 14323 + }, + { + "epoch": 0.5608421052631579, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 14324 + }, + { + "epoch": 0.5609022556390978, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 14325 + }, + { + "epoch": 0.5609624060150376, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 14326 + }, + { + "epoch": 0.5610225563909774, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 14327 + }, + { + "epoch": 0.5610827067669173, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 14328 + }, + { + "epoch": 0.5611428571428572, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 14329 + }, + { + "epoch": 0.561203007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2225, + "step": 14330 + }, + { + "epoch": 0.5612631578947368, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 14331 + }, + { + "epoch": 0.5613233082706767, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2017, + "step": 14332 + }, + { + "epoch": 0.5613834586466165, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2329, + "step": 14333 + }, + { + "epoch": 0.5614436090225564, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1785, + "step": 14334 + }, + { + "epoch": 0.5615037593984963, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 14335 + }, + { + "epoch": 0.561563909774436, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2952, + "step": 14336 + }, + { + "epoch": 0.5616240601503759, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 14337 + }, + { + "epoch": 0.5616842105263158, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 14338 + }, + { + "epoch": 0.5617443609022557, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 14339 + }, + { + "epoch": 0.5618045112781955, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 14340 + }, + { + "epoch": 0.5618646616541353, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 14341 + }, + { + "epoch": 0.5619248120300752, + "grad_norm": 1.9375, + "learning_rate": 3e-05, + "loss": 1.1772, + "step": 14342 + }, + { + "epoch": 0.561984962406015, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2868, + "step": 14343 + }, + { + "epoch": 0.5620451127819549, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 14344 + }, + { + "epoch": 0.5621052631578948, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 14345 + }, + { + "epoch": 0.5621654135338345, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 14346 + }, + { + "epoch": 0.5622255639097744, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 14347 + }, + { + "epoch": 0.5622857142857143, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 14348 + }, + { + "epoch": 0.5623458646616541, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 14349 + }, + { + "epoch": 0.562406015037594, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2312, + "step": 14350 + }, + { + "epoch": 0.5624661654135338, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 14351 + }, + { + "epoch": 0.5625263157894737, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 14352 + }, + { + "epoch": 0.5625864661654135, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 14353 + }, + { + "epoch": 0.5626466165413534, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 14354 + }, + { + "epoch": 0.5627067669172933, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 14355 + }, + { + "epoch": 0.5627669172932331, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 14356 + }, + { + "epoch": 0.5628270676691729, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2094, + "step": 14357 + }, + { + "epoch": 0.5628872180451128, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1436, + "step": 14358 + }, + { + "epoch": 0.5629473684210526, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1907, + "step": 14359 + }, + { + "epoch": 0.5630075187969925, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 14360 + }, + { + "epoch": 0.5630676691729324, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 14361 + }, + { + "epoch": 0.5631278195488721, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2127, + "step": 14362 + }, + { + "epoch": 0.563187969924812, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 14363 + }, + { + "epoch": 0.5632481203007519, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 14364 + }, + { + "epoch": 0.5633082706766918, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 14365 + }, + { + "epoch": 0.5633684210526316, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 14366 + }, + { + "epoch": 0.5634285714285714, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 14367 + }, + { + "epoch": 0.5634887218045113, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1582, + "step": 14368 + }, + { + "epoch": 0.5635488721804511, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 14369 + }, + { + "epoch": 0.563609022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1839, + "step": 14370 + }, + { + "epoch": 0.5636691729323309, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1688, + "step": 14371 + }, + { + "epoch": 0.5637293233082706, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 14372 + }, + { + "epoch": 0.5637894736842105, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 14373 + }, + { + "epoch": 0.5638496240601504, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 14374 + }, + { + "epoch": 0.5639097744360902, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 14375 + }, + { + "epoch": 0.5639699248120301, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 14376 + }, + { + "epoch": 0.5640300751879699, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 14377 + }, + { + "epoch": 0.5640902255639098, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 14378 + }, + { + "epoch": 0.5641503759398496, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2738, + "step": 14379 + }, + { + "epoch": 0.5642105263157895, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1949, + "step": 14380 + }, + { + "epoch": 0.5642706766917294, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1852, + "step": 14381 + }, + { + "epoch": 0.5643308270676691, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 14382 + }, + { + "epoch": 0.564390977443609, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 14383 + }, + { + "epoch": 0.5644511278195489, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 14384 + }, + { + "epoch": 0.5645112781954887, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2209, + "step": 14385 + }, + { + "epoch": 0.5645714285714286, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 14386 + }, + { + "epoch": 0.5646315789473684, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 14387 + }, + { + "epoch": 0.5646917293233082, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 14388 + }, + { + "epoch": 0.5647518796992481, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2155, + "step": 14389 + }, + { + "epoch": 0.564812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 14390 + }, + { + "epoch": 0.5648721804511279, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 14391 + }, + { + "epoch": 0.5649323308270676, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2582, + "step": 14392 + }, + { + "epoch": 0.5649924812030075, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 14393 + }, + { + "epoch": 0.5650526315789474, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 14394 + }, + { + "epoch": 0.5651127819548872, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2088, + "step": 14395 + }, + { + "epoch": 0.5651729323308271, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 14396 + }, + { + "epoch": 0.5652330827067669, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1651, + "step": 14397 + }, + { + "epoch": 0.5652932330827067, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 14398 + }, + { + "epoch": 0.5653533834586466, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 14399 + }, + { + "epoch": 0.5654135338345865, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1774, + "step": 14400 + }, + { + "epoch": 0.5654736842105264, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 14401 + }, + { + "epoch": 0.5655338345864661, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 14402 + }, + { + "epoch": 0.565593984962406, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 14403 + }, + { + "epoch": 0.5656541353383459, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 14404 + }, + { + "epoch": 0.5657142857142857, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 14405 + }, + { + "epoch": 0.5657744360902256, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 14406 + }, + { + "epoch": 0.5658345864661655, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 14407 + }, + { + "epoch": 0.5658947368421052, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2196, + "step": 14408 + }, + { + "epoch": 0.5659548872180451, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 14409 + }, + { + "epoch": 0.566015037593985, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 14410 + }, + { + "epoch": 0.5660751879699248, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 14411 + }, + { + "epoch": 0.5661353383458647, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 14412 + }, + { + "epoch": 0.5661954887218045, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 14413 + }, + { + "epoch": 0.5662556390977443, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 14414 + }, + { + "epoch": 0.5663157894736842, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 14415 + }, + { + "epoch": 0.5663759398496241, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2208, + "step": 14416 + }, + { + "epoch": 0.566436090225564, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 14417 + }, + { + "epoch": 0.5664962406015037, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 14418 + }, + { + "epoch": 0.5665563909774436, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 14419 + }, + { + "epoch": 0.5666165413533835, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 14420 + }, + { + "epoch": 0.5666766917293233, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 14421 + }, + { + "epoch": 0.5667368421052632, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 14422 + }, + { + "epoch": 0.566796992481203, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 14423 + }, + { + "epoch": 0.5668571428571428, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 14424 + }, + { + "epoch": 0.5669172932330827, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.1326, + "step": 14425 + }, + { + "epoch": 0.5669774436090226, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 14426 + }, + { + "epoch": 0.5670375939849625, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 14427 + }, + { + "epoch": 0.5670977443609022, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 14428 + }, + { + "epoch": 0.5671578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 14429 + }, + { + "epoch": 0.567218045112782, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 14430 + }, + { + "epoch": 0.5672781954887218, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 14431 + }, + { + "epoch": 0.5673383458646617, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 14432 + }, + { + "epoch": 0.5673984962406015, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 14433 + }, + { + "epoch": 0.5674586466165413, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1567, + "step": 14434 + }, + { + "epoch": 0.5675187969924812, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 14435 + }, + { + "epoch": 0.5675789473684211, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2259, + "step": 14436 + }, + { + "epoch": 0.5676390977443609, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 14437 + }, + { + "epoch": 0.5676992481203007, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 14438 + }, + { + "epoch": 0.5677593984962406, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 14439 + }, + { + "epoch": 0.5678195488721804, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 14440 + }, + { + "epoch": 0.5678796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 14441 + }, + { + "epoch": 0.5679398496240602, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 14442 + }, + { + "epoch": 0.568, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 14443 + }, + { + "epoch": 0.5680601503759398, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 14444 + }, + { + "epoch": 0.5681203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 14445 + }, + { + "epoch": 0.5681804511278196, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2157, + "step": 14446 + }, + { + "epoch": 0.5682406015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 14447 + }, + { + "epoch": 0.5683007518796992, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 14448 + }, + { + "epoch": 0.5683609022556391, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2343, + "step": 14449 + }, + { + "epoch": 0.5684210526315789, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 14450 + }, + { + "epoch": 0.5684812030075188, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 14451 + }, + { + "epoch": 0.5685413533834587, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 14452 + }, + { + "epoch": 0.5686015037593984, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 14453 + }, + { + "epoch": 0.5686616541353383, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 14454 + }, + { + "epoch": 0.5687218045112782, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 14455 + }, + { + "epoch": 0.568781954887218, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 14456 + }, + { + "epoch": 0.5688421052631579, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 14457 + }, + { + "epoch": 0.5689022556390978, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2968, + "step": 14458 + }, + { + "epoch": 0.5689624060150376, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 14459 + }, + { + "epoch": 0.5690225563909774, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 14460 + }, + { + "epoch": 0.5690827067669173, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 14461 + }, + { + "epoch": 0.5691428571428572, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 14462 + }, + { + "epoch": 0.569203007518797, + "grad_norm": 1.4765625, + "learning_rate": 3e-05, + "loss": 1.1837, + "step": 14463 + }, + { + "epoch": 0.5692631578947368, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 14464 + }, + { + "epoch": 0.5693233082706767, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 14465 + }, + { + "epoch": 0.5693834586466165, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 14466 + }, + { + "epoch": 0.5694436090225564, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2053, + "step": 14467 + }, + { + "epoch": 0.5695037593984963, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 14468 + }, + { + "epoch": 0.569563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 14469 + }, + { + "epoch": 0.5696240601503759, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 14470 + }, + { + "epoch": 0.5696842105263158, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2467, + "step": 14471 + }, + { + "epoch": 0.5697443609022557, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 14472 + }, + { + "epoch": 0.5698045112781955, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1357, + "step": 14473 + }, + { + "epoch": 0.5698646616541353, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 14474 + }, + { + "epoch": 0.5699248120300752, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 14475 + }, + { + "epoch": 0.569984962406015, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 14476 + }, + { + "epoch": 0.5700451127819549, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.3286, + "step": 14477 + }, + { + "epoch": 0.5701052631578948, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 14478 + }, + { + "epoch": 0.5701654135338345, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 14479 + }, + { + "epoch": 0.5702255639097744, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 14480 + }, + { + "epoch": 0.5702857142857143, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 14481 + }, + { + "epoch": 0.5703458646616542, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 14482 + }, + { + "epoch": 0.570406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 14483 + }, + { + "epoch": 0.5704661654135338, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 14484 + }, + { + "epoch": 0.5705263157894737, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 14485 + }, + { + "epoch": 0.5705864661654135, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 14486 + }, + { + "epoch": 0.5706466165413534, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 14487 + }, + { + "epoch": 0.5707067669172933, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 14488 + }, + { + "epoch": 0.570766917293233, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 14489 + }, + { + "epoch": 0.5708270676691729, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1791, + "step": 14490 + }, + { + "epoch": 0.5708872180451128, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 14491 + }, + { + "epoch": 0.5709473684210526, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 14492 + }, + { + "epoch": 0.5710075187969925, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 14493 + }, + { + "epoch": 0.5710676691729323, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 14494 + }, + { + "epoch": 0.5711278195488722, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 14495 + }, + { + "epoch": 0.571187969924812, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2017, + "step": 14496 + }, + { + "epoch": 0.5712481203007519, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 14497 + }, + { + "epoch": 0.5713082706766918, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 14498 + }, + { + "epoch": 0.5713684210526315, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1632, + "step": 14499 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 14500 + }, + { + "epoch": 0.5714887218045113, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 14501 + }, + { + "epoch": 0.5715488721804511, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 14502 + }, + { + "epoch": 0.571609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 14503 + }, + { + "epoch": 0.5716691729323309, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.3484, + "step": 14504 + }, + { + "epoch": 0.5717293233082706, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 14505 + }, + { + "epoch": 0.5717894736842105, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2225, + "step": 14506 + }, + { + "epoch": 0.5718496240601504, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 14507 + }, + { + "epoch": 0.5719097744360903, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 14508 + }, + { + "epoch": 0.5719699248120301, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2127, + "step": 14509 + }, + { + "epoch": 0.5720300751879699, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 14510 + }, + { + "epoch": 0.5720902255639098, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2128, + "step": 14511 + }, + { + "epoch": 0.5721503759398496, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 14512 + }, + { + "epoch": 0.5722105263157895, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 14513 + }, + { + "epoch": 0.5722706766917294, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 14514 + }, + { + "epoch": 0.5723308270676691, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 14515 + }, + { + "epoch": 0.572390977443609, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 14516 + }, + { + "epoch": 0.5724511278195489, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 14517 + }, + { + "epoch": 0.5725112781954887, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 14518 + }, + { + "epoch": 0.5725714285714286, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2267, + "step": 14519 + }, + { + "epoch": 0.5726315789473684, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 14520 + }, + { + "epoch": 0.5726917293233083, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 14521 + }, + { + "epoch": 0.5727518796992481, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 14522 + }, + { + "epoch": 0.572812030075188, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2164, + "step": 14523 + }, + { + "epoch": 0.5728721804511279, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 14524 + }, + { + "epoch": 0.5729323308270676, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 14525 + }, + { + "epoch": 0.5729924812030075, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 14526 + }, + { + "epoch": 0.5730526315789474, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 14527 + }, + { + "epoch": 0.5731127819548872, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 14528 + }, + { + "epoch": 0.5731729323308271, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 14529 + }, + { + "epoch": 0.5732330827067669, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 14530 + }, + { + "epoch": 0.5732932330827067, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 14531 + }, + { + "epoch": 0.5733533834586466, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.224, + "step": 14532 + }, + { + "epoch": 0.5734135338345865, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2133, + "step": 14533 + }, + { + "epoch": 0.5734736842105264, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 14534 + }, + { + "epoch": 0.5735338345864661, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 14535 + }, + { + "epoch": 0.573593984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2174, + "step": 14536 + }, + { + "epoch": 0.5736541353383459, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 14537 + }, + { + "epoch": 0.5737142857142857, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2515, + "step": 14538 + }, + { + "epoch": 0.5737744360902256, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 14539 + }, + { + "epoch": 0.5738345864661654, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 14540 + }, + { + "epoch": 0.5738947368421052, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 14541 + }, + { + "epoch": 0.5739548872180451, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 14542 + }, + { + "epoch": 0.574015037593985, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 14543 + }, + { + "epoch": 0.5740751879699248, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 14544 + }, + { + "epoch": 0.5741353383458646, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 14545 + }, + { + "epoch": 0.5741954887218045, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 14546 + }, + { + "epoch": 0.5742556390977444, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 14547 + }, + { + "epoch": 0.5743157894736842, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 14548 + }, + { + "epoch": 0.5743759398496241, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 14549 + }, + { + "epoch": 0.5744360902255639, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1597, + "step": 14550 + }, + { + "epoch": 0.5744962406015037, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 14551 + }, + { + "epoch": 0.5745563909774436, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1894, + "step": 14552 + }, + { + "epoch": 0.5746165413533835, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2211, + "step": 14553 + }, + { + "epoch": 0.5746766917293233, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 14554 + }, + { + "epoch": 0.5747368421052632, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 14555 + }, + { + "epoch": 0.574796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 14556 + }, + { + "epoch": 0.5748571428571428, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 14557 + }, + { + "epoch": 0.5749172932330827, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2332, + "step": 14558 + }, + { + "epoch": 0.5749774436090226, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 14559 + }, + { + "epoch": 0.5750375939849625, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1951, + "step": 14560 + }, + { + "epoch": 0.5750977443609022, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1658, + "step": 14561 + }, + { + "epoch": 0.5751578947368421, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 14562 + }, + { + "epoch": 0.575218045112782, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 14563 + }, + { + "epoch": 0.5752781954887218, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 14564 + }, + { + "epoch": 0.5753383458646617, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2548, + "step": 14565 + }, + { + "epoch": 0.5753984962406015, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 14566 + }, + { + "epoch": 0.5754586466165413, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 14567 + }, + { + "epoch": 0.5755187969924812, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 14568 + }, + { + "epoch": 0.5755789473684211, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1905, + "step": 14569 + }, + { + "epoch": 0.575639097744361, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.2256, + "step": 14570 + }, + { + "epoch": 0.5756992481203007, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2712, + "step": 14571 + }, + { + "epoch": 0.5757593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 14572 + }, + { + "epoch": 0.5758195488721805, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 14573 + }, + { + "epoch": 0.5758796992481203, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2053, + "step": 14574 + }, + { + "epoch": 0.5759398496240602, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1333, + "step": 14575 + }, + { + "epoch": 0.576, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.2037, + "step": 14576 + }, + { + "epoch": 0.5760601503759398, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2149, + "step": 14577 + }, + { + "epoch": 0.5761203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1537, + "step": 14578 + }, + { + "epoch": 0.5761804511278196, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1254, + "step": 14579 + }, + { + "epoch": 0.5762406015037594, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2307, + "step": 14580 + }, + { + "epoch": 0.5763007518796992, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2039, + "step": 14581 + }, + { + "epoch": 0.5763609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 14582 + }, + { + "epoch": 0.576421052631579, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 14583 + }, + { + "epoch": 0.5764812030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 14584 + }, + { + "epoch": 0.5765413533834587, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 14585 + }, + { + "epoch": 0.5766015037593984, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2072, + "step": 14586 + }, + { + "epoch": 0.5766616541353383, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.3906, + "step": 14587 + }, + { + "epoch": 0.5767218045112782, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 14588 + }, + { + "epoch": 0.5767819548872181, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1804, + "step": 14589 + }, + { + "epoch": 0.5768421052631579, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 14590 + }, + { + "epoch": 0.5769022556390977, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 14591 + }, + { + "epoch": 0.5769624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 14592 + }, + { + "epoch": 0.5770225563909774, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 14593 + }, + { + "epoch": 0.5770827067669173, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 14594 + }, + { + "epoch": 0.5771428571428572, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 14595 + }, + { + "epoch": 0.5772030075187969, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 14596 + }, + { + "epoch": 0.5772631578947368, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.3216, + "step": 14597 + }, + { + "epoch": 0.5773233082706767, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 14598 + }, + { + "epoch": 0.5773834586466166, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 14599 + }, + { + "epoch": 0.5774436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 14600 + }, + { + "epoch": 0.5775037593984962, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 14601 + }, + { + "epoch": 0.5775639097744361, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1811, + "step": 14602 + }, + { + "epoch": 0.5776240601503759, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 14603 + }, + { + "epoch": 0.5776842105263158, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 14604 + }, + { + "epoch": 0.5777443609022557, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1872, + "step": 14605 + }, + { + "epoch": 0.5778045112781955, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 14606 + }, + { + "epoch": 0.5778646616541353, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 14607 + }, + { + "epoch": 0.5779248120300752, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 14608 + }, + { + "epoch": 0.577984962406015, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2615, + "step": 14609 + }, + { + "epoch": 0.5780451127819549, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1643, + "step": 14610 + }, + { + "epoch": 0.5781052631578948, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 14611 + }, + { + "epoch": 0.5781654135338345, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 14612 + }, + { + "epoch": 0.5782255639097744, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 14613 + }, + { + "epoch": 0.5782857142857143, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.3552, + "step": 14614 + }, + { + "epoch": 0.5783458646616542, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1357, + "step": 14615 + }, + { + "epoch": 0.578406015037594, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 14616 + }, + { + "epoch": 0.5784661654135338, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 14617 + }, + { + "epoch": 0.5785263157894737, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1718, + "step": 14618 + }, + { + "epoch": 0.5785864661654135, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 14619 + }, + { + "epoch": 0.5786466165413534, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 14620 + }, + { + "epoch": 0.5787067669172933, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2017, + "step": 14621 + }, + { + "epoch": 0.578766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 14622 + }, + { + "epoch": 0.5788270676691729, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2398, + "step": 14623 + }, + { + "epoch": 0.5788872180451128, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 14624 + }, + { + "epoch": 0.5789473684210527, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 14625 + }, + { + "epoch": 0.5790075187969925, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 14626 + }, + { + "epoch": 0.5790676691729323, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2362, + "step": 14627 + }, + { + "epoch": 0.5791278195488722, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 14628 + }, + { + "epoch": 0.579187969924812, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2213, + "step": 14629 + }, + { + "epoch": 0.5792481203007519, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2367, + "step": 14630 + }, + { + "epoch": 0.5793082706766918, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 14631 + }, + { + "epoch": 0.5793684210526315, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 14632 + }, + { + "epoch": 0.5794285714285714, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 14633 + }, + { + "epoch": 0.5794887218045113, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 14634 + }, + { + "epoch": 0.5795488721804511, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 14635 + }, + { + "epoch": 0.579609022556391, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 14636 + }, + { + "epoch": 0.5796691729323308, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 14637 + }, + { + "epoch": 0.5797293233082706, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 14638 + }, + { + "epoch": 0.5797894736842105, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 14639 + }, + { + "epoch": 0.5798496240601504, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2171, + "step": 14640 + }, + { + "epoch": 0.5799097744360903, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 14641 + }, + { + "epoch": 0.57996992481203, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 14642 + }, + { + "epoch": 0.5800300751879699, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3941, + "step": 14643 + }, + { + "epoch": 0.5800902255639098, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 14644 + }, + { + "epoch": 0.5801503759398496, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 14645 + }, + { + "epoch": 0.5802105263157895, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 14646 + }, + { + "epoch": 0.5802706766917293, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 14647 + }, + { + "epoch": 0.5803308270676691, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2, + "step": 14648 + }, + { + "epoch": 0.580390977443609, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 14649 + }, + { + "epoch": 0.5804511278195489, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2024, + "step": 14650 + }, + { + "epoch": 0.5805112781954888, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 14651 + }, + { + "epoch": 0.5805714285714285, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 14652 + }, + { + "epoch": 0.5806315789473684, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1636, + "step": 14653 + }, + { + "epoch": 0.5806917293233083, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 14654 + }, + { + "epoch": 0.5807518796992481, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 14655 + }, + { + "epoch": 0.580812030075188, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2701, + "step": 14656 + }, + { + "epoch": 0.5808721804511279, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 14657 + }, + { + "epoch": 0.5809323308270676, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 14658 + }, + { + "epoch": 0.5809924812030075, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 14659 + }, + { + "epoch": 0.5810526315789474, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1725, + "step": 14660 + }, + { + "epoch": 0.5811127819548872, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.163, + "step": 14661 + }, + { + "epoch": 0.5811729323308271, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 14662 + }, + { + "epoch": 0.5812330827067669, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 14663 + }, + { + "epoch": 0.5812932330827068, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 14664 + }, + { + "epoch": 0.5813533834586466, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 14665 + }, + { + "epoch": 0.5814135338345865, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 14666 + }, + { + "epoch": 0.5814736842105264, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 14667 + }, + { + "epoch": 0.5815338345864661, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2108, + "step": 14668 + }, + { + "epoch": 0.581593984962406, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 14669 + }, + { + "epoch": 0.5816541353383459, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3159, + "step": 14670 + }, + { + "epoch": 0.5817142857142857, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 14671 + }, + { + "epoch": 0.5817744360902256, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1833, + "step": 14672 + }, + { + "epoch": 0.5818345864661654, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 14673 + }, + { + "epoch": 0.5818947368421052, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 14674 + }, + { + "epoch": 0.5819548872180451, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 14675 + }, + { + "epoch": 0.582015037593985, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 14676 + }, + { + "epoch": 0.5820751879699249, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 14677 + }, + { + "epoch": 0.5821353383458646, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 14678 + }, + { + "epoch": 0.5821954887218045, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1723, + "step": 14679 + }, + { + "epoch": 0.5822556390977444, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 14680 + }, + { + "epoch": 0.5823157894736842, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 14681 + }, + { + "epoch": 0.5823759398496241, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 14682 + }, + { + "epoch": 0.5824360902255639, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 14683 + }, + { + "epoch": 0.5824962406015037, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 14684 + }, + { + "epoch": 0.5825563909774436, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 14685 + }, + { + "epoch": 0.5826165413533835, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 14686 + }, + { + "epoch": 0.5826766917293233, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 14687 + }, + { + "epoch": 0.5827368421052631, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1326, + "step": 14688 + }, + { + "epoch": 0.582796992481203, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 14689 + }, + { + "epoch": 0.5828571428571429, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1994, + "step": 14690 + }, + { + "epoch": 0.5829172932330827, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 14691 + }, + { + "epoch": 0.5829774436090226, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 14692 + }, + { + "epoch": 0.5830375939849624, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 14693 + }, + { + "epoch": 0.5830977443609022, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 14694 + }, + { + "epoch": 0.5831578947368421, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 14695 + }, + { + "epoch": 0.583218045112782, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1933, + "step": 14696 + }, + { + "epoch": 0.5832781954887218, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 14697 + }, + { + "epoch": 0.5833383458646616, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 14698 + }, + { + "epoch": 0.5833984962406015, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 14699 + }, + { + "epoch": 0.5834586466165413, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2875, + "step": 14700 + }, + { + "epoch": 0.5835187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 14701 + }, + { + "epoch": 0.5835789473684211, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 14702 + }, + { + "epoch": 0.583639097744361, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 14703 + }, + { + "epoch": 0.5836992481203007, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.235, + "step": 14704 + }, + { + "epoch": 0.5837593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 14705 + }, + { + "epoch": 0.5838195488721805, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 14706 + }, + { + "epoch": 0.5838796992481203, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 14707 + }, + { + "epoch": 0.5839398496240602, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1725, + "step": 14708 + }, + { + "epoch": 0.584, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 14709 + }, + { + "epoch": 0.5840601503759398, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 14710 + }, + { + "epoch": 0.5841203007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 14711 + }, + { + "epoch": 0.5841804511278196, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 14712 + }, + { + "epoch": 0.5842406015037594, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 14713 + }, + { + "epoch": 0.5843007518796992, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 14714 + }, + { + "epoch": 0.5843609022556391, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 14715 + }, + { + "epoch": 0.584421052631579, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 14716 + }, + { + "epoch": 0.5844812030075188, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 14717 + }, + { + "epoch": 0.5845413533834587, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 14718 + }, + { + "epoch": 0.5846015037593985, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 14719 + }, + { + "epoch": 0.5846616541353383, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 14720 + }, + { + "epoch": 0.5847218045112782, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 14721 + }, + { + "epoch": 0.5847819548872181, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 14722 + }, + { + "epoch": 0.5848421052631579, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2823, + "step": 14723 + }, + { + "epoch": 0.5849022556390977, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.4256, + "step": 14724 + }, + { + "epoch": 0.5849624060150376, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 14725 + }, + { + "epoch": 0.5850225563909774, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 14726 + }, + { + "epoch": 0.5850827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 14727 + }, + { + "epoch": 0.5851428571428572, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 14728 + }, + { + "epoch": 0.585203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2662, + "step": 14729 + }, + { + "epoch": 0.5852631578947368, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 14730 + }, + { + "epoch": 0.5853233082706767, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 14731 + }, + { + "epoch": 0.5853834586466166, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 14732 + }, + { + "epoch": 0.5854436090225564, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3061, + "step": 14733 + }, + { + "epoch": 0.5855037593984962, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1526, + "step": 14734 + }, + { + "epoch": 0.5855639097744361, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 14735 + }, + { + "epoch": 0.5856240601503759, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 14736 + }, + { + "epoch": 0.5856842105263158, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 14737 + }, + { + "epoch": 0.5857443609022557, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1643, + "step": 14738 + }, + { + "epoch": 0.5858045112781954, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 14739 + }, + { + "epoch": 0.5858646616541353, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 14740 + }, + { + "epoch": 0.5859248120300752, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 14741 + }, + { + "epoch": 0.585984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 14742 + }, + { + "epoch": 0.5860451127819549, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 14743 + }, + { + "epoch": 0.5861052631578947, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 14744 + }, + { + "epoch": 0.5861654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.187, + "step": 14745 + }, + { + "epoch": 0.5862255639097744, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 14746 + }, + { + "epoch": 0.5862857142857143, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1757, + "step": 14747 + }, + { + "epoch": 0.5863458646616542, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 14748 + }, + { + "epoch": 0.5864060150375939, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2743, + "step": 14749 + }, + { + "epoch": 0.5864661654135338, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2299, + "step": 14750 + }, + { + "epoch": 0.5865263157894737, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 14751 + }, + { + "epoch": 0.5865864661654135, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 14752 + }, + { + "epoch": 0.5866466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2662, + "step": 14753 + }, + { + "epoch": 0.5867067669172933, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 14754 + }, + { + "epoch": 0.586766917293233, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 14755 + }, + { + "epoch": 0.5868270676691729, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 14756 + }, + { + "epoch": 0.5868872180451128, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2205, + "step": 14757 + }, + { + "epoch": 0.5869473684210527, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 14758 + }, + { + "epoch": 0.5870075187969925, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 14759 + }, + { + "epoch": 0.5870676691729323, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1832, + "step": 14760 + }, + { + "epoch": 0.5871278195488722, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 14761 + }, + { + "epoch": 0.587187969924812, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 14762 + }, + { + "epoch": 0.5872481203007519, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 14763 + }, + { + "epoch": 0.5873082706766918, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 14764 + }, + { + "epoch": 0.5873684210526315, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 14765 + }, + { + "epoch": 0.5874285714285714, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 14766 + }, + { + "epoch": 0.5874887218045113, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2067, + "step": 14767 + }, + { + "epoch": 0.5875488721804512, + "grad_norm": 4.40625, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 14768 + }, + { + "epoch": 0.587609022556391, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 14769 + }, + { + "epoch": 0.5876691729323308, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 14770 + }, + { + "epoch": 0.5877293233082707, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1707, + "step": 14771 + }, + { + "epoch": 0.5877894736842105, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 14772 + }, + { + "epoch": 0.5878496240601504, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 14773 + }, + { + "epoch": 0.5879097744360903, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 14774 + }, + { + "epoch": 0.58796992481203, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 14775 + }, + { + "epoch": 0.5880300751879699, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.3551, + "step": 14776 + }, + { + "epoch": 0.5880902255639098, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 14777 + }, + { + "epoch": 0.5881503759398496, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 14778 + }, + { + "epoch": 0.5882105263157895, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 14779 + }, + { + "epoch": 0.5882706766917293, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2405, + "step": 14780 + }, + { + "epoch": 0.5883308270676691, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 14781 + }, + { + "epoch": 0.588390977443609, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 14782 + }, + { + "epoch": 0.5884511278195489, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 14783 + }, + { + "epoch": 0.5885112781954888, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.3486, + "step": 14784 + }, + { + "epoch": 0.5885714285714285, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 14785 + }, + { + "epoch": 0.5886315789473684, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 14786 + }, + { + "epoch": 0.5886917293233083, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 14787 + }, + { + "epoch": 0.5887518796992481, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 14788 + }, + { + "epoch": 0.588812030075188, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2337, + "step": 14789 + }, + { + "epoch": 0.5888721804511278, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 14790 + }, + { + "epoch": 0.5889323308270676, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 14791 + }, + { + "epoch": 0.5889924812030075, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2991, + "step": 14792 + }, + { + "epoch": 0.5890526315789474, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 14793 + }, + { + "epoch": 0.5891127819548873, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 14794 + }, + { + "epoch": 0.589172932330827, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 14795 + }, + { + "epoch": 0.5892330827067669, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 14796 + }, + { + "epoch": 0.5892932330827068, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 14797 + }, + { + "epoch": 0.5893533834586466, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 14798 + }, + { + "epoch": 0.5894135338345865, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 14799 + }, + { + "epoch": 0.5894736842105263, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 14800 + }, + { + "epoch": 0.5895338345864661, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 14801 + }, + { + "epoch": 0.589593984962406, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 14802 + }, + { + "epoch": 0.5896541353383459, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.4133, + "step": 14803 + }, + { + "epoch": 0.5897142857142857, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 14804 + }, + { + "epoch": 0.5897744360902256, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2469, + "step": 14805 + }, + { + "epoch": 0.5898345864661654, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 14806 + }, + { + "epoch": 0.5898947368421052, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 14807 + }, + { + "epoch": 0.5899548872180451, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 14808 + }, + { + "epoch": 0.590015037593985, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 14809 + }, + { + "epoch": 0.5900751879699249, + "grad_norm": 1.8671875, + "learning_rate": 3e-05, + "loss": 1.2321, + "step": 14810 + }, + { + "epoch": 0.5901353383458646, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 14811 + }, + { + "epoch": 0.5901954887218045, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 14812 + }, + { + "epoch": 0.5902556390977444, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 14813 + }, + { + "epoch": 0.5903157894736842, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 14814 + }, + { + "epoch": 0.5903759398496241, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 14815 + }, + { + "epoch": 0.5904360902255639, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 14816 + }, + { + "epoch": 0.5904962406015037, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 14817 + }, + { + "epoch": 0.5905563909774436, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 14818 + }, + { + "epoch": 0.5906165413533835, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 14819 + }, + { + "epoch": 0.5906766917293234, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 14820 + }, + { + "epoch": 0.5907368421052631, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 14821 + }, + { + "epoch": 0.590796992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.211, + "step": 14822 + }, + { + "epoch": 0.5908571428571429, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 14823 + }, + { + "epoch": 0.5909172932330827, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 14824 + }, + { + "epoch": 0.5909774436090226, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 14825 + }, + { + "epoch": 0.5910375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2079, + "step": 14826 + }, + { + "epoch": 0.5910977443609022, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 14827 + }, + { + "epoch": 0.5911578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 14828 + }, + { + "epoch": 0.591218045112782, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1972, + "step": 14829 + }, + { + "epoch": 0.5912781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.3719, + "step": 14830 + }, + { + "epoch": 0.5913383458646616, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 14831 + }, + { + "epoch": 0.5913984962406015, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 14832 + }, + { + "epoch": 0.5914586466165414, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 14833 + }, + { + "epoch": 0.5915187969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 14834 + }, + { + "epoch": 0.5915789473684211, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 14835 + }, + { + "epoch": 0.5916390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 14836 + }, + { + "epoch": 0.5916992481203007, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 14837 + }, + { + "epoch": 0.5917593984962406, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1996, + "step": 14838 + }, + { + "epoch": 0.5918195488721805, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 14839 + }, + { + "epoch": 0.5918796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1496, + "step": 14840 + }, + { + "epoch": 0.5919398496240601, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 14841 + }, + { + "epoch": 0.592, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 14842 + }, + { + "epoch": 0.5920601503759398, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 14843 + }, + { + "epoch": 0.5921203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.19, + "step": 14844 + }, + { + "epoch": 0.5921804511278196, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 14845 + }, + { + "epoch": 0.5922406015037593, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1314, + "step": 14846 + }, + { + "epoch": 0.5923007518796992, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 14847 + }, + { + "epoch": 0.5923609022556391, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2042, + "step": 14848 + }, + { + "epoch": 0.592421052631579, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 14849 + }, + { + "epoch": 0.5924812030075188, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 14850 + }, + { + "epoch": 0.5925413533834587, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 14851 + }, + { + "epoch": 0.5926015037593985, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2452, + "step": 14852 + }, + { + "epoch": 0.5926616541353383, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2444, + "step": 14853 + }, + { + "epoch": 0.5927218045112782, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 14854 + }, + { + "epoch": 0.5927819548872181, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 14855 + }, + { + "epoch": 0.592842105263158, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 14856 + }, + { + "epoch": 0.5929022556390977, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 14857 + }, + { + "epoch": 0.5929624060150376, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 14858 + }, + { + "epoch": 0.5930225563909775, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 14859 + }, + { + "epoch": 0.5930827067669173, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2181, + "step": 14860 + }, + { + "epoch": 0.5931428571428572, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 14861 + }, + { + "epoch": 0.593203007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2147, + "step": 14862 + }, + { + "epoch": 0.5932631578947368, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 14863 + }, + { + "epoch": 0.5933233082706767, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2281, + "step": 14864 + }, + { + "epoch": 0.5933834586466166, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 14865 + }, + { + "epoch": 0.5934436090225564, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 14866 + }, + { + "epoch": 0.5935037593984962, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 14867 + }, + { + "epoch": 0.5935639097744361, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 14868 + }, + { + "epoch": 0.5936240601503759, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 14869 + }, + { + "epoch": 0.5936842105263158, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 14870 + }, + { + "epoch": 0.5937443609022557, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 14871 + }, + { + "epoch": 0.5938045112781954, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 14872 + }, + { + "epoch": 0.5938646616541353, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 14873 + }, + { + "epoch": 0.5939248120300752, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 14874 + }, + { + "epoch": 0.5939849624060151, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 14875 + }, + { + "epoch": 0.5940451127819549, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 14876 + }, + { + "epoch": 0.5941052631578947, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 14877 + }, + { + "epoch": 0.5941654135338346, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 14878 + }, + { + "epoch": 0.5942255639097744, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1776, + "step": 14879 + }, + { + "epoch": 0.5942857142857143, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 14880 + }, + { + "epoch": 0.5943458646616542, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2841, + "step": 14881 + }, + { + "epoch": 0.5944060150375939, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 14882 + }, + { + "epoch": 0.5944661654135338, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 14883 + }, + { + "epoch": 0.5945263157894737, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 14884 + }, + { + "epoch": 0.5945864661654136, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 14885 + }, + { + "epoch": 0.5946466165413534, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 14886 + }, + { + "epoch": 0.5947067669172932, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 14887 + }, + { + "epoch": 0.594766917293233, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 14888 + }, + { + "epoch": 0.5948270676691729, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 14889 + }, + { + "epoch": 0.5948872180451128, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 14890 + }, + { + "epoch": 0.5949473684210527, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2505, + "step": 14891 + }, + { + "epoch": 0.5950075187969924, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1529, + "step": 14892 + }, + { + "epoch": 0.5950676691729323, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 14893 + }, + { + "epoch": 0.5951278195488722, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 14894 + }, + { + "epoch": 0.595187969924812, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 14895 + }, + { + "epoch": 0.5952481203007519, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 14896 + }, + { + "epoch": 0.5953082706766917, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 14897 + }, + { + "epoch": 0.5953684210526315, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 14898 + }, + { + "epoch": 0.5954285714285714, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 14899 + }, + { + "epoch": 0.5954887218045113, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 14900 + }, + { + "epoch": 0.5955488721804512, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 14901 + }, + { + "epoch": 0.595609022556391, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 14902 + }, + { + "epoch": 0.5956691729323308, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 14903 + }, + { + "epoch": 0.5957293233082707, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 14904 + }, + { + "epoch": 0.5957894736842105, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.235, + "step": 14905 + }, + { + "epoch": 0.5958496240601504, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 14906 + }, + { + "epoch": 0.5959097744360903, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 14907 + }, + { + "epoch": 0.59596992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1708, + "step": 14908 + }, + { + "epoch": 0.5960300751879699, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 14909 + }, + { + "epoch": 0.5960902255639098, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 14910 + }, + { + "epoch": 0.5961503759398497, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 14911 + }, + { + "epoch": 0.5962105263157895, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 14912 + }, + { + "epoch": 0.5962706766917293, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 14913 + }, + { + "epoch": 0.5963308270676692, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2654, + "step": 14914 + }, + { + "epoch": 0.596390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1933, + "step": 14915 + }, + { + "epoch": 0.5964511278195489, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 14916 + }, + { + "epoch": 0.5965112781954888, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3116, + "step": 14917 + }, + { + "epoch": 0.5965714285714285, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 14918 + }, + { + "epoch": 0.5966315789473684, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 14919 + }, + { + "epoch": 0.5966917293233083, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2138, + "step": 14920 + }, + { + "epoch": 0.5967518796992481, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1732, + "step": 14921 + }, + { + "epoch": 0.596812030075188, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2305, + "step": 14922 + }, + { + "epoch": 0.5968721804511278, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2058, + "step": 14923 + }, + { + "epoch": 0.5969323308270676, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 14924 + }, + { + "epoch": 0.5969924812030075, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 14925 + }, + { + "epoch": 0.5970526315789474, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 14926 + }, + { + "epoch": 0.5971127819548873, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2333, + "step": 14927 + }, + { + "epoch": 0.597172932330827, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 14928 + }, + { + "epoch": 0.5972330827067669, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1965, + "step": 14929 + }, + { + "epoch": 0.5972932330827068, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 14930 + }, + { + "epoch": 0.5973533834586466, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 14931 + }, + { + "epoch": 0.5974135338345865, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 14932 + }, + { + "epoch": 0.5974736842105263, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 14933 + }, + { + "epoch": 0.5975338345864661, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 14934 + }, + { + "epoch": 0.597593984962406, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 14935 + }, + { + "epoch": 0.5976541353383459, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2732, + "step": 14936 + }, + { + "epoch": 0.5977142857142858, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 14937 + }, + { + "epoch": 0.5977744360902255, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 14938 + }, + { + "epoch": 0.5978345864661654, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2283, + "step": 14939 + }, + { + "epoch": 0.5978947368421053, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 14940 + }, + { + "epoch": 0.5979548872180451, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 14941 + }, + { + "epoch": 0.598015037593985, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 14942 + }, + { + "epoch": 0.5980751879699248, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1429, + "step": 14943 + }, + { + "epoch": 0.5981353383458646, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 14944 + }, + { + "epoch": 0.5981954887218045, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 14945 + }, + { + "epoch": 0.5982556390977444, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.226, + "step": 14946 + }, + { + "epoch": 0.5983157894736842, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2243, + "step": 14947 + }, + { + "epoch": 0.598375939849624, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 14948 + }, + { + "epoch": 0.5984360902255639, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 14949 + }, + { + "epoch": 0.5984962406015037, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 14950 + }, + { + "epoch": 0.5985563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 14951 + }, + { + "epoch": 0.5986165413533835, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 14952 + }, + { + "epoch": 0.5986766917293234, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1832, + "step": 14953 + }, + { + "epoch": 0.5987368421052631, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 14954 + }, + { + "epoch": 0.598796992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 14955 + }, + { + "epoch": 0.5988571428571429, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 14956 + }, + { + "epoch": 0.5989172932330827, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 14957 + }, + { + "epoch": 0.5989774436090226, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 14958 + }, + { + "epoch": 0.5990375939849624, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 14959 + }, + { + "epoch": 0.5990977443609022, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 14960 + }, + { + "epoch": 0.5991578947368421, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 14961 + }, + { + "epoch": 0.599218045112782, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 14962 + }, + { + "epoch": 0.5992781954887219, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 14963 + }, + { + "epoch": 0.5993383458646616, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.3006, + "step": 14964 + }, + { + "epoch": 0.5993984962406015, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 14965 + }, + { + "epoch": 0.5994586466165414, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 14966 + }, + { + "epoch": 0.5995187969924812, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 14967 + }, + { + "epoch": 0.5995789473684211, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 14968 + }, + { + "epoch": 0.5996390977443609, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 14969 + }, + { + "epoch": 0.5996992481203007, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 14970 + }, + { + "epoch": 0.5997593984962406, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 14971 + }, + { + "epoch": 0.5998195488721805, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 14972 + }, + { + "epoch": 0.5998796992481203, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1965, + "step": 14973 + }, + { + "epoch": 0.5999398496240601, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 14974 + }, + { + "epoch": 0.6, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 14975 + }, + { + "epoch": 0.6000601503759398, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 14976 + }, + { + "epoch": 0.6001203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 14977 + }, + { + "epoch": 0.6001804511278196, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 14978 + }, + { + "epoch": 0.6002406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 14979 + }, + { + "epoch": 0.6003007518796992, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 14980 + }, + { + "epoch": 0.6003609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1254, + "step": 14981 + }, + { + "epoch": 0.600421052631579, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 14982 + }, + { + "epoch": 0.6004812030075188, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 14983 + }, + { + "epoch": 0.6005413533834586, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 14984 + }, + { + "epoch": 0.6006015037593985, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 14985 + }, + { + "epoch": 0.6006616541353383, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 14986 + }, + { + "epoch": 0.6007218045112782, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 14987 + }, + { + "epoch": 0.6007819548872181, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 14988 + }, + { + "epoch": 0.6008421052631578, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 14989 + }, + { + "epoch": 0.6009022556390977, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1837, + "step": 14990 + }, + { + "epoch": 0.6009624060150376, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 14991 + }, + { + "epoch": 0.6010225563909775, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1254, + "step": 14992 + }, + { + "epoch": 0.6010827067669173, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.202, + "step": 14993 + }, + { + "epoch": 0.6011428571428571, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 14994 + }, + { + "epoch": 0.601203007518797, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1414, + "step": 14995 + }, + { + "epoch": 0.6012631578947368, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 14996 + }, + { + "epoch": 0.6013233082706767, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 14997 + }, + { + "epoch": 0.6013834586466166, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.2278, + "step": 14998 + }, + { + "epoch": 0.6014436090225563, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.2304, + "step": 14999 + }, + { + "epoch": 0.6015037593984962, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 15000 + }, + { + "epoch": 0.6015639097744361, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 15001 + }, + { + "epoch": 0.601624060150376, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1852, + "step": 15002 + }, + { + "epoch": 0.6016842105263158, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 15003 + }, + { + "epoch": 0.6017443609022557, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2072, + "step": 15004 + }, + { + "epoch": 0.6018045112781955, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 15005 + }, + { + "epoch": 0.6018646616541353, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 15006 + }, + { + "epoch": 0.6019248120300752, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1865, + "step": 15007 + }, + { + "epoch": 0.6019849624060151, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 15008 + }, + { + "epoch": 0.6020451127819549, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2173, + "step": 15009 + }, + { + "epoch": 0.6021052631578947, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 15010 + }, + { + "epoch": 0.6021654135338346, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2544, + "step": 15011 + }, + { + "epoch": 0.6022255639097744, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 15012 + }, + { + "epoch": 0.6022857142857143, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 15013 + }, + { + "epoch": 0.6023458646616542, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 15014 + }, + { + "epoch": 0.602406015037594, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 15015 + }, + { + "epoch": 0.6024661654135338, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 15016 + }, + { + "epoch": 0.6025263157894737, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 15017 + }, + { + "epoch": 0.6025864661654136, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 15018 + }, + { + "epoch": 0.6026466165413534, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 15019 + }, + { + "epoch": 0.6027067669172932, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 15020 + }, + { + "epoch": 0.6027669172932331, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 15021 + }, + { + "epoch": 0.6028270676691729, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 15022 + }, + { + "epoch": 0.6028872180451128, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 15023 + }, + { + "epoch": 0.6029473684210527, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 15024 + }, + { + "epoch": 0.6030075187969924, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 15025 + }, + { + "epoch": 0.6030676691729323, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 15026 + }, + { + "epoch": 0.6031278195488722, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 15027 + }, + { + "epoch": 0.603187969924812, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 15028 + }, + { + "epoch": 0.6032481203007519, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 15029 + }, + { + "epoch": 0.6033082706766917, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 15030 + }, + { + "epoch": 0.6033684210526316, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 15031 + }, + { + "epoch": 0.6034285714285714, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2048, + "step": 15032 + }, + { + "epoch": 0.6034887218045113, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 15033 + }, + { + "epoch": 0.6035488721804512, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 15034 + }, + { + "epoch": 0.6036090225563909, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 15035 + }, + { + "epoch": 0.6036691729323308, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2772, + "step": 15036 + }, + { + "epoch": 0.6037293233082707, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 15037 + }, + { + "epoch": 0.6037894736842105, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 15038 + }, + { + "epoch": 0.6038496240601504, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 15039 + }, + { + "epoch": 0.6039097744360902, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 15040 + }, + { + "epoch": 0.60396992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 15041 + }, + { + "epoch": 0.6040300751879699, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 15042 + }, + { + "epoch": 0.6040902255639098, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1911, + "step": 15043 + }, + { + "epoch": 0.6041503759398497, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 15044 + }, + { + "epoch": 0.6042105263157894, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 15045 + }, + { + "epoch": 0.6042706766917293, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 15046 + }, + { + "epoch": 0.6043308270676692, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 15047 + }, + { + "epoch": 0.604390977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 15048 + }, + { + "epoch": 0.6044511278195489, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 15049 + }, + { + "epoch": 0.6045112781954888, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 15050 + }, + { + "epoch": 0.6045714285714285, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 15051 + }, + { + "epoch": 0.6046315789473684, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2518, + "step": 15052 + }, + { + "epoch": 0.6046917293233083, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 15053 + }, + { + "epoch": 0.6047518796992482, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 15054 + }, + { + "epoch": 0.604812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 15055 + }, + { + "epoch": 0.6048721804511278, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1597, + "step": 15056 + }, + { + "epoch": 0.6049323308270677, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 15057 + }, + { + "epoch": 0.6049924812030075, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 15058 + }, + { + "epoch": 0.6050526315789474, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 15059 + }, + { + "epoch": 0.6051127819548873, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 15060 + }, + { + "epoch": 0.605172932330827, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 15061 + }, + { + "epoch": 0.6052330827067669, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 15062 + }, + { + "epoch": 0.6052932330827068, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1822, + "step": 15063 + }, + { + "epoch": 0.6053533834586466, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 15064 + }, + { + "epoch": 0.6054135338345865, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 15065 + }, + { + "epoch": 0.6054736842105263, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 15066 + }, + { + "epoch": 0.6055338345864661, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 15067 + }, + { + "epoch": 0.605593984962406, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 15068 + }, + { + "epoch": 0.6056541353383459, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 15069 + }, + { + "epoch": 0.6057142857142858, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 15070 + }, + { + "epoch": 0.6057744360902255, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 15071 + }, + { + "epoch": 0.6058345864661654, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 15072 + }, + { + "epoch": 0.6058947368421053, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 15073 + }, + { + "epoch": 0.6059548872180451, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1823, + "step": 15074 + }, + { + "epoch": 0.606015037593985, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2542, + "step": 15075 + }, + { + "epoch": 0.6060751879699248, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 15076 + }, + { + "epoch": 0.6061353383458646, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 15077 + }, + { + "epoch": 0.6061954887218045, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 15078 + }, + { + "epoch": 0.6062556390977444, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 15079 + }, + { + "epoch": 0.6063157894736843, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 15080 + }, + { + "epoch": 0.606375939849624, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 15081 + }, + { + "epoch": 0.6064360902255639, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 15082 + }, + { + "epoch": 0.6064962406015038, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 15083 + }, + { + "epoch": 0.6065563909774436, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 15084 + }, + { + "epoch": 0.6066165413533835, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 15085 + }, + { + "epoch": 0.6066766917293233, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1537, + "step": 15086 + }, + { + "epoch": 0.6067368421052631, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 15087 + }, + { + "epoch": 0.606796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 15088 + }, + { + "epoch": 0.6068571428571429, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1886, + "step": 15089 + }, + { + "epoch": 0.6069172932330827, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 15090 + }, + { + "epoch": 0.6069774436090225, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2265, + "step": 15091 + }, + { + "epoch": 0.6070375939849624, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 15092 + }, + { + "epoch": 0.6070977443609022, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 15093 + }, + { + "epoch": 0.6071578947368421, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 15094 + }, + { + "epoch": 0.607218045112782, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 15095 + }, + { + "epoch": 0.6072781954887218, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 15096 + }, + { + "epoch": 0.6073383458646616, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 15097 + }, + { + "epoch": 0.6073984962406015, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2147, + "step": 15098 + }, + { + "epoch": 0.6074586466165414, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 15099 + }, + { + "epoch": 0.6075187969924812, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2401, + "step": 15100 + }, + { + "epoch": 0.6075789473684211, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 15101 + }, + { + "epoch": 0.6076390977443609, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 15102 + }, + { + "epoch": 0.6076992481203007, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 15103 + }, + { + "epoch": 0.6077593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1959, + "step": 15104 + }, + { + "epoch": 0.6078195488721805, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 15105 + }, + { + "epoch": 0.6078796992481204, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1966, + "step": 15106 + }, + { + "epoch": 0.6079398496240601, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 15107 + }, + { + "epoch": 0.608, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 15108 + }, + { + "epoch": 0.6080601503759399, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 15109 + }, + { + "epoch": 0.6081203007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 15110 + }, + { + "epoch": 0.6081804511278196, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 15111 + }, + { + "epoch": 0.6082406015037594, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 15112 + }, + { + "epoch": 0.6083007518796992, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 15113 + }, + { + "epoch": 0.6083609022556391, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 15114 + }, + { + "epoch": 0.608421052631579, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2114, + "step": 15115 + }, + { + "epoch": 0.6084812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 15116 + }, + { + "epoch": 0.6085413533834586, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 15117 + }, + { + "epoch": 0.6086015037593985, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 15118 + }, + { + "epoch": 0.6086616541353383, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 15119 + }, + { + "epoch": 0.6087218045112782, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 15120 + }, + { + "epoch": 0.6087819548872181, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2133, + "step": 15121 + }, + { + "epoch": 0.6088421052631579, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 15122 + }, + { + "epoch": 0.6089022556390977, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 15123 + }, + { + "epoch": 0.6089624060150376, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 15124 + }, + { + "epoch": 0.6090225563909775, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 15125 + }, + { + "epoch": 0.6090827067669173, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 15126 + }, + { + "epoch": 0.6091428571428571, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 15127 + }, + { + "epoch": 0.609203007518797, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 15128 + }, + { + "epoch": 0.6092631578947368, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 15129 + }, + { + "epoch": 0.6093233082706767, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 15130 + }, + { + "epoch": 0.6093834586466166, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 15131 + }, + { + "epoch": 0.6094436090225563, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 15132 + }, + { + "epoch": 0.6095037593984962, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 15133 + }, + { + "epoch": 0.6095639097744361, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 15134 + }, + { + "epoch": 0.609624060150376, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 15135 + }, + { + "epoch": 0.6096842105263158, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 15136 + }, + { + "epoch": 0.6097443609022556, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 15137 + }, + { + "epoch": 0.6098045112781955, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 15138 + }, + { + "epoch": 0.6098646616541353, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1785, + "step": 15139 + }, + { + "epoch": 0.6099248120300752, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 15140 + }, + { + "epoch": 0.6099849624060151, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.3036, + "step": 15141 + }, + { + "epoch": 0.6100451127819548, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 15142 + }, + { + "epoch": 0.6101052631578947, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 15143 + }, + { + "epoch": 0.6101654135338346, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 15144 + }, + { + "epoch": 0.6102255639097744, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 15145 + }, + { + "epoch": 0.6102857142857143, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 15146 + }, + { + "epoch": 0.6103458646616541, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 15147 + }, + { + "epoch": 0.610406015037594, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 15148 + }, + { + "epoch": 0.6104661654135338, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1923, + "step": 15149 + }, + { + "epoch": 0.6105263157894737, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 15150 + }, + { + "epoch": 0.6105864661654136, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 15151 + }, + { + "epoch": 0.6106466165413534, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 15152 + }, + { + "epoch": 0.6107067669172932, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 15153 + }, + { + "epoch": 0.6107669172932331, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 15154 + }, + { + "epoch": 0.6108270676691729, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 15155 + }, + { + "epoch": 0.6108872180451128, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 15156 + }, + { + "epoch": 0.6109473684210527, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 15157 + }, + { + "epoch": 0.6110075187969924, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 15158 + }, + { + "epoch": 0.6110676691729323, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 15159 + }, + { + "epoch": 0.6111278195488722, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 15160 + }, + { + "epoch": 0.6111879699248121, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 15161 + }, + { + "epoch": 0.6112481203007519, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 15162 + }, + { + "epoch": 0.6113082706766917, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 15163 + }, + { + "epoch": 0.6113684210526316, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2355, + "step": 15164 + }, + { + "epoch": 0.6114285714285714, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 15165 + }, + { + "epoch": 0.6114887218045113, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2334, + "step": 15166 + }, + { + "epoch": 0.6115488721804512, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 15167 + }, + { + "epoch": 0.6116090225563909, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1731, + "step": 15168 + }, + { + "epoch": 0.6116691729323308, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 15169 + }, + { + "epoch": 0.6117293233082707, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 15170 + }, + { + "epoch": 0.6117894736842105, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 15171 + }, + { + "epoch": 0.6118496240601504, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 15172 + }, + { + "epoch": 0.6119097744360902, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 15173 + }, + { + "epoch": 0.61196992481203, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 15174 + }, + { + "epoch": 0.6120300751879699, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 15175 + }, + { + "epoch": 0.6120902255639098, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1482, + "step": 15176 + }, + { + "epoch": 0.6121503759398497, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 15177 + }, + { + "epoch": 0.6122105263157894, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 15178 + }, + { + "epoch": 0.6122706766917293, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 15179 + }, + { + "epoch": 0.6123308270676692, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1704, + "step": 15180 + }, + { + "epoch": 0.612390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 15181 + }, + { + "epoch": 0.6124511278195489, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 15182 + }, + { + "epoch": 0.6125112781954887, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 15183 + }, + { + "epoch": 0.6125714285714285, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 15184 + }, + { + "epoch": 0.6126315789473684, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 15185 + }, + { + "epoch": 0.6126917293233083, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 15186 + }, + { + "epoch": 0.6127518796992482, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 15187 + }, + { + "epoch": 0.6128120300751879, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 15188 + }, + { + "epoch": 0.6128721804511278, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 15189 + }, + { + "epoch": 0.6129323308270677, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2345, + "step": 15190 + }, + { + "epoch": 0.6129924812030075, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 15191 + }, + { + "epoch": 0.6130526315789474, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 15192 + }, + { + "epoch": 0.6131127819548872, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2217, + "step": 15193 + }, + { + "epoch": 0.613172932330827, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2086, + "step": 15194 + }, + { + "epoch": 0.6132330827067669, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 15195 + }, + { + "epoch": 0.6132932330827068, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 15196 + }, + { + "epoch": 0.6133533834586467, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 15197 + }, + { + "epoch": 0.6134135338345864, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 15198 + }, + { + "epoch": 0.6134736842105263, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 15199 + }, + { + "epoch": 0.6135338345864662, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 15200 + }, + { + "epoch": 0.613593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 15201 + }, + { + "epoch": 0.6136541353383459, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 15202 + }, + { + "epoch": 0.6137142857142858, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 15203 + }, + { + "epoch": 0.6137744360902255, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.155, + "step": 15204 + }, + { + "epoch": 0.6138345864661654, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 15205 + }, + { + "epoch": 0.6138947368421053, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2509, + "step": 15206 + }, + { + "epoch": 0.6139548872180451, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 15207 + }, + { + "epoch": 0.614015037593985, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 15208 + }, + { + "epoch": 0.6140751879699248, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 15209 + }, + { + "epoch": 0.6141353383458646, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 15210 + }, + { + "epoch": 0.6141954887218045, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 15211 + }, + { + "epoch": 0.6142556390977444, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 15212 + }, + { + "epoch": 0.6143157894736843, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 15213 + }, + { + "epoch": 0.614375939849624, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 15214 + }, + { + "epoch": 0.6144360902255639, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 15215 + }, + { + "epoch": 0.6144962406015038, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 15216 + }, + { + "epoch": 0.6145563909774436, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 15217 + }, + { + "epoch": 0.6146165413533835, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 15218 + }, + { + "epoch": 0.6146766917293233, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 15219 + }, + { + "epoch": 0.6147368421052631, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 15220 + }, + { + "epoch": 0.614796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 15221 + }, + { + "epoch": 0.6148571428571429, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 15222 + }, + { + "epoch": 0.6149172932330828, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 15223 + }, + { + "epoch": 0.6149774436090225, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1888, + "step": 15224 + }, + { + "epoch": 0.6150375939849624, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 15225 + }, + { + "epoch": 0.6150977443609023, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 15226 + }, + { + "epoch": 0.6151578947368421, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 15227 + }, + { + "epoch": 0.615218045112782, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 15228 + }, + { + "epoch": 0.6152781954887218, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 15229 + }, + { + "epoch": 0.6153383458646616, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1865, + "step": 15230 + }, + { + "epoch": 0.6153984962406015, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 15231 + }, + { + "epoch": 0.6154586466165414, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 15232 + }, + { + "epoch": 0.6155187969924812, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 15233 + }, + { + "epoch": 0.615578947368421, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1719, + "step": 15234 + }, + { + "epoch": 0.6156390977443609, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2075, + "step": 15235 + }, + { + "epoch": 0.6156992481203007, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2492, + "step": 15236 + }, + { + "epoch": 0.6157593984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 15237 + }, + { + "epoch": 0.6158195488721805, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 15238 + }, + { + "epoch": 0.6158796992481202, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2075, + "step": 15239 + }, + { + "epoch": 0.6159398496240601, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 15240 + }, + { + "epoch": 0.616, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 15241 + }, + { + "epoch": 0.6160601503759399, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 15242 + }, + { + "epoch": 0.6161203007518797, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1864, + "step": 15243 + }, + { + "epoch": 0.6161804511278195, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 15244 + }, + { + "epoch": 0.6162406015037594, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 15245 + }, + { + "epoch": 0.6163007518796992, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 15246 + }, + { + "epoch": 0.6163609022556391, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 15247 + }, + { + "epoch": 0.616421052631579, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 15248 + }, + { + "epoch": 0.6164812030075189, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 15249 + }, + { + "epoch": 0.6165413533834586, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 15250 + }, + { + "epoch": 0.6166015037593985, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 15251 + }, + { + "epoch": 0.6166616541353384, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.3658, + "step": 15252 + }, + { + "epoch": 0.6167218045112782, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 15253 + }, + { + "epoch": 0.6167819548872181, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 15254 + }, + { + "epoch": 0.6168421052631579, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 15255 + }, + { + "epoch": 0.6169022556390977, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 15256 + }, + { + "epoch": 0.6169624060150376, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2185, + "step": 15257 + }, + { + "epoch": 0.6170225563909775, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 15258 + }, + { + "epoch": 0.6170827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 15259 + }, + { + "epoch": 0.6171428571428571, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 15260 + }, + { + "epoch": 0.617203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 15261 + }, + { + "epoch": 0.6172631578947368, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2229, + "step": 15262 + }, + { + "epoch": 0.6173233082706767, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 15263 + }, + { + "epoch": 0.6173834586466166, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 15264 + }, + { + "epoch": 0.6174436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 15265 + }, + { + "epoch": 0.6175037593984962, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 15266 + }, + { + "epoch": 0.6175639097744361, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 15267 + }, + { + "epoch": 0.617624060150376, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 15268 + }, + { + "epoch": 0.6176842105263158, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 15269 + }, + { + "epoch": 0.6177443609022556, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 15270 + }, + { + "epoch": 0.6178045112781955, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 15271 + }, + { + "epoch": 0.6178646616541353, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 15272 + }, + { + "epoch": 0.6179248120300752, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 15273 + }, + { + "epoch": 0.6179849624060151, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 15274 + }, + { + "epoch": 0.6180451127819548, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 15275 + }, + { + "epoch": 0.6181052631578947, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 15276 + }, + { + "epoch": 0.6181654135338346, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2388, + "step": 15277 + }, + { + "epoch": 0.6182255639097745, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 15278 + }, + { + "epoch": 0.6182857142857143, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 15279 + }, + { + "epoch": 0.6183458646616541, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 15280 + }, + { + "epoch": 0.618406015037594, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 15281 + }, + { + "epoch": 0.6184661654135338, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.307, + "step": 15282 + }, + { + "epoch": 0.6185263157894737, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 15283 + }, + { + "epoch": 0.6185864661654136, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 15284 + }, + { + "epoch": 0.6186466165413533, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 15285 + }, + { + "epoch": 0.6187067669172932, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 15286 + }, + { + "epoch": 0.6187669172932331, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 15287 + }, + { + "epoch": 0.618827067669173, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 15288 + }, + { + "epoch": 0.6188872180451128, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 15289 + }, + { + "epoch": 0.6189473684210526, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 15290 + }, + { + "epoch": 0.6190075187969925, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 15291 + }, + { + "epoch": 0.6190676691729323, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 15292 + }, + { + "epoch": 0.6191278195488722, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1836, + "step": 15293 + }, + { + "epoch": 0.6191879699248121, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 15294 + }, + { + "epoch": 0.6192481203007518, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 15295 + }, + { + "epoch": 0.6193082706766917, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 15296 + }, + { + "epoch": 0.6193684210526316, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 15297 + }, + { + "epoch": 0.6194285714285714, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 15298 + }, + { + "epoch": 0.6194887218045113, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 15299 + }, + { + "epoch": 0.6195488721804512, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 15300 + }, + { + "epoch": 0.6196090225563909, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 15301 + }, + { + "epoch": 0.6196691729323308, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 15302 + }, + { + "epoch": 0.6197293233082707, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 15303 + }, + { + "epoch": 0.6197894736842106, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 15304 + }, + { + "epoch": 0.6198496240601504, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1635, + "step": 15305 + }, + { + "epoch": 0.6199097744360902, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 15306 + }, + { + "epoch": 0.6199699248120301, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 15307 + }, + { + "epoch": 0.6200300751879699, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 15308 + }, + { + "epoch": 0.6200902255639098, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2572, + "step": 15309 + }, + { + "epoch": 0.6201503759398497, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 15310 + }, + { + "epoch": 0.6202105263157894, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 15311 + }, + { + "epoch": 0.6202706766917293, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 15312 + }, + { + "epoch": 0.6203308270676692, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 15313 + }, + { + "epoch": 0.620390977443609, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 15314 + }, + { + "epoch": 0.6204511278195489, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 15315 + }, + { + "epoch": 0.6205112781954887, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 15316 + }, + { + "epoch": 0.6205714285714286, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 15317 + }, + { + "epoch": 0.6206315789473684, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 15318 + }, + { + "epoch": 0.6206917293233083, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 15319 + }, + { + "epoch": 0.6207518796992482, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 15320 + }, + { + "epoch": 0.6208120300751879, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 15321 + }, + { + "epoch": 0.6208721804511278, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 15322 + }, + { + "epoch": 0.6209323308270677, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 15323 + }, + { + "epoch": 0.6209924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 15324 + }, + { + "epoch": 0.6210526315789474, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 15325 + }, + { + "epoch": 0.6211127819548872, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1334, + "step": 15326 + }, + { + "epoch": 0.621172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.232, + "step": 15327 + }, + { + "epoch": 0.6212330827067669, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 15328 + }, + { + "epoch": 0.6212932330827068, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2491, + "step": 15329 + }, + { + "epoch": 0.6213533834586467, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 15330 + }, + { + "epoch": 0.6214135338345864, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.3052, + "step": 15331 + }, + { + "epoch": 0.6214736842105263, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 15332 + }, + { + "epoch": 0.6215338345864662, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2425, + "step": 15333 + }, + { + "epoch": 0.621593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 15334 + }, + { + "epoch": 0.6216541353383459, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 15335 + }, + { + "epoch": 0.6217142857142857, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 15336 + }, + { + "epoch": 0.6217744360902255, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 15337 + }, + { + "epoch": 0.6218345864661654, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 15338 + }, + { + "epoch": 0.6218947368421053, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 15339 + }, + { + "epoch": 0.6219548872180451, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1757, + "step": 15340 + }, + { + "epoch": 0.6220150375939849, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 15341 + }, + { + "epoch": 0.6220751879699248, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 15342 + }, + { + "epoch": 0.6221353383458647, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 15343 + }, + { + "epoch": 0.6221954887218045, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 15344 + }, + { + "epoch": 0.6222556390977444, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 15345 + }, + { + "epoch": 0.6223157894736842, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 15346 + }, + { + "epoch": 0.622375939849624, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 15347 + }, + { + "epoch": 0.6224360902255639, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 15348 + }, + { + "epoch": 0.6224962406015038, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 15349 + }, + { + "epoch": 0.6225563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 15350 + }, + { + "epoch": 0.6226165413533835, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1738, + "step": 15351 + }, + { + "epoch": 0.6226766917293233, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 15352 + }, + { + "epoch": 0.6227368421052631, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 15353 + }, + { + "epoch": 0.622796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 15354 + }, + { + "epoch": 0.6228571428571429, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1823, + "step": 15355 + }, + { + "epoch": 0.6229172932330828, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 15356 + }, + { + "epoch": 0.6229774436090225, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 15357 + }, + { + "epoch": 0.6230375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 15358 + }, + { + "epoch": 0.6230977443609023, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 15359 + }, + { + "epoch": 0.6231578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 15360 + }, + { + "epoch": 0.623218045112782, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 15361 + }, + { + "epoch": 0.6232781954887218, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1635, + "step": 15362 + }, + { + "epoch": 0.6233383458646616, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 15363 + }, + { + "epoch": 0.6233984962406015, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2844, + "step": 15364 + }, + { + "epoch": 0.6234586466165414, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 15365 + }, + { + "epoch": 0.6235187969924813, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 15366 + }, + { + "epoch": 0.623578947368421, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 15367 + }, + { + "epoch": 0.6236390977443609, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 15368 + }, + { + "epoch": 0.6236992481203008, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.2024, + "step": 15369 + }, + { + "epoch": 0.6237593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1482, + "step": 15370 + }, + { + "epoch": 0.6238195488721805, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.218, + "step": 15371 + }, + { + "epoch": 0.6238796992481203, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 15372 + }, + { + "epoch": 0.6239398496240601, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 15373 + }, + { + "epoch": 0.624, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 15374 + }, + { + "epoch": 0.6240601503759399, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 15375 + }, + { + "epoch": 0.6241203007518797, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 15376 + }, + { + "epoch": 0.6241804511278195, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 15377 + }, + { + "epoch": 0.6242406015037594, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2497, + "step": 15378 + }, + { + "epoch": 0.6243007518796992, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2352, + "step": 15379 + }, + { + "epoch": 0.6243609022556391, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 15380 + }, + { + "epoch": 0.624421052631579, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 15381 + }, + { + "epoch": 0.6244812030075187, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 15382 + }, + { + "epoch": 0.6245413533834586, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 15383 + }, + { + "epoch": 0.6246015037593985, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 15384 + }, + { + "epoch": 0.6246616541353384, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3667, + "step": 15385 + }, + { + "epoch": 0.6247218045112782, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 15386 + }, + { + "epoch": 0.624781954887218, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 15387 + }, + { + "epoch": 0.6248421052631579, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 15388 + }, + { + "epoch": 0.6249022556390977, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 15389 + }, + { + "epoch": 0.6249624060150376, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 15390 + }, + { + "epoch": 0.6250225563909775, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2435, + "step": 15391 + }, + { + "epoch": 0.6250827067669172, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2189, + "step": 15392 + }, + { + "epoch": 0.6251428571428571, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2008, + "step": 15393 + }, + { + "epoch": 0.625203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 15394 + }, + { + "epoch": 0.6252631578947369, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3436, + "step": 15395 + }, + { + "epoch": 0.6253233082706767, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 15396 + }, + { + "epoch": 0.6253834586466166, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 15397 + }, + { + "epoch": 0.6254436090225564, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 15398 + }, + { + "epoch": 0.6255037593984962, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2249, + "step": 15399 + }, + { + "epoch": 0.6255639097744361, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 15400 + }, + { + "epoch": 0.625624060150376, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 15401 + }, + { + "epoch": 0.6256842105263158, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 15402 + }, + { + "epoch": 0.6257443609022556, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 15403 + }, + { + "epoch": 0.6258045112781955, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 15404 + }, + { + "epoch": 0.6258646616541353, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 15405 + }, + { + "epoch": 0.6259248120300752, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 15406 + }, + { + "epoch": 0.6259849624060151, + "grad_norm": 1.3671875, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 15407 + }, + { + "epoch": 0.6260451127819548, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 15408 + }, + { + "epoch": 0.6261052631578947, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 15409 + }, + { + "epoch": 0.6261654135338346, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 15410 + }, + { + "epoch": 0.6262255639097745, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.3588, + "step": 15411 + }, + { + "epoch": 0.6262857142857143, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 15412 + }, + { + "epoch": 0.6263458646616541, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 15413 + }, + { + "epoch": 0.626406015037594, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.2077, + "step": 15414 + }, + { + "epoch": 0.6264661654135338, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 15415 + }, + { + "epoch": 0.6265263157894737, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 15416 + }, + { + "epoch": 0.6265864661654136, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 15417 + }, + { + "epoch": 0.6266466165413533, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2468, + "step": 15418 + }, + { + "epoch": 0.6267067669172932, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 15419 + }, + { + "epoch": 0.6267669172932331, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 15420 + }, + { + "epoch": 0.626827067669173, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 15421 + }, + { + "epoch": 0.6268872180451128, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 15422 + }, + { + "epoch": 0.6269473684210526, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 15423 + }, + { + "epoch": 0.6270075187969925, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 15424 + }, + { + "epoch": 0.6270676691729323, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1564, + "step": 15425 + }, + { + "epoch": 0.6271278195488722, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2172, + "step": 15426 + }, + { + "epoch": 0.6271879699248121, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 15427 + }, + { + "epoch": 0.6272481203007518, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 15428 + }, + { + "epoch": 0.6273082706766917, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 15429 + }, + { + "epoch": 0.6273684210526316, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 15430 + }, + { + "epoch": 0.6274285714285714, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1793, + "step": 15431 + }, + { + "epoch": 0.6274887218045113, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 15432 + }, + { + "epoch": 0.6275488721804511, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 15433 + }, + { + "epoch": 0.627609022556391, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 15434 + }, + { + "epoch": 0.6276691729323308, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2987, + "step": 15435 + }, + { + "epoch": 0.6277293233082707, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 15436 + }, + { + "epoch": 0.6277894736842106, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 15437 + }, + { + "epoch": 0.6278496240601503, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 15438 + }, + { + "epoch": 0.6279097744360902, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 15439 + }, + { + "epoch": 0.6279699248120301, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 15440 + }, + { + "epoch": 0.6280300751879699, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1708, + "step": 15441 + }, + { + "epoch": 0.6280902255639098, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 15442 + }, + { + "epoch": 0.6281503759398496, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 15443 + }, + { + "epoch": 0.6282105263157894, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 15444 + }, + { + "epoch": 0.6282706766917293, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 15445 + }, + { + "epoch": 0.6283308270676692, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 15446 + }, + { + "epoch": 0.6283909774436091, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 15447 + }, + { + "epoch": 0.6284511278195489, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 15448 + }, + { + "epoch": 0.6285112781954887, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2384, + "step": 15449 + }, + { + "epoch": 0.6285714285714286, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1725, + "step": 15450 + }, + { + "epoch": 0.6286315789473684, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 15451 + }, + { + "epoch": 0.6286917293233083, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.16, + "step": 15452 + }, + { + "epoch": 0.6287518796992482, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 15453 + }, + { + "epoch": 0.6288120300751879, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 15454 + }, + { + "epoch": 0.6288721804511278, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 15455 + }, + { + "epoch": 0.6289323308270677, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 15456 + }, + { + "epoch": 0.6289924812030075, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 15457 + }, + { + "epoch": 0.6290526315789474, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 15458 + }, + { + "epoch": 0.6291127819548872, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 15459 + }, + { + "epoch": 0.629172932330827, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1635, + "step": 15460 + }, + { + "epoch": 0.6292330827067669, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 15461 + }, + { + "epoch": 0.6292932330827068, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 15462 + }, + { + "epoch": 0.6293533834586467, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 15463 + }, + { + "epoch": 0.6294135338345864, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 15464 + }, + { + "epoch": 0.6294736842105263, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 15465 + }, + { + "epoch": 0.6295338345864662, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 15466 + }, + { + "epoch": 0.629593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 15467 + }, + { + "epoch": 0.6296541353383459, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 15468 + }, + { + "epoch": 0.6297142857142857, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 15469 + }, + { + "epoch": 0.6297744360902255, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 15470 + }, + { + "epoch": 0.6298345864661654, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 15471 + }, + { + "epoch": 0.6298947368421053, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 15472 + }, + { + "epoch": 0.6299548872180452, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 15473 + }, + { + "epoch": 0.6300150375939849, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 15474 + }, + { + "epoch": 0.6300751879699248, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 15475 + }, + { + "epoch": 0.6301353383458647, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 15476 + }, + { + "epoch": 0.6301954887218045, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 15477 + }, + { + "epoch": 0.6302556390977444, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 15478 + }, + { + "epoch": 0.6303157894736842, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 15479 + }, + { + "epoch": 0.630375939849624, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 15480 + }, + { + "epoch": 0.6304360902255639, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 15481 + }, + { + "epoch": 0.6304962406015038, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2467, + "step": 15482 + }, + { + "epoch": 0.6305563909774436, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 15483 + }, + { + "epoch": 0.6306165413533834, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 15484 + }, + { + "epoch": 0.6306766917293233, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2167, + "step": 15485 + }, + { + "epoch": 0.6307368421052632, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2002, + "step": 15486 + }, + { + "epoch": 0.630796992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 15487 + }, + { + "epoch": 0.6308571428571429, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2151, + "step": 15488 + }, + { + "epoch": 0.6309172932330827, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 15489 + }, + { + "epoch": 0.6309774436090225, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 15490 + }, + { + "epoch": 0.6310375939849624, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 15491 + }, + { + "epoch": 0.6310977443609023, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.197, + "step": 15492 + }, + { + "epoch": 0.6311578947368421, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 15493 + }, + { + "epoch": 0.6312180451127819, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1682, + "step": 15494 + }, + { + "epoch": 0.6312781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 15495 + }, + { + "epoch": 0.6313383458646616, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1913, + "step": 15496 + }, + { + "epoch": 0.6313984962406015, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 15497 + }, + { + "epoch": 0.6314586466165414, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 15498 + }, + { + "epoch": 0.6315187969924813, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 15499 + }, + { + "epoch": 0.631578947368421, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 15500 + }, + { + "epoch": 0.6316390977443609, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 15501 + }, + { + "epoch": 0.6316992481203008, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1467, + "step": 15502 + }, + { + "epoch": 0.6317593984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 15503 + }, + { + "epoch": 0.6318195488721805, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1908, + "step": 15504 + }, + { + "epoch": 0.6318796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 15505 + }, + { + "epoch": 0.6319398496240601, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 15506 + }, + { + "epoch": 0.632, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 15507 + }, + { + "epoch": 0.6320601503759399, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1543, + "step": 15508 + }, + { + "epoch": 0.6321203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 15509 + }, + { + "epoch": 0.6321804511278195, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 15510 + }, + { + "epoch": 0.6322406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1597, + "step": 15511 + }, + { + "epoch": 0.6323007518796993, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 15512 + }, + { + "epoch": 0.6323609022556391, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 15513 + }, + { + "epoch": 0.632421052631579, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 15514 + }, + { + "epoch": 0.6324812030075188, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1785, + "step": 15515 + }, + { + "epoch": 0.6325413533834586, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 15516 + }, + { + "epoch": 0.6326015037593985, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 15517 + }, + { + "epoch": 0.6326616541353384, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 15518 + }, + { + "epoch": 0.6327218045112782, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 15519 + }, + { + "epoch": 0.632781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 15520 + }, + { + "epoch": 0.6328421052631579, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 15521 + }, + { + "epoch": 0.6329022556390977, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 15522 + }, + { + "epoch": 0.6329624060150376, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 15523 + }, + { + "epoch": 0.6330225563909775, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 15524 + }, + { + "epoch": 0.6330827067669172, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.3006, + "step": 15525 + }, + { + "epoch": 0.6331428571428571, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 15526 + }, + { + "epoch": 0.633203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 15527 + }, + { + "epoch": 0.6332631578947369, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 15528 + }, + { + "epoch": 0.6333233082706767, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 15529 + }, + { + "epoch": 0.6333834586466165, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 15530 + }, + { + "epoch": 0.6334436090225564, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 15531 + }, + { + "epoch": 0.6335037593984962, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 15532 + }, + { + "epoch": 0.6335639097744361, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 15533 + }, + { + "epoch": 0.633624060150376, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1705, + "step": 15534 + }, + { + "epoch": 0.6336842105263157, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 15535 + }, + { + "epoch": 0.6337443609022556, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 15536 + }, + { + "epoch": 0.6338045112781955, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 15537 + }, + { + "epoch": 0.6338646616541354, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 15538 + }, + { + "epoch": 0.6339248120300752, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 15539 + }, + { + "epoch": 0.633984962406015, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 15540 + }, + { + "epoch": 0.6340451127819549, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 15541 + }, + { + "epoch": 0.6341052631578947, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 15542 + }, + { + "epoch": 0.6341654135338346, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 15543 + }, + { + "epoch": 0.6342255639097745, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 15544 + }, + { + "epoch": 0.6342857142857142, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 15545 + }, + { + "epoch": 0.6343458646616541, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 15546 + }, + { + "epoch": 0.634406015037594, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 15547 + }, + { + "epoch": 0.6344661654135338, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 15548 + }, + { + "epoch": 0.6345263157894737, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 15549 + }, + { + "epoch": 0.6345864661654136, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1781, + "step": 15550 + }, + { + "epoch": 0.6346466165413533, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 15551 + }, + { + "epoch": 0.6347067669172932, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 15552 + }, + { + "epoch": 0.6347669172932331, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 15553 + }, + { + "epoch": 0.634827067669173, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 15554 + }, + { + "epoch": 0.6348872180451128, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 15555 + }, + { + "epoch": 0.6349473684210526, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 15556 + }, + { + "epoch": 0.6350075187969925, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 15557 + }, + { + "epoch": 0.6350676691729323, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 15558 + }, + { + "epoch": 0.6351278195488722, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2483, + "step": 15559 + }, + { + "epoch": 0.6351879699248121, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 15560 + }, + { + "epoch": 0.6352481203007518, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 15561 + }, + { + "epoch": 0.6353082706766917, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2495, + "step": 15562 + }, + { + "epoch": 0.6353684210526316, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 15563 + }, + { + "epoch": 0.6354285714285715, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 15564 + }, + { + "epoch": 0.6354887218045113, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 15565 + }, + { + "epoch": 0.6355488721804511, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2665, + "step": 15566 + }, + { + "epoch": 0.635609022556391, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1881, + "step": 15567 + }, + { + "epoch": 0.6356691729323308, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1894, + "step": 15568 + }, + { + "epoch": 0.6357293233082707, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 15569 + }, + { + "epoch": 0.6357894736842106, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2411, + "step": 15570 + }, + { + "epoch": 0.6358496240601503, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1767, + "step": 15571 + }, + { + "epoch": 0.6359097744360902, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 15572 + }, + { + "epoch": 0.6359699248120301, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 15573 + }, + { + "epoch": 0.63603007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 15574 + }, + { + "epoch": 0.6360902255639098, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 15575 + }, + { + "epoch": 0.6361503759398496, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 15576 + }, + { + "epoch": 0.6362105263157894, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 15577 + }, + { + "epoch": 0.6362706766917293, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 15578 + }, + { + "epoch": 0.6363308270676692, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 15579 + }, + { + "epoch": 0.6363909774436091, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 15580 + }, + { + "epoch": 0.6364511278195488, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 15581 + }, + { + "epoch": 0.6365112781954887, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 15582 + }, + { + "epoch": 0.6365714285714286, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 15583 + }, + { + "epoch": 0.6366315789473684, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 15584 + }, + { + "epoch": 0.6366917293233083, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 15585 + }, + { + "epoch": 0.6367518796992481, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1764, + "step": 15586 + }, + { + "epoch": 0.6368120300751879, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 15587 + }, + { + "epoch": 0.6368721804511278, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 15588 + }, + { + "epoch": 0.6369323308270677, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 15589 + }, + { + "epoch": 0.6369924812030076, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.2432, + "step": 15590 + }, + { + "epoch": 0.6370526315789473, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 15591 + }, + { + "epoch": 0.6371127819548872, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1691, + "step": 15592 + }, + { + "epoch": 0.6371729323308271, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2177, + "step": 15593 + }, + { + "epoch": 0.6372330827067669, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 15594 + }, + { + "epoch": 0.6372932330827068, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 15595 + }, + { + "epoch": 0.6373533834586467, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 15596 + }, + { + "epoch": 0.6374135338345864, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 15597 + }, + { + "epoch": 0.6374736842105263, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 15598 + }, + { + "epoch": 0.6375338345864662, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 15599 + }, + { + "epoch": 0.637593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 15600 + }, + { + "epoch": 0.6376541353383459, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 15601 + }, + { + "epoch": 0.6377142857142857, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 15602 + }, + { + "epoch": 0.6377744360902255, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 15603 + }, + { + "epoch": 0.6378345864661654, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 15604 + }, + { + "epoch": 0.6378947368421053, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 15605 + }, + { + "epoch": 0.6379548872180452, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 15606 + }, + { + "epoch": 0.6380150375939849, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2245, + "step": 15607 + }, + { + "epoch": 0.6380751879699248, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2007, + "step": 15608 + }, + { + "epoch": 0.6381353383458647, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2164, + "step": 15609 + }, + { + "epoch": 0.6381954887218045, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 15610 + }, + { + "epoch": 0.6382556390977444, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 15611 + }, + { + "epoch": 0.6383157894736842, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 15612 + }, + { + "epoch": 0.638375939849624, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 15613 + }, + { + "epoch": 0.6384360902255639, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 15614 + }, + { + "epoch": 0.6384962406015038, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 15615 + }, + { + "epoch": 0.6385563909774437, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1761, + "step": 15616 + }, + { + "epoch": 0.6386165413533834, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 15617 + }, + { + "epoch": 0.6386766917293233, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 15618 + }, + { + "epoch": 0.6387368421052632, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.2152, + "step": 15619 + }, + { + "epoch": 0.638796992481203, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1991, + "step": 15620 + }, + { + "epoch": 0.6388571428571429, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 15621 + }, + { + "epoch": 0.6389172932330827, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2436, + "step": 15622 + }, + { + "epoch": 0.6389774436090225, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 15623 + }, + { + "epoch": 0.6390375939849624, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 15624 + }, + { + "epoch": 0.6390977443609023, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 15625 + }, + { + "epoch": 0.6391578947368421, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 15626 + }, + { + "epoch": 0.6392180451127819, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 15627 + }, + { + "epoch": 0.6392781954887218, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 15628 + }, + { + "epoch": 0.6393383458646617, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 15629 + }, + { + "epoch": 0.6393984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 15630 + }, + { + "epoch": 0.6394586466165414, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2122, + "step": 15631 + }, + { + "epoch": 0.6395187969924812, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1983, + "step": 15632 + }, + { + "epoch": 0.639578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 15633 + }, + { + "epoch": 0.6396390977443609, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2365, + "step": 15634 + }, + { + "epoch": 0.6396992481203008, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2086, + "step": 15635 + }, + { + "epoch": 0.6397593984962406, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 15636 + }, + { + "epoch": 0.6398195488721804, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 15637 + }, + { + "epoch": 0.6398796992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 15638 + }, + { + "epoch": 0.6399398496240601, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 15639 + }, + { + "epoch": 0.64, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 15640 + }, + { + "epoch": 0.6400601503759399, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 15641 + }, + { + "epoch": 0.6401203007518796, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 15642 + }, + { + "epoch": 0.6401804511278195, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 15643 + }, + { + "epoch": 0.6402406015037594, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 15644 + }, + { + "epoch": 0.6403007518796993, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 15645 + }, + { + "epoch": 0.6403609022556391, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 15646 + }, + { + "epoch": 0.640421052631579, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 15647 + }, + { + "epoch": 0.6404812030075188, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.2639, + "step": 15648 + }, + { + "epoch": 0.6405413533834586, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 15649 + }, + { + "epoch": 0.6406015037593985, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1727, + "step": 15650 + }, + { + "epoch": 0.6406616541353384, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 15651 + }, + { + "epoch": 0.6407218045112782, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 15652 + }, + { + "epoch": 0.640781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 15653 + }, + { + "epoch": 0.6408421052631579, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.156, + "step": 15654 + }, + { + "epoch": 0.6409022556390978, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 15655 + }, + { + "epoch": 0.6409624060150376, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 15656 + }, + { + "epoch": 0.6410225563909775, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 15657 + }, + { + "epoch": 0.6410827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 15658 + }, + { + "epoch": 0.6411428571428571, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 15659 + }, + { + "epoch": 0.641203007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1861, + "step": 15660 + }, + { + "epoch": 0.6412631578947369, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 15661 + }, + { + "epoch": 0.6413233082706767, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1626, + "step": 15662 + }, + { + "epoch": 0.6413834586466165, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 15663 + }, + { + "epoch": 0.6414436090225564, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 15664 + }, + { + "epoch": 0.6415037593984962, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 15665 + }, + { + "epoch": 0.6415639097744361, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 15666 + }, + { + "epoch": 0.641624060150376, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 15667 + }, + { + "epoch": 0.6416842105263157, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 15668 + }, + { + "epoch": 0.6417443609022556, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 15669 + }, + { + "epoch": 0.6418045112781955, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 15670 + }, + { + "epoch": 0.6418646616541354, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 15671 + }, + { + "epoch": 0.6419248120300752, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1637, + "step": 15672 + }, + { + "epoch": 0.641984962406015, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1988, + "step": 15673 + }, + { + "epoch": 0.6420451127819549, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1761, + "step": 15674 + }, + { + "epoch": 0.6421052631578947, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 15675 + }, + { + "epoch": 0.6421654135338346, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 15676 + }, + { + "epoch": 0.6422255639097745, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2664, + "step": 15677 + }, + { + "epoch": 0.6422857142857142, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 15678 + }, + { + "epoch": 0.6423458646616541, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 15679 + }, + { + "epoch": 0.642406015037594, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 15680 + }, + { + "epoch": 0.6424661654135339, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 15681 + }, + { + "epoch": 0.6425263157894737, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 15682 + }, + { + "epoch": 0.6425864661654135, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 15683 + }, + { + "epoch": 0.6426466165413534, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 15684 + }, + { + "epoch": 0.6427067669172932, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 15685 + }, + { + "epoch": 0.6427669172932331, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 15686 + }, + { + "epoch": 0.642827067669173, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 15687 + }, + { + "epoch": 0.6428872180451127, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 15688 + }, + { + "epoch": 0.6429473684210526, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 15689 + }, + { + "epoch": 0.6430075187969925, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 15690 + }, + { + "epoch": 0.6430676691729323, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 15691 + }, + { + "epoch": 0.6431278195488722, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.2965, + "step": 15692 + }, + { + "epoch": 0.643187969924812, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 15693 + }, + { + "epoch": 0.6432481203007518, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 15694 + }, + { + "epoch": 0.6433082706766917, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 15695 + }, + { + "epoch": 0.6433684210526316, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 15696 + }, + { + "epoch": 0.6434285714285715, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 15697 + }, + { + "epoch": 0.6434887218045113, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 15698 + }, + { + "epoch": 0.6435488721804511, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 15699 + }, + { + "epoch": 0.643609022556391, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 15700 + }, + { + "epoch": 0.6436691729323308, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 15701 + }, + { + "epoch": 0.6437293233082707, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 15702 + }, + { + "epoch": 0.6437894736842106, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 15703 + }, + { + "epoch": 0.6438496240601503, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 15704 + }, + { + "epoch": 0.6439097744360902, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 15705 + }, + { + "epoch": 0.6439699248120301, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 15706 + }, + { + "epoch": 0.64403007518797, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1781, + "step": 15707 + }, + { + "epoch": 0.6440902255639098, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 15708 + }, + { + "epoch": 0.6441503759398496, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 15709 + }, + { + "epoch": 0.6442105263157895, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 15710 + }, + { + "epoch": 0.6442706766917293, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 15711 + }, + { + "epoch": 0.6443308270676692, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 15712 + }, + { + "epoch": 0.6443909774436091, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 15713 + }, + { + "epoch": 0.6444511278195488, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 15714 + }, + { + "epoch": 0.6445112781954887, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 15715 + }, + { + "epoch": 0.6445714285714286, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 15716 + }, + { + "epoch": 0.6446315789473684, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 15717 + }, + { + "epoch": 0.6446917293233083, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 15718 + }, + { + "epoch": 0.6447518796992481, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 15719 + }, + { + "epoch": 0.644812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 15720 + }, + { + "epoch": 0.6448721804511278, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 15721 + }, + { + "epoch": 0.6449323308270677, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 15722 + }, + { + "epoch": 0.6449924812030076, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 15723 + }, + { + "epoch": 0.6450526315789473, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 15724 + }, + { + "epoch": 0.6451127819548872, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 15725 + }, + { + "epoch": 0.6451729323308271, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 15726 + }, + { + "epoch": 0.6452330827067669, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 15727 + }, + { + "epoch": 0.6452932330827068, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 15728 + }, + { + "epoch": 0.6453533834586466, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 15729 + }, + { + "epoch": 0.6454135338345864, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 15730 + }, + { + "epoch": 0.6454736842105263, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2442, + "step": 15731 + }, + { + "epoch": 0.6455338345864662, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 15732 + }, + { + "epoch": 0.645593984962406, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 15733 + }, + { + "epoch": 0.6456541353383458, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 15734 + }, + { + "epoch": 0.6457142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 15735 + }, + { + "epoch": 0.6457744360902256, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 15736 + }, + { + "epoch": 0.6458345864661654, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 15737 + }, + { + "epoch": 0.6458947368421053, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 15738 + }, + { + "epoch": 0.6459548872180451, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 15739 + }, + { + "epoch": 0.6460150375939849, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 15740 + }, + { + "epoch": 0.6460751879699248, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 15741 + }, + { + "epoch": 0.6461353383458647, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 15742 + }, + { + "epoch": 0.6461954887218045, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 15743 + }, + { + "epoch": 0.6462556390977443, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 15744 + }, + { + "epoch": 0.6463157894736842, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.213, + "step": 15745 + }, + { + "epoch": 0.646375939849624, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 15746 + }, + { + "epoch": 0.6464360902255639, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 15747 + }, + { + "epoch": 0.6464962406015038, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 15748 + }, + { + "epoch": 0.6465563909774437, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 15749 + }, + { + "epoch": 0.6466165413533834, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 15750 + }, + { + "epoch": 0.6466766917293233, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2307, + "step": 15751 + }, + { + "epoch": 0.6467368421052632, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 15752 + }, + { + "epoch": 0.646796992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 15753 + }, + { + "epoch": 0.6468571428571429, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 15754 + }, + { + "epoch": 0.6469172932330827, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 15755 + }, + { + "epoch": 0.6469774436090225, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 15756 + }, + { + "epoch": 0.6470375939849624, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2457, + "step": 15757 + }, + { + "epoch": 0.6470977443609023, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 15758 + }, + { + "epoch": 0.6471578947368422, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 15759 + }, + { + "epoch": 0.6472180451127819, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1541, + "step": 15760 + }, + { + "epoch": 0.6472781954887218, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 15761 + }, + { + "epoch": 0.6473383458646617, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 15762 + }, + { + "epoch": 0.6473984962406015, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 15763 + }, + { + "epoch": 0.6474586466165414, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 15764 + }, + { + "epoch": 0.6475187969924812, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 15765 + }, + { + "epoch": 0.647578947368421, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 15766 + }, + { + "epoch": 0.6476390977443609, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.3652, + "step": 15767 + }, + { + "epoch": 0.6476992481203008, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 15768 + }, + { + "epoch": 0.6477593984962406, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 15769 + }, + { + "epoch": 0.6478195488721804, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 15770 + }, + { + "epoch": 0.6478796992481203, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.192, + "step": 15771 + }, + { + "epoch": 0.6479398496240601, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 15772 + }, + { + "epoch": 0.648, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 15773 + }, + { + "epoch": 0.6480601503759399, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 15774 + }, + { + "epoch": 0.6481203007518797, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.2635, + "step": 15775 + }, + { + "epoch": 0.6481804511278195, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 15776 + }, + { + "epoch": 0.6482406015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 15777 + }, + { + "epoch": 0.6483007518796993, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 15778 + }, + { + "epoch": 0.6483609022556391, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 15779 + }, + { + "epoch": 0.6484210526315789, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 15780 + }, + { + "epoch": 0.6484812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2655, + "step": 15781 + }, + { + "epoch": 0.6485413533834586, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 15782 + }, + { + "epoch": 0.6486015037593985, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.204, + "step": 15783 + }, + { + "epoch": 0.6486616541353384, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 15784 + }, + { + "epoch": 0.6487218045112781, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 15785 + }, + { + "epoch": 0.648781954887218, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 15786 + }, + { + "epoch": 0.6488421052631579, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 15787 + }, + { + "epoch": 0.6489022556390978, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 15788 + }, + { + "epoch": 0.6489624060150376, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 15789 + }, + { + "epoch": 0.6490225563909774, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 15790 + }, + { + "epoch": 0.6490827067669173, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 15791 + }, + { + "epoch": 0.6491428571428571, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 15792 + }, + { + "epoch": 0.649203007518797, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 15793 + }, + { + "epoch": 0.6492631578947369, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 15794 + }, + { + "epoch": 0.6493233082706767, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 15795 + }, + { + "epoch": 0.6493834586466165, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 15796 + }, + { + "epoch": 0.6494436090225564, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1541, + "step": 15797 + }, + { + "epoch": 0.6495037593984962, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 15798 + }, + { + "epoch": 0.6495639097744361, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 15799 + }, + { + "epoch": 0.649624060150376, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 15800 + }, + { + "epoch": 0.6496842105263158, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 15801 + }, + { + "epoch": 0.6497443609022556, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 15802 + }, + { + "epoch": 0.6498045112781955, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 15803 + }, + { + "epoch": 0.6498646616541354, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 15804 + }, + { + "epoch": 0.6499248120300752, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 15805 + }, + { + "epoch": 0.649984962406015, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 15806 + }, + { + "epoch": 0.6500451127819549, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 15807 + }, + { + "epoch": 0.6501052631578947, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 15808 + }, + { + "epoch": 0.6501654135338346, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 15809 + }, + { + "epoch": 0.6502255639097745, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 15810 + }, + { + "epoch": 0.6502857142857142, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 15811 + }, + { + "epoch": 0.6503458646616541, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 15812 + }, + { + "epoch": 0.650406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 15813 + }, + { + "epoch": 0.6504661654135339, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 15814 + }, + { + "epoch": 0.6505263157894737, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 15815 + }, + { + "epoch": 0.6505864661654135, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 15816 + }, + { + "epoch": 0.6506466165413534, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2063, + "step": 15817 + }, + { + "epoch": 0.6507067669172932, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 15818 + }, + { + "epoch": 0.6507669172932331, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 15819 + }, + { + "epoch": 0.650827067669173, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 15820 + }, + { + "epoch": 0.6508872180451127, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 15821 + }, + { + "epoch": 0.6509473684210526, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2134, + "step": 15822 + }, + { + "epoch": 0.6510075187969925, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 15823 + }, + { + "epoch": 0.6510676691729324, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1766, + "step": 15824 + }, + { + "epoch": 0.6511278195488722, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1414, + "step": 15825 + }, + { + "epoch": 0.651187969924812, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 15826 + }, + { + "epoch": 0.6512481203007519, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 15827 + }, + { + "epoch": 0.6513082706766917, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 15828 + }, + { + "epoch": 0.6513684210526316, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2215, + "step": 15829 + }, + { + "epoch": 0.6514285714285715, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 15830 + }, + { + "epoch": 0.6514887218045112, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 15831 + }, + { + "epoch": 0.6515488721804511, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 15832 + }, + { + "epoch": 0.651609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 15833 + }, + { + "epoch": 0.6516691729323308, + "grad_norm": 1.3671875, + "learning_rate": 3e-05, + "loss": 1.2013, + "step": 15834 + }, + { + "epoch": 0.6517293233082707, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2107, + "step": 15835 + }, + { + "epoch": 0.6517894736842105, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 15836 + }, + { + "epoch": 0.6518496240601503, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 15837 + }, + { + "epoch": 0.6519097744360902, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 15838 + }, + { + "epoch": 0.6519699248120301, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 15839 + }, + { + "epoch": 0.65203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 15840 + }, + { + "epoch": 0.6520902255639097, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 15841 + }, + { + "epoch": 0.6521503759398496, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.2034, + "step": 15842 + }, + { + "epoch": 0.6522105263157895, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 15843 + }, + { + "epoch": 0.6522706766917293, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 15844 + }, + { + "epoch": 0.6523308270676692, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 15845 + }, + { + "epoch": 0.6523909774436091, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 15846 + }, + { + "epoch": 0.6524511278195488, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1836, + "step": 15847 + }, + { + "epoch": 0.6525112781954887, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 15848 + }, + { + "epoch": 0.6525714285714286, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1766, + "step": 15849 + }, + { + "epoch": 0.6526315789473685, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1817, + "step": 15850 + }, + { + "epoch": 0.6526917293233083, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 15851 + }, + { + "epoch": 0.6527518796992481, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 15852 + }, + { + "epoch": 0.652812030075188, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 15853 + }, + { + "epoch": 0.6528721804511278, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 15854 + }, + { + "epoch": 0.6529323308270677, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 15855 + }, + { + "epoch": 0.6529924812030076, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1721, + "step": 15856 + }, + { + "epoch": 0.6530526315789473, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 15857 + }, + { + "epoch": 0.6531127819548872, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 15858 + }, + { + "epoch": 0.6531729323308271, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 15859 + }, + { + "epoch": 0.6532330827067669, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.2902, + "step": 15860 + }, + { + "epoch": 0.6532932330827068, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.155, + "step": 15861 + }, + { + "epoch": 0.6533533834586466, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 15862 + }, + { + "epoch": 0.6534135338345864, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 15863 + }, + { + "epoch": 0.6534736842105263, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 15864 + }, + { + "epoch": 0.6535338345864662, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 15865 + }, + { + "epoch": 0.6535939849624061, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 15866 + }, + { + "epoch": 0.6536541353383458, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 15867 + }, + { + "epoch": 0.6537142857142857, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 15868 + }, + { + "epoch": 0.6537744360902256, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 15869 + }, + { + "epoch": 0.6538345864661654, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 15870 + }, + { + "epoch": 0.6538947368421053, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.4905, + "step": 15871 + }, + { + "epoch": 0.6539548872180451, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 15872 + }, + { + "epoch": 0.6540150375939849, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 15873 + }, + { + "epoch": 0.6540751879699248, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 15874 + }, + { + "epoch": 0.6541353383458647, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 15875 + }, + { + "epoch": 0.6541954887218046, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1124, + "step": 15876 + }, + { + "epoch": 0.6542556390977443, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 15877 + }, + { + "epoch": 0.6543157894736842, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 15878 + }, + { + "epoch": 0.654375939849624, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 15879 + }, + { + "epoch": 0.6544360902255639, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 15880 + }, + { + "epoch": 0.6544962406015038, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 15881 + }, + { + "epoch": 0.6545563909774436, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.2361, + "step": 15882 + }, + { + "epoch": 0.6546165413533834, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 15883 + }, + { + "epoch": 0.6546766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 15884 + }, + { + "epoch": 0.6547368421052632, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 15885 + }, + { + "epoch": 0.654796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 15886 + }, + { + "epoch": 0.6548571428571428, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 15887 + }, + { + "epoch": 0.6549172932330827, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 15888 + }, + { + "epoch": 0.6549774436090225, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 15889 + }, + { + "epoch": 0.6550375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 15890 + }, + { + "epoch": 0.6550977443609023, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 15891 + }, + { + "epoch": 0.655157894736842, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 15892 + }, + { + "epoch": 0.6552180451127819, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 15893 + }, + { + "epoch": 0.6552781954887218, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 15894 + }, + { + "epoch": 0.6553383458646617, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 15895 + }, + { + "epoch": 0.6553984962406015, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 15896 + }, + { + "epoch": 0.6554586466165414, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2216, + "step": 15897 + }, + { + "epoch": 0.6555187969924812, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 15898 + }, + { + "epoch": 0.655578947368421, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 15899 + }, + { + "epoch": 0.6556390977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1917, + "step": 15900 + }, + { + "epoch": 0.6556992481203008, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 15901 + }, + { + "epoch": 0.6557593984962407, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 15902 + }, + { + "epoch": 0.6558195488721804, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 15903 + }, + { + "epoch": 0.6558796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1868, + "step": 15904 + }, + { + "epoch": 0.6559398496240602, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 15905 + }, + { + "epoch": 0.656, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 15906 + }, + { + "epoch": 0.6560601503759399, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 15907 + }, + { + "epoch": 0.6561203007518797, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 15908 + }, + { + "epoch": 0.6561804511278195, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 15909 + }, + { + "epoch": 0.6562406015037594, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 15910 + }, + { + "epoch": 0.6563007518796993, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 15911 + }, + { + "epoch": 0.6563609022556391, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2337, + "step": 15912 + }, + { + "epoch": 0.6564210526315789, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 15913 + }, + { + "epoch": 0.6564812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 15914 + }, + { + "epoch": 0.6565413533834586, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 15915 + }, + { + "epoch": 0.6566015037593985, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 15916 + }, + { + "epoch": 0.6566616541353384, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 15917 + }, + { + "epoch": 0.6567218045112782, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 15918 + }, + { + "epoch": 0.656781954887218, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 15919 + }, + { + "epoch": 0.6568421052631579, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2168, + "step": 15920 + }, + { + "epoch": 0.6569022556390978, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1896, + "step": 15921 + }, + { + "epoch": 0.6569624060150376, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 15922 + }, + { + "epoch": 0.6570225563909774, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 15923 + }, + { + "epoch": 0.6570827067669173, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 15924 + }, + { + "epoch": 0.6571428571428571, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 15925 + }, + { + "epoch": 0.657203007518797, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 15926 + }, + { + "epoch": 0.6572631578947369, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 15927 + }, + { + "epoch": 0.6573233082706766, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 15928 + }, + { + "epoch": 0.6573834586466165, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 15929 + }, + { + "epoch": 0.6574436090225564, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 15930 + }, + { + "epoch": 0.6575037593984963, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 15931 + }, + { + "epoch": 0.6575639097744361, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 15932 + }, + { + "epoch": 0.6576240601503759, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 15933 + }, + { + "epoch": 0.6576842105263158, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 15934 + }, + { + "epoch": 0.6577443609022556, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 15935 + }, + { + "epoch": 0.6578045112781955, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 15936 + }, + { + "epoch": 0.6578646616541354, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 15937 + }, + { + "epoch": 0.6579248120300751, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 15938 + }, + { + "epoch": 0.657984962406015, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 15939 + }, + { + "epoch": 0.6580451127819549, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 15940 + }, + { + "epoch": 0.6581052631578947, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.213, + "step": 15941 + }, + { + "epoch": 0.6581654135338346, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 15942 + }, + { + "epoch": 0.6582255639097745, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 15943 + }, + { + "epoch": 0.6582857142857143, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 15944 + }, + { + "epoch": 0.6583458646616541, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 15945 + }, + { + "epoch": 0.658406015037594, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 15946 + }, + { + "epoch": 0.6584661654135339, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 15947 + }, + { + "epoch": 0.6585263157894737, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 15948 + }, + { + "epoch": 0.6585864661654135, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 15949 + }, + { + "epoch": 0.6586466165413534, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 15950 + }, + { + "epoch": 0.6587067669172932, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 15951 + }, + { + "epoch": 0.6587669172932331, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 15952 + }, + { + "epoch": 0.658827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 15953 + }, + { + "epoch": 0.6588872180451127, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 15954 + }, + { + "epoch": 0.6589473684210526, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 15955 + }, + { + "epoch": 0.6590075187969925, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2876, + "step": 15956 + }, + { + "epoch": 0.6590676691729324, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2163, + "step": 15957 + }, + { + "epoch": 0.6591278195488722, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2134, + "step": 15958 + }, + { + "epoch": 0.659187969924812, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2028, + "step": 15959 + }, + { + "epoch": 0.6592481203007519, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2039, + "step": 15960 + }, + { + "epoch": 0.6593082706766917, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2397, + "step": 15961 + }, + { + "epoch": 0.6593684210526316, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 15962 + }, + { + "epoch": 0.6594285714285715, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 15963 + }, + { + "epoch": 0.6594887218045112, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2176, + "step": 15964 + }, + { + "epoch": 0.6595488721804511, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 15965 + }, + { + "epoch": 0.659609022556391, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2284, + "step": 15966 + }, + { + "epoch": 0.6596691729323308, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 15967 + }, + { + "epoch": 0.6597293233082707, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 15968 + }, + { + "epoch": 0.6597894736842105, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2464, + "step": 15969 + }, + { + "epoch": 0.6598496240601504, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 15970 + }, + { + "epoch": 0.6599097744360902, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 15971 + }, + { + "epoch": 0.6599699248120301, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2361, + "step": 15972 + }, + { + "epoch": 0.66003007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 15973 + }, + { + "epoch": 0.6600902255639097, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 15974 + }, + { + "epoch": 0.6601503759398496, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 15975 + }, + { + "epoch": 0.6602105263157895, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 15976 + }, + { + "epoch": 0.6602706766917293, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 15977 + }, + { + "epoch": 0.6603308270676692, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 15978 + }, + { + "epoch": 0.660390977443609, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 15979 + }, + { + "epoch": 0.6604511278195488, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2084, + "step": 15980 + }, + { + "epoch": 0.6605112781954887, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1512, + "step": 15981 + }, + { + "epoch": 0.6605714285714286, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 15982 + }, + { + "epoch": 0.6606315789473685, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1552, + "step": 15983 + }, + { + "epoch": 0.6606917293233082, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2579, + "step": 15984 + }, + { + "epoch": 0.6607518796992481, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 15985 + }, + { + "epoch": 0.660812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 15986 + }, + { + "epoch": 0.6608721804511278, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 15987 + }, + { + "epoch": 0.6609323308270677, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 15988 + }, + { + "epoch": 0.6609924812030075, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 15989 + }, + { + "epoch": 0.6610526315789473, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 15990 + }, + { + "epoch": 0.6611127819548872, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 15991 + }, + { + "epoch": 0.6611729323308271, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1496, + "step": 15992 + }, + { + "epoch": 0.661233082706767, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 15993 + }, + { + "epoch": 0.6612932330827068, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1707, + "step": 15994 + }, + { + "epoch": 0.6613533834586466, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 15995 + }, + { + "epoch": 0.6614135338345865, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 15996 + }, + { + "epoch": 0.6614736842105263, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 15997 + }, + { + "epoch": 0.6615338345864662, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 15998 + }, + { + "epoch": 0.6615939849624061, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 15999 + }, + { + "epoch": 0.6616541353383458, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 16000 + }, + { + "epoch": 0.6617142857142857, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2801, + "step": 16001 + }, + { + "epoch": 0.6617744360902256, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.2425, + "step": 16002 + }, + { + "epoch": 0.6618345864661654, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 16003 + }, + { + "epoch": 0.6618947368421053, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 16004 + }, + { + "epoch": 0.6619548872180451, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 16005 + }, + { + "epoch": 0.662015037593985, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 16006 + }, + { + "epoch": 0.6620751879699248, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 16007 + }, + { + "epoch": 0.6621353383458647, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 16008 + }, + { + "epoch": 0.6621954887218046, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 16009 + }, + { + "epoch": 0.6622556390977443, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 16010 + }, + { + "epoch": 0.6623157894736842, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 16011 + }, + { + "epoch": 0.6623759398496241, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 16012 + }, + { + "epoch": 0.6624360902255639, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1646, + "step": 16013 + }, + { + "epoch": 0.6624962406015038, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1982, + "step": 16014 + }, + { + "epoch": 0.6625563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 16015 + }, + { + "epoch": 0.6626165413533834, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 16016 + }, + { + "epoch": 0.6626766917293233, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 16017 + }, + { + "epoch": 0.6627368421052632, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1801, + "step": 16018 + }, + { + "epoch": 0.662796992481203, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 16019 + }, + { + "epoch": 0.6628571428571428, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 16020 + }, + { + "epoch": 0.6629172932330827, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 16021 + }, + { + "epoch": 0.6629774436090226, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 16022 + }, + { + "epoch": 0.6630375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 16023 + }, + { + "epoch": 0.6630977443609023, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 16024 + }, + { + "epoch": 0.6631578947368421, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 16025 + }, + { + "epoch": 0.6632180451127819, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1626, + "step": 16026 + }, + { + "epoch": 0.6632781954887218, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 16027 + }, + { + "epoch": 0.6633383458646617, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 16028 + }, + { + "epoch": 0.6633984962406015, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 16029 + }, + { + "epoch": 0.6634586466165413, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 16030 + }, + { + "epoch": 0.6635187969924812, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 16031 + }, + { + "epoch": 0.663578947368421, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 16032 + }, + { + "epoch": 0.6636390977443609, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 16033 + }, + { + "epoch": 0.6636992481203008, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 16034 + }, + { + "epoch": 0.6637593984962405, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 16035 + }, + { + "epoch": 0.6638195488721804, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 16036 + }, + { + "epoch": 0.6638796992481203, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 16037 + }, + { + "epoch": 0.6639398496240602, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 16038 + }, + { + "epoch": 0.664, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 16039 + }, + { + "epoch": 0.6640601503759398, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 16040 + }, + { + "epoch": 0.6641203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1643, + "step": 16041 + }, + { + "epoch": 0.6641804511278195, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 16042 + }, + { + "epoch": 0.6642406015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 16043 + }, + { + "epoch": 0.6643007518796993, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 16044 + }, + { + "epoch": 0.6643609022556392, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 16045 + }, + { + "epoch": 0.6644210526315789, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 16046 + }, + { + "epoch": 0.6644812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 16047 + }, + { + "epoch": 0.6645413533834587, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 16048 + }, + { + "epoch": 0.6646015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 16049 + }, + { + "epoch": 0.6646616541353384, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 16050 + }, + { + "epoch": 0.6647218045112782, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 16051 + }, + { + "epoch": 0.664781954887218, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 16052 + }, + { + "epoch": 0.6648421052631579, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 16053 + }, + { + "epoch": 0.6649022556390978, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 16054 + }, + { + "epoch": 0.6649624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 16055 + }, + { + "epoch": 0.6650225563909774, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 16056 + }, + { + "epoch": 0.6650827067669173, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 16057 + }, + { + "epoch": 0.6651428571428571, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 16058 + }, + { + "epoch": 0.665203007518797, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 16059 + }, + { + "epoch": 0.6652631578947369, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 16060 + }, + { + "epoch": 0.6653233082706766, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 16061 + }, + { + "epoch": 0.6653834586466165, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 16062 + }, + { + "epoch": 0.6654436090225564, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 16063 + }, + { + "epoch": 0.6655037593984963, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 16064 + }, + { + "epoch": 0.6655639097744361, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 16065 + }, + { + "epoch": 0.6656240601503759, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 16066 + }, + { + "epoch": 0.6656842105263158, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 16067 + }, + { + "epoch": 0.6657443609022556, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 16068 + }, + { + "epoch": 0.6658045112781955, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 16069 + }, + { + "epoch": 0.6658646616541354, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 16070 + }, + { + "epoch": 0.6659248120300751, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1726, + "step": 16071 + }, + { + "epoch": 0.665984962406015, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 16072 + }, + { + "epoch": 0.6660451127819549, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 16073 + }, + { + "epoch": 0.6661052631578948, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.26, + "step": 16074 + }, + { + "epoch": 0.6661654135338346, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 16075 + }, + { + "epoch": 0.6662255639097744, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 16076 + }, + { + "epoch": 0.6662857142857143, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2414, + "step": 16077 + }, + { + "epoch": 0.6663458646616541, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 16078 + }, + { + "epoch": 0.666406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 16079 + }, + { + "epoch": 0.6664661654135339, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 16080 + }, + { + "epoch": 0.6665263157894736, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 16081 + }, + { + "epoch": 0.6665864661654135, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2927, + "step": 16082 + }, + { + "epoch": 0.6666466165413534, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2631, + "step": 16083 + }, + { + "epoch": 0.6667067669172932, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1738, + "step": 16084 + }, + { + "epoch": 0.6667669172932331, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 16085 + }, + { + "epoch": 0.6668270676691729, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2697, + "step": 16086 + }, + { + "epoch": 0.6668872180451128, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 16087 + }, + { + "epoch": 0.6669473684210526, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 16088 + }, + { + "epoch": 0.6670075187969925, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2155, + "step": 16089 + }, + { + "epoch": 0.6670676691729324, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 16090 + }, + { + "epoch": 0.6671278195488721, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1702, + "step": 16091 + }, + { + "epoch": 0.667187969924812, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 16092 + }, + { + "epoch": 0.6672481203007519, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2291, + "step": 16093 + }, + { + "epoch": 0.6673082706766917, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 16094 + }, + { + "epoch": 0.6673684210526316, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 16095 + }, + { + "epoch": 0.6674285714285715, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 16096 + }, + { + "epoch": 0.6674887218045112, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.136, + "step": 16097 + }, + { + "epoch": 0.6675488721804511, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 16098 + }, + { + "epoch": 0.667609022556391, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 16099 + }, + { + "epoch": 0.6676691729323309, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 16100 + }, + { + "epoch": 0.6677293233082707, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.3553, + "step": 16101 + }, + { + "epoch": 0.6677894736842105, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 16102 + }, + { + "epoch": 0.6678496240601504, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 16103 + }, + { + "epoch": 0.6679097744360902, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 16104 + }, + { + "epoch": 0.6679699248120301, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 16105 + }, + { + "epoch": 0.66803007518797, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.189, + "step": 16106 + }, + { + "epoch": 0.6680902255639097, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1467, + "step": 16107 + }, + { + "epoch": 0.6681503759398496, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.3247, + "step": 16108 + }, + { + "epoch": 0.6682105263157895, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2127, + "step": 16109 + }, + { + "epoch": 0.6682706766917293, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 16110 + }, + { + "epoch": 0.6683308270676692, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.4759, + "step": 16111 + }, + { + "epoch": 0.668390977443609, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1679, + "step": 16112 + }, + { + "epoch": 0.6684511278195489, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 16113 + }, + { + "epoch": 0.6685112781954887, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 16114 + }, + { + "epoch": 0.6685714285714286, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 16115 + }, + { + "epoch": 0.6686315789473685, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 16116 + }, + { + "epoch": 0.6686917293233082, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 16117 + }, + { + "epoch": 0.6687518796992481, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 16118 + }, + { + "epoch": 0.668812030075188, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1671, + "step": 16119 + }, + { + "epoch": 0.6688721804511278, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2188, + "step": 16120 + }, + { + "epoch": 0.6689323308270677, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 16121 + }, + { + "epoch": 0.6689924812030075, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 16122 + }, + { + "epoch": 0.6690526315789473, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 16123 + }, + { + "epoch": 0.6691127819548872, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 16124 + }, + { + "epoch": 0.6691729323308271, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 16125 + }, + { + "epoch": 0.669233082706767, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 16126 + }, + { + "epoch": 0.6692932330827067, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 16127 + }, + { + "epoch": 0.6693533834586466, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2064, + "step": 16128 + }, + { + "epoch": 0.6694135338345865, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.234, + "step": 16129 + }, + { + "epoch": 0.6694736842105263, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.3012, + "step": 16130 + }, + { + "epoch": 0.6695338345864662, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 16131 + }, + { + "epoch": 0.669593984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 16132 + }, + { + "epoch": 0.6696541353383458, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 16133 + }, + { + "epoch": 0.6697142857142857, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2509, + "step": 16134 + }, + { + "epoch": 0.6697744360902256, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1841, + "step": 16135 + }, + { + "epoch": 0.6698345864661654, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 16136 + }, + { + "epoch": 0.6698947368421052, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 16137 + }, + { + "epoch": 0.6699548872180451, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 16138 + }, + { + "epoch": 0.670015037593985, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 16139 + }, + { + "epoch": 0.6700751879699248, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 16140 + }, + { + "epoch": 0.6701353383458647, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2544, + "step": 16141 + }, + { + "epoch": 0.6701954887218046, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 16142 + }, + { + "epoch": 0.6702556390977443, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 16143 + }, + { + "epoch": 0.6703157894736842, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.3077, + "step": 16144 + }, + { + "epoch": 0.6703759398496241, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2162, + "step": 16145 + }, + { + "epoch": 0.6704360902255639, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 16146 + }, + { + "epoch": 0.6704962406015038, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 16147 + }, + { + "epoch": 0.6705563909774436, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 16148 + }, + { + "epoch": 0.6706165413533834, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 16149 + }, + { + "epoch": 0.6706766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 16150 + }, + { + "epoch": 0.6707368421052632, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 16151 + }, + { + "epoch": 0.6707969924812031, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 16152 + }, + { + "epoch": 0.6708571428571428, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1646, + "step": 16153 + }, + { + "epoch": 0.6709172932330827, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 16154 + }, + { + "epoch": 0.6709774436090226, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 16155 + }, + { + "epoch": 0.6710375939849624, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 16156 + }, + { + "epoch": 0.6710977443609023, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 16157 + }, + { + "epoch": 0.6711578947368421, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1721, + "step": 16158 + }, + { + "epoch": 0.6712180451127819, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2431, + "step": 16159 + }, + { + "epoch": 0.6712781954887218, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 16160 + }, + { + "epoch": 0.6713383458646617, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 16161 + }, + { + "epoch": 0.6713984962406015, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 16162 + }, + { + "epoch": 0.6714586466165413, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 16163 + }, + { + "epoch": 0.6715187969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 16164 + }, + { + "epoch": 0.671578947368421, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 16165 + }, + { + "epoch": 0.6716390977443609, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 16166 + }, + { + "epoch": 0.6716992481203008, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 16167 + }, + { + "epoch": 0.6717593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 16168 + }, + { + "epoch": 0.6718195488721804, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 16169 + }, + { + "epoch": 0.6718796992481203, + "grad_norm": 2.046875, + "learning_rate": 3e-05, + "loss": 1.26, + "step": 16170 + }, + { + "epoch": 0.6719398496240602, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2093, + "step": 16171 + }, + { + "epoch": 0.672, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1708, + "step": 16172 + }, + { + "epoch": 0.6720601503759398, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 16173 + }, + { + "epoch": 0.6721203007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 16174 + }, + { + "epoch": 0.6721804511278195, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 16175 + }, + { + "epoch": 0.6722406015037594, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 16176 + }, + { + "epoch": 0.6723007518796993, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 16177 + }, + { + "epoch": 0.672360902255639, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 16178 + }, + { + "epoch": 0.6724210526315789, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2067, + "step": 16179 + }, + { + "epoch": 0.6724812030075188, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 16180 + }, + { + "epoch": 0.6725413533834587, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 16181 + }, + { + "epoch": 0.6726015037593985, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1545, + "step": 16182 + }, + { + "epoch": 0.6726616541353383, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 16183 + }, + { + "epoch": 0.6727218045112782, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 16184 + }, + { + "epoch": 0.672781954887218, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 16185 + }, + { + "epoch": 0.6728421052631579, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 16186 + }, + { + "epoch": 0.6729022556390978, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 16187 + }, + { + "epoch": 0.6729624060150375, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 16188 + }, + { + "epoch": 0.6730225563909774, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 16189 + }, + { + "epoch": 0.6730827067669173, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.207, + "step": 16190 + }, + { + "epoch": 0.6731428571428572, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 16191 + }, + { + "epoch": 0.673203007518797, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 16192 + }, + { + "epoch": 0.6732631578947369, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 16193 + }, + { + "epoch": 0.6733233082706767, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 16194 + }, + { + "epoch": 0.6733834586466165, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 16195 + }, + { + "epoch": 0.6734436090225564, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 16196 + }, + { + "epoch": 0.6735037593984963, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2083, + "step": 16197 + }, + { + "epoch": 0.6735639097744361, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 16198 + }, + { + "epoch": 0.6736240601503759, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 16199 + }, + { + "epoch": 0.6736842105263158, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 16200 + }, + { + "epoch": 0.6737443609022556, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 16201 + }, + { + "epoch": 0.6738045112781955, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 16202 + }, + { + "epoch": 0.6738646616541354, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 16203 + }, + { + "epoch": 0.6739248120300751, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 16204 + }, + { + "epoch": 0.673984962406015, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 16205 + }, + { + "epoch": 0.6740451127819549, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 16206 + }, + { + "epoch": 0.6741052631578948, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 16207 + }, + { + "epoch": 0.6741654135338346, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 16208 + }, + { + "epoch": 0.6742255639097744, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1641, + "step": 16209 + }, + { + "epoch": 0.6742857142857143, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.236, + "step": 16210 + }, + { + "epoch": 0.6743458646616541, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 16211 + }, + { + "epoch": 0.674406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2435, + "step": 16212 + }, + { + "epoch": 0.6744661654135339, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 16213 + }, + { + "epoch": 0.6745263157894736, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.2554, + "step": 16214 + }, + { + "epoch": 0.6745864661654135, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1791, + "step": 16215 + }, + { + "epoch": 0.6746466165413534, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 16216 + }, + { + "epoch": 0.6747067669172933, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2678, + "step": 16217 + }, + { + "epoch": 0.6747669172932331, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 16218 + }, + { + "epoch": 0.6748270676691729, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 16219 + }, + { + "epoch": 0.6748872180451128, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 16220 + }, + { + "epoch": 0.6749473684210526, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 16221 + }, + { + "epoch": 0.6750075187969925, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 16222 + }, + { + "epoch": 0.6750676691729324, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 16223 + }, + { + "epoch": 0.6751278195488721, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 16224 + }, + { + "epoch": 0.675187969924812, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1965, + "step": 16225 + }, + { + "epoch": 0.6752481203007519, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 16226 + }, + { + "epoch": 0.6753082706766917, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 16227 + }, + { + "epoch": 0.6753684210526316, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2154, + "step": 16228 + }, + { + "epoch": 0.6754285714285714, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 16229 + }, + { + "epoch": 0.6754887218045112, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 16230 + }, + { + "epoch": 0.6755488721804511, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1688, + "step": 16231 + }, + { + "epoch": 0.675609022556391, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 16232 + }, + { + "epoch": 0.6756691729323309, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 16233 + }, + { + "epoch": 0.6757293233082706, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 16234 + }, + { + "epoch": 0.6757894736842105, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 16235 + }, + { + "epoch": 0.6758496240601504, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2595, + "step": 16236 + }, + { + "epoch": 0.6759097744360902, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 16237 + }, + { + "epoch": 0.6759699248120301, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 16238 + }, + { + "epoch": 0.6760300751879699, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 16239 + }, + { + "epoch": 0.6760902255639097, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 16240 + }, + { + "epoch": 0.6761503759398496, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 16241 + }, + { + "epoch": 0.6762105263157895, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2294, + "step": 16242 + }, + { + "epoch": 0.6762706766917294, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 16243 + }, + { + "epoch": 0.6763308270676692, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 16244 + }, + { + "epoch": 0.676390977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 16245 + }, + { + "epoch": 0.6764511278195489, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 16246 + }, + { + "epoch": 0.6765112781954887, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 16247 + }, + { + "epoch": 0.6765714285714286, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 16248 + }, + { + "epoch": 0.6766315789473685, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 16249 + }, + { + "epoch": 0.6766917293233082, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2251, + "step": 16250 + }, + { + "epoch": 0.6767518796992481, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 16251 + }, + { + "epoch": 0.676812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 16252 + }, + { + "epoch": 0.6768721804511278, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 16253 + }, + { + "epoch": 0.6769323308270677, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 16254 + }, + { + "epoch": 0.6769924812030075, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 16255 + }, + { + "epoch": 0.6770526315789474, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 16256 + }, + { + "epoch": 0.6771127819548872, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 16257 + }, + { + "epoch": 0.6771729323308271, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 16258 + }, + { + "epoch": 0.677233082706767, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 16259 + }, + { + "epoch": 0.6772932330827067, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 16260 + }, + { + "epoch": 0.6773533834586466, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2399, + "step": 16261 + }, + { + "epoch": 0.6774135338345865, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 16262 + }, + { + "epoch": 0.6774736842105263, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 16263 + }, + { + "epoch": 0.6775338345864662, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 16264 + }, + { + "epoch": 0.677593984962406, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2252, + "step": 16265 + }, + { + "epoch": 0.6776541353383458, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 16266 + }, + { + "epoch": 0.6777142857142857, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 16267 + }, + { + "epoch": 0.6777744360902256, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 16268 + }, + { + "epoch": 0.6778345864661655, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 16269 + }, + { + "epoch": 0.6778947368421052, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2349, + "step": 16270 + }, + { + "epoch": 0.6779548872180451, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 16271 + }, + { + "epoch": 0.678015037593985, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1908, + "step": 16272 + }, + { + "epoch": 0.6780751879699248, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 16273 + }, + { + "epoch": 0.6781353383458647, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 16274 + }, + { + "epoch": 0.6781954887218045, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 16275 + }, + { + "epoch": 0.6782556390977443, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 16276 + }, + { + "epoch": 0.6783157894736842, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 16277 + }, + { + "epoch": 0.6783759398496241, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 16278 + }, + { + "epoch": 0.678436090225564, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 16279 + }, + { + "epoch": 0.6784962406015037, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.3893, + "step": 16280 + }, + { + "epoch": 0.6785563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 16281 + }, + { + "epoch": 0.6786165413533835, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 16282 + }, + { + "epoch": 0.6786766917293233, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 16283 + }, + { + "epoch": 0.6787368421052632, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 16284 + }, + { + "epoch": 0.678796992481203, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1543, + "step": 16285 + }, + { + "epoch": 0.6788571428571428, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 16286 + }, + { + "epoch": 0.6789172932330827, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 16287 + }, + { + "epoch": 0.6789774436090226, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 16288 + }, + { + "epoch": 0.6790375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 16289 + }, + { + "epoch": 0.6790977443609022, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2464, + "step": 16290 + }, + { + "epoch": 0.6791578947368421, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 16291 + }, + { + "epoch": 0.6792180451127819, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 16292 + }, + { + "epoch": 0.6792781954887218, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 16293 + }, + { + "epoch": 0.6793383458646617, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 16294 + }, + { + "epoch": 0.6793984962406016, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 16295 + }, + { + "epoch": 0.6794586466165413, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2559, + "step": 16296 + }, + { + "epoch": 0.6795187969924812, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 16297 + }, + { + "epoch": 0.6795789473684211, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 16298 + }, + { + "epoch": 0.6796390977443609, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.3534, + "step": 16299 + }, + { + "epoch": 0.6796992481203008, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 16300 + }, + { + "epoch": 0.6797593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 16301 + }, + { + "epoch": 0.6798195488721804, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 16302 + }, + { + "epoch": 0.6798796992481203, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 16303 + }, + { + "epoch": 0.6799398496240602, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1791, + "step": 16304 + }, + { + "epoch": 0.68, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2166, + "step": 16305 + }, + { + "epoch": 0.6800601503759398, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3114, + "step": 16306 + }, + { + "epoch": 0.6801203007518797, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 16307 + }, + { + "epoch": 0.6801804511278196, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 16308 + }, + { + "epoch": 0.6802406015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 16309 + }, + { + "epoch": 0.6803007518796993, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 16310 + }, + { + "epoch": 0.680360902255639, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 16311 + }, + { + "epoch": 0.6804210526315789, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 16312 + }, + { + "epoch": 0.6804812030075188, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 16313 + }, + { + "epoch": 0.6805413533834587, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 16314 + }, + { + "epoch": 0.6806015037593985, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 16315 + }, + { + "epoch": 0.6806616541353383, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 16316 + }, + { + "epoch": 0.6807218045112782, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 16317 + }, + { + "epoch": 0.680781954887218, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 16318 + }, + { + "epoch": 0.6808421052631579, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 16319 + }, + { + "epoch": 0.6809022556390978, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 16320 + }, + { + "epoch": 0.6809624060150375, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2632, + "step": 16321 + }, + { + "epoch": 0.6810225563909774, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 16322 + }, + { + "epoch": 0.6810827067669173, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2046, + "step": 16323 + }, + { + "epoch": 0.6811428571428572, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 16324 + }, + { + "epoch": 0.681203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1774, + "step": 16325 + }, + { + "epoch": 0.6812631578947368, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 16326 + }, + { + "epoch": 0.6813233082706767, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 16327 + }, + { + "epoch": 0.6813834586466165, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 16328 + }, + { + "epoch": 0.6814436090225564, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 16329 + }, + { + "epoch": 0.6815037593984963, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 16330 + }, + { + "epoch": 0.681563909774436, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 16331 + }, + { + "epoch": 0.6816240601503759, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 16332 + }, + { + "epoch": 0.6816842105263158, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 16333 + }, + { + "epoch": 0.6817443609022557, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 16334 + }, + { + "epoch": 0.6818045112781955, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 16335 + }, + { + "epoch": 0.6818646616541353, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 16336 + }, + { + "epoch": 0.6819248120300752, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.16, + "step": 16337 + }, + { + "epoch": 0.681984962406015, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2645, + "step": 16338 + }, + { + "epoch": 0.6820451127819549, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 16339 + }, + { + "epoch": 0.6821052631578948, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 16340 + }, + { + "epoch": 0.6821654135338346, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 16341 + }, + { + "epoch": 0.6822255639097744, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1436, + "step": 16342 + }, + { + "epoch": 0.6822857142857143, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 16343 + }, + { + "epoch": 0.6823458646616541, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 16344 + }, + { + "epoch": 0.682406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 16345 + }, + { + "epoch": 0.6824661654135339, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 16346 + }, + { + "epoch": 0.6825263157894736, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 16347 + }, + { + "epoch": 0.6825864661654135, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 16348 + }, + { + "epoch": 0.6826466165413534, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 16349 + }, + { + "epoch": 0.6827067669172933, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 16350 + }, + { + "epoch": 0.6827669172932331, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 16351 + }, + { + "epoch": 0.6828270676691729, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 16352 + }, + { + "epoch": 0.6828872180451128, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 16353 + }, + { + "epoch": 0.6829473684210526, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2313, + "step": 16354 + }, + { + "epoch": 0.6830075187969925, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 16355 + }, + { + "epoch": 0.6830676691729324, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 16356 + }, + { + "epoch": 0.6831278195488721, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 16357 + }, + { + "epoch": 0.683187969924812, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 16358 + }, + { + "epoch": 0.6832481203007519, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 16359 + }, + { + "epoch": 0.6833082706766918, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1814, + "step": 16360 + }, + { + "epoch": 0.6833684210526316, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 16361 + }, + { + "epoch": 0.6834285714285714, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 16362 + }, + { + "epoch": 0.6834887218045113, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1988, + "step": 16363 + }, + { + "epoch": 0.6835488721804511, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2398, + "step": 16364 + }, + { + "epoch": 0.683609022556391, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 16365 + }, + { + "epoch": 0.6836691729323309, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 16366 + }, + { + "epoch": 0.6837293233082706, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 16367 + }, + { + "epoch": 0.6837894736842105, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 16368 + }, + { + "epoch": 0.6838496240601504, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 16369 + }, + { + "epoch": 0.6839097744360902, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1911, + "step": 16370 + }, + { + "epoch": 0.6839699248120301, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 16371 + }, + { + "epoch": 0.6840300751879699, + "grad_norm": 2.09375, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 16372 + }, + { + "epoch": 0.6840902255639097, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 16373 + }, + { + "epoch": 0.6841503759398496, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1692, + "step": 16374 + }, + { + "epoch": 0.6842105263157895, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1996, + "step": 16375 + }, + { + "epoch": 0.6842706766917294, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.16, + "step": 16376 + }, + { + "epoch": 0.6843308270676691, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 16377 + }, + { + "epoch": 0.684390977443609, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 16378 + }, + { + "epoch": 0.6844511278195489, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 16379 + }, + { + "epoch": 0.6845112781954887, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2294, + "step": 16380 + }, + { + "epoch": 0.6845714285714286, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 16381 + }, + { + "epoch": 0.6846315789473684, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 16382 + }, + { + "epoch": 0.6846917293233082, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 16383 + }, + { + "epoch": 0.6847518796992481, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 16384 + }, + { + "epoch": 0.684812030075188, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2143, + "step": 16385 + }, + { + "epoch": 0.6848721804511279, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 16386 + }, + { + "epoch": 0.6849323308270676, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 16387 + }, + { + "epoch": 0.6849924812030075, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 16388 + }, + { + "epoch": 0.6850526315789474, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1333, + "step": 16389 + }, + { + "epoch": 0.6851127819548872, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 16390 + }, + { + "epoch": 0.6851729323308271, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 16391 + }, + { + "epoch": 0.685233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 16392 + }, + { + "epoch": 0.6852932330827067, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3263, + "step": 16393 + }, + { + "epoch": 0.6853533834586466, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 16394 + }, + { + "epoch": 0.6854135338345865, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 16395 + }, + { + "epoch": 0.6854736842105263, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 16396 + }, + { + "epoch": 0.6855338345864662, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1793, + "step": 16397 + }, + { + "epoch": 0.685593984962406, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 16398 + }, + { + "epoch": 0.6856541353383458, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 16399 + }, + { + "epoch": 0.6857142857142857, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 16400 + }, + { + "epoch": 0.6857744360902256, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 16401 + }, + { + "epoch": 0.6858345864661655, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.2509, + "step": 16402 + }, + { + "epoch": 0.6858947368421052, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.16, + "step": 16403 + }, + { + "epoch": 0.6859548872180451, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 16404 + }, + { + "epoch": 0.686015037593985, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 16405 + }, + { + "epoch": 0.6860751879699248, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 16406 + }, + { + "epoch": 0.6861353383458647, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 16407 + }, + { + "epoch": 0.6861954887218045, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 16408 + }, + { + "epoch": 0.6862556390977443, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1989, + "step": 16409 + }, + { + "epoch": 0.6863157894736842, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 16410 + }, + { + "epoch": 0.6863759398496241, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 16411 + }, + { + "epoch": 0.686436090225564, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2336, + "step": 16412 + }, + { + "epoch": 0.6864962406015037, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1781, + "step": 16413 + }, + { + "epoch": 0.6865563909774436, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 16414 + }, + { + "epoch": 0.6866165413533835, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 16415 + }, + { + "epoch": 0.6866766917293233, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2058, + "step": 16416 + }, + { + "epoch": 0.6867368421052632, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 16417 + }, + { + "epoch": 0.686796992481203, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 16418 + }, + { + "epoch": 0.6868571428571428, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 16419 + }, + { + "epoch": 0.6869172932330827, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 16420 + }, + { + "epoch": 0.6869774436090226, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 16421 + }, + { + "epoch": 0.6870375939849624, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 16422 + }, + { + "epoch": 0.6870977443609022, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 16423 + }, + { + "epoch": 0.6871578947368421, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 16424 + }, + { + "epoch": 0.687218045112782, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 16425 + }, + { + "epoch": 0.6872781954887218, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.207, + "step": 16426 + }, + { + "epoch": 0.6873383458646617, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 16427 + }, + { + "epoch": 0.6873984962406015, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 16428 + }, + { + "epoch": 0.6874586466165413, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 16429 + }, + { + "epoch": 0.6875187969924812, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 16430 + }, + { + "epoch": 0.6875789473684211, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 16431 + }, + { + "epoch": 0.6876390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 16432 + }, + { + "epoch": 0.6876992481203007, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 16433 + }, + { + "epoch": 0.6877593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 16434 + }, + { + "epoch": 0.6878195488721804, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2067, + "step": 16435 + }, + { + "epoch": 0.6878796992481203, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 16436 + }, + { + "epoch": 0.6879398496240602, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 16437 + }, + { + "epoch": 0.688, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 16438 + }, + { + "epoch": 0.6880601503759398, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 16439 + }, + { + "epoch": 0.6881203007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 16440 + }, + { + "epoch": 0.6881804511278196, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 16441 + }, + { + "epoch": 0.6882406015037594, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 16442 + }, + { + "epoch": 0.6883007518796993, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 16443 + }, + { + "epoch": 0.6883609022556391, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 16444 + }, + { + "epoch": 0.6884210526315789, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 16445 + }, + { + "epoch": 0.6884812030075188, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 16446 + }, + { + "epoch": 0.6885413533834587, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 16447 + }, + { + "epoch": 0.6886015037593985, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1754, + "step": 16448 + }, + { + "epoch": 0.6886616541353383, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2152, + "step": 16449 + }, + { + "epoch": 0.6887218045112782, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 16450 + }, + { + "epoch": 0.688781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 16451 + }, + { + "epoch": 0.6888421052631579, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 16452 + }, + { + "epoch": 0.6889022556390978, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2188, + "step": 16453 + }, + { + "epoch": 0.6889624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 16454 + }, + { + "epoch": 0.6890225563909774, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 16455 + }, + { + "epoch": 0.6890827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2261, + "step": 16456 + }, + { + "epoch": 0.6891428571428572, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 16457 + }, + { + "epoch": 0.689203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 16458 + }, + { + "epoch": 0.6892631578947368, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 16459 + }, + { + "epoch": 0.6893233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 16460 + }, + { + "epoch": 0.6893834586466165, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9302, + "step": 16461 + }, + { + "epoch": 0.6894436090225564, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 16462 + }, + { + "epoch": 0.6895037593984963, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 16463 + }, + { + "epoch": 0.689563909774436, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 16464 + }, + { + "epoch": 0.6896240601503759, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 16465 + }, + { + "epoch": 0.6896842105263158, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 16466 + }, + { + "epoch": 0.6897443609022557, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 16467 + }, + { + "epoch": 0.6898045112781955, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.3234, + "step": 16468 + }, + { + "epoch": 0.6898646616541353, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.5134, + "step": 16469 + }, + { + "epoch": 0.6899248120300752, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 16470 + }, + { + "epoch": 0.689984962406015, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 16471 + }, + { + "epoch": 0.6900451127819549, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 16472 + }, + { + "epoch": 0.6901052631578948, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 16473 + }, + { + "epoch": 0.6901654135338345, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 16474 + }, + { + "epoch": 0.6902255639097744, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 16475 + }, + { + "epoch": 0.6902857142857143, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 16476 + }, + { + "epoch": 0.6903458646616542, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 16477 + }, + { + "epoch": 0.690406015037594, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 16478 + }, + { + "epoch": 0.6904661654135338, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 16479 + }, + { + "epoch": 0.6905263157894737, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 16480 + }, + { + "epoch": 0.6905864661654135, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 16481 + }, + { + "epoch": 0.6906466165413534, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1891, + "step": 16482 + }, + { + "epoch": 0.6907067669172933, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 16483 + }, + { + "epoch": 0.690766917293233, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 16484 + }, + { + "epoch": 0.6908270676691729, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 16485 + }, + { + "epoch": 0.6908872180451128, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 16486 + }, + { + "epoch": 0.6909473684210526, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 16487 + }, + { + "epoch": 0.6910075187969925, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 16488 + }, + { + "epoch": 0.6910676691729324, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 16489 + }, + { + "epoch": 0.6911278195488721, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2571, + "step": 16490 + }, + { + "epoch": 0.691187969924812, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 16491 + }, + { + "epoch": 0.6912481203007519, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 16492 + }, + { + "epoch": 0.6913082706766918, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1923, + "step": 16493 + }, + { + "epoch": 0.6913684210526316, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 16494 + }, + { + "epoch": 0.6914285714285714, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 16495 + }, + { + "epoch": 0.6914887218045113, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 16496 + }, + { + "epoch": 0.6915488721804511, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 16497 + }, + { + "epoch": 0.691609022556391, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 16498 + }, + { + "epoch": 0.6916691729323309, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 16499 + }, + { + "epoch": 0.6917293233082706, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 16500 + }, + { + "epoch": 0.6917894736842105, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2471, + "step": 16501 + }, + { + "epoch": 0.6918496240601504, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 16502 + }, + { + "epoch": 0.6919097744360903, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 16503 + }, + { + "epoch": 0.6919699248120301, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 16504 + }, + { + "epoch": 0.6920300751879699, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 16505 + }, + { + "epoch": 0.6920902255639098, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 16506 + }, + { + "epoch": 0.6921503759398496, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 16507 + }, + { + "epoch": 0.6922105263157895, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 16508 + }, + { + "epoch": 0.6922706766917294, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 16509 + }, + { + "epoch": 0.6923308270676691, + "grad_norm": 1.40625, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 16510 + }, + { + "epoch": 0.692390977443609, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 16511 + }, + { + "epoch": 0.6924511278195489, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 16512 + }, + { + "epoch": 0.6925112781954887, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 16513 + }, + { + "epoch": 0.6925714285714286, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 16514 + }, + { + "epoch": 0.6926315789473684, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 16515 + }, + { + "epoch": 0.6926917293233082, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 16516 + }, + { + "epoch": 0.6927518796992481, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 16517 + }, + { + "epoch": 0.692812030075188, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 16518 + }, + { + "epoch": 0.6928721804511279, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.3182, + "step": 16519 + }, + { + "epoch": 0.6929323308270676, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.2065, + "step": 16520 + }, + { + "epoch": 0.6929924812030075, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 16521 + }, + { + "epoch": 0.6930526315789474, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 16522 + }, + { + "epoch": 0.6931127819548872, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 16523 + }, + { + "epoch": 0.6931729323308271, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 16524 + }, + { + "epoch": 0.6932330827067669, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1537, + "step": 16525 + }, + { + "epoch": 0.6932932330827067, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 16526 + }, + { + "epoch": 0.6933533834586466, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 16527 + }, + { + "epoch": 0.6934135338345865, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1929, + "step": 16528 + }, + { + "epoch": 0.6934736842105264, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 16529 + }, + { + "epoch": 0.6935338345864661, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 16530 + }, + { + "epoch": 0.693593984962406, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 16531 + }, + { + "epoch": 0.6936541353383459, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2107, + "step": 16532 + }, + { + "epoch": 0.6937142857142857, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 16533 + }, + { + "epoch": 0.6937744360902256, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 16534 + }, + { + "epoch": 0.6938345864661654, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 16535 + }, + { + "epoch": 0.6938947368421052, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 16536 + }, + { + "epoch": 0.6939548872180451, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 16537 + }, + { + "epoch": 0.694015037593985, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 16538 + }, + { + "epoch": 0.6940751879699248, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 16539 + }, + { + "epoch": 0.6941353383458647, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 16540 + }, + { + "epoch": 0.6941954887218045, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 16541 + }, + { + "epoch": 0.6942556390977443, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 16542 + }, + { + "epoch": 0.6943157894736842, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2359, + "step": 16543 + }, + { + "epoch": 0.6943759398496241, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 16544 + }, + { + "epoch": 0.694436090225564, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 16545 + }, + { + "epoch": 0.6944962406015037, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 16546 + }, + { + "epoch": 0.6945563909774436, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 16547 + }, + { + "epoch": 0.6946165413533835, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 16548 + }, + { + "epoch": 0.6946766917293233, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2607, + "step": 16549 + }, + { + "epoch": 0.6947368421052632, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.233, + "step": 16550 + }, + { + "epoch": 0.694796992481203, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 16551 + }, + { + "epoch": 0.6948571428571428, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1897, + "step": 16552 + }, + { + "epoch": 0.6949172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 16553 + }, + { + "epoch": 0.6949774436090226, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 16554 + }, + { + "epoch": 0.6950375939849625, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 16555 + }, + { + "epoch": 0.6950977443609022, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 16556 + }, + { + "epoch": 0.6951578947368421, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 16557 + }, + { + "epoch": 0.695218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 16558 + }, + { + "epoch": 0.6952781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1537, + "step": 16559 + }, + { + "epoch": 0.6953383458646617, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 16560 + }, + { + "epoch": 0.6953984962406015, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 16561 + }, + { + "epoch": 0.6954586466165413, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 16562 + }, + { + "epoch": 0.6955187969924812, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 16563 + }, + { + "epoch": 0.6955789473684211, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 16564 + }, + { + "epoch": 0.695639097744361, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 16565 + }, + { + "epoch": 0.6956992481203007, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2047, + "step": 16566 + }, + { + "epoch": 0.6957593984962406, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 16567 + }, + { + "epoch": 0.6958195488721804, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 16568 + }, + { + "epoch": 0.6958796992481203, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.209, + "step": 16569 + }, + { + "epoch": 0.6959398496240602, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 16570 + }, + { + "epoch": 0.696, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 16571 + }, + { + "epoch": 0.6960601503759398, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2676, + "step": 16572 + }, + { + "epoch": 0.6961203007518797, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 16573 + }, + { + "epoch": 0.6961804511278196, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.3452, + "step": 16574 + }, + { + "epoch": 0.6962406015037594, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 16575 + }, + { + "epoch": 0.6963007518796992, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 16576 + }, + { + "epoch": 0.6963609022556391, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 16577 + }, + { + "epoch": 0.6964210526315789, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 16578 + }, + { + "epoch": 0.6964812030075188, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2082, + "step": 16579 + }, + { + "epoch": 0.6965413533834587, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 16580 + }, + { + "epoch": 0.6966015037593984, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 16581 + }, + { + "epoch": 0.6966616541353383, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 16582 + }, + { + "epoch": 0.6967218045112782, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 16583 + }, + { + "epoch": 0.6967819548872181, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1917, + "step": 16584 + }, + { + "epoch": 0.6968421052631579, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 16585 + }, + { + "epoch": 0.6969022556390977, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 16586 + }, + { + "epoch": 0.6969624060150376, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 16587 + }, + { + "epoch": 0.6970225563909774, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 16588 + }, + { + "epoch": 0.6970827067669173, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 16589 + }, + { + "epoch": 0.6971428571428572, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 16590 + }, + { + "epoch": 0.697203007518797, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 16591 + }, + { + "epoch": 0.6972631578947368, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 16592 + }, + { + "epoch": 0.6973233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2255, + "step": 16593 + }, + { + "epoch": 0.6973834586466165, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.2237, + "step": 16594 + }, + { + "epoch": 0.6974436090225564, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.148, + "step": 16595 + }, + { + "epoch": 0.6975037593984963, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 16596 + }, + { + "epoch": 0.697563909774436, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 16597 + }, + { + "epoch": 0.6976240601503759, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1641, + "step": 16598 + }, + { + "epoch": 0.6976842105263158, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 16599 + }, + { + "epoch": 0.6977443609022557, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 16600 + }, + { + "epoch": 0.6978045112781955, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 16601 + }, + { + "epoch": 0.6978646616541353, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 16602 + }, + { + "epoch": 0.6979248120300752, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.163, + "step": 16603 + }, + { + "epoch": 0.697984962406015, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 16604 + }, + { + "epoch": 0.6980451127819549, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 16605 + }, + { + "epoch": 0.6981052631578948, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 16606 + }, + { + "epoch": 0.6981654135338345, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 16607 + }, + { + "epoch": 0.6982255639097744, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 16608 + }, + { + "epoch": 0.6982857142857143, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 16609 + }, + { + "epoch": 0.6983458646616542, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 16610 + }, + { + "epoch": 0.698406015037594, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 16611 + }, + { + "epoch": 0.6984661654135338, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 16612 + }, + { + "epoch": 0.6985263157894737, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.4531, + "step": 16613 + }, + { + "epoch": 0.6985864661654135, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 16614 + }, + { + "epoch": 0.6986466165413534, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 16615 + }, + { + "epoch": 0.6987067669172933, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 16616 + }, + { + "epoch": 0.698766917293233, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 16617 + }, + { + "epoch": 0.6988270676691729, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 16618 + }, + { + "epoch": 0.6988872180451128, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 16619 + }, + { + "epoch": 0.6989473684210527, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 16620 + }, + { + "epoch": 0.6990075187969925, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2332, + "step": 16621 + }, + { + "epoch": 0.6990676691729323, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 16622 + }, + { + "epoch": 0.6991278195488722, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 16623 + }, + { + "epoch": 0.699187969924812, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 16624 + }, + { + "epoch": 0.6992481203007519, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 16625 + }, + { + "epoch": 0.6993082706766918, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 16626 + }, + { + "epoch": 0.6993684210526315, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 16627 + }, + { + "epoch": 0.6994285714285714, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2227, + "step": 16628 + }, + { + "epoch": 0.6994887218045113, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1852, + "step": 16629 + }, + { + "epoch": 0.6995488721804511, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 16630 + }, + { + "epoch": 0.699609022556391, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 16631 + }, + { + "epoch": 0.6996691729323308, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2423, + "step": 16632 + }, + { + "epoch": 0.6997293233082706, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 16633 + }, + { + "epoch": 0.6997894736842105, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2734, + "step": 16634 + }, + { + "epoch": 0.6998496240601504, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 16635 + }, + { + "epoch": 0.6999097744360903, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 16636 + }, + { + "epoch": 0.69996992481203, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 16637 + }, + { + "epoch": 0.7000300751879699, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 16638 + }, + { + "epoch": 0.7000902255639098, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 16639 + }, + { + "epoch": 0.7001503759398496, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 16640 + }, + { + "epoch": 0.7002105263157895, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 16641 + }, + { + "epoch": 0.7002706766917294, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 16642 + }, + { + "epoch": 0.7003308270676691, + "grad_norm": 1.4921875, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 16643 + }, + { + "epoch": 0.700390977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 16644 + }, + { + "epoch": 0.7004511278195489, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2115, + "step": 16645 + }, + { + "epoch": 0.7005112781954888, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 16646 + }, + { + "epoch": 0.7005714285714286, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 16647 + }, + { + "epoch": 0.7006315789473684, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 16648 + }, + { + "epoch": 0.7006917293233083, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 16649 + }, + { + "epoch": 0.7007518796992481, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 16650 + }, + { + "epoch": 0.700812030075188, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 16651 + }, + { + "epoch": 0.7008721804511279, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 16652 + }, + { + "epoch": 0.7009323308270676, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 16653 + }, + { + "epoch": 0.7009924812030075, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 16654 + }, + { + "epoch": 0.7010526315789474, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 16655 + }, + { + "epoch": 0.7011127819548872, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 16656 + }, + { + "epoch": 0.7011729323308271, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 16657 + }, + { + "epoch": 0.7012330827067669, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 16658 + }, + { + "epoch": 0.7012932330827067, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 16659 + }, + { + "epoch": 0.7013533834586466, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 16660 + }, + { + "epoch": 0.7014135338345865, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 16661 + }, + { + "epoch": 0.7014736842105264, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 16662 + }, + { + "epoch": 0.7015338345864661, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.3142, + "step": 16663 + }, + { + "epoch": 0.701593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 16664 + }, + { + "epoch": 0.7016541353383459, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1861, + "step": 16665 + }, + { + "epoch": 0.7017142857142857, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 16666 + }, + { + "epoch": 0.7017744360902256, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.2386, + "step": 16667 + }, + { + "epoch": 0.7018345864661654, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.2228, + "step": 16668 + }, + { + "epoch": 0.7018947368421052, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 16669 + }, + { + "epoch": 0.7019548872180451, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 16670 + }, + { + "epoch": 0.702015037593985, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 16671 + }, + { + "epoch": 0.7020751879699249, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 16672 + }, + { + "epoch": 0.7021353383458646, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 16673 + }, + { + "epoch": 0.7021954887218045, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 16674 + }, + { + "epoch": 0.7022556390977444, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 16675 + }, + { + "epoch": 0.7023157894736842, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 16676 + }, + { + "epoch": 0.7023759398496241, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2156, + "step": 16677 + }, + { + "epoch": 0.7024360902255639, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 16678 + }, + { + "epoch": 0.7024962406015037, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 16679 + }, + { + "epoch": 0.7025563909774436, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 16680 + }, + { + "epoch": 0.7026165413533835, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 16681 + }, + { + "epoch": 0.7026766917293233, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 16682 + }, + { + "epoch": 0.7027368421052631, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2121, + "step": 16683 + }, + { + "epoch": 0.702796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 16684 + }, + { + "epoch": 0.7028571428571428, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 16685 + }, + { + "epoch": 0.7029172932330827, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2041, + "step": 16686 + }, + { + "epoch": 0.7029774436090226, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 16687 + }, + { + "epoch": 0.7030375939849625, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 16688 + }, + { + "epoch": 0.7030977443609022, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 16689 + }, + { + "epoch": 0.7031578947368421, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 16690 + }, + { + "epoch": 0.703218045112782, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 16691 + }, + { + "epoch": 0.7032781954887218, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 16692 + }, + { + "epoch": 0.7033383458646617, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 16693 + }, + { + "epoch": 0.7033984962406015, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 16694 + }, + { + "epoch": 0.7034586466165413, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 16695 + }, + { + "epoch": 0.7035187969924812, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 16696 + }, + { + "epoch": 0.7035789473684211, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 16697 + }, + { + "epoch": 0.703639097744361, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 16698 + }, + { + "epoch": 0.7036992481203007, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 16699 + }, + { + "epoch": 0.7037593984962406, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 16700 + }, + { + "epoch": 0.7038195488721805, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 16701 + }, + { + "epoch": 0.7038796992481203, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2139, + "step": 16702 + }, + { + "epoch": 0.7039398496240602, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2157, + "step": 16703 + }, + { + "epoch": 0.704, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 16704 + }, + { + "epoch": 0.7040601503759398, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 16705 + }, + { + "epoch": 0.7041203007518797, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.3542, + "step": 16706 + }, + { + "epoch": 0.7041804511278196, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 16707 + }, + { + "epoch": 0.7042406015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 16708 + }, + { + "epoch": 0.7043007518796992, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 16709 + }, + { + "epoch": 0.7043609022556391, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 16710 + }, + { + "epoch": 0.704421052631579, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 16711 + }, + { + "epoch": 0.7044812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 16712 + }, + { + "epoch": 0.7045413533834587, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 16713 + }, + { + "epoch": 0.7046015037593985, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 16714 + }, + { + "epoch": 0.7046616541353383, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 16715 + }, + { + "epoch": 0.7047218045112782, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 16716 + }, + { + "epoch": 0.7047819548872181, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 16717 + }, + { + "epoch": 0.7048421052631579, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 16718 + }, + { + "epoch": 0.7049022556390977, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 16719 + }, + { + "epoch": 0.7049624060150376, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 16720 + }, + { + "epoch": 0.7050225563909774, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 16721 + }, + { + "epoch": 0.7050827067669173, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 16722 + }, + { + "epoch": 0.7051428571428572, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1985, + "step": 16723 + }, + { + "epoch": 0.7052030075187969, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1073, + "step": 16724 + }, + { + "epoch": 0.7052631578947368, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 16725 + }, + { + "epoch": 0.7053233082706767, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 16726 + }, + { + "epoch": 0.7053834586466166, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 16727 + }, + { + "epoch": 0.7054436090225564, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 16728 + }, + { + "epoch": 0.7055037593984962, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 16729 + }, + { + "epoch": 0.7055639097744361, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 16730 + }, + { + "epoch": 0.7056240601503759, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 16731 + }, + { + "epoch": 0.7056842105263158, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 16732 + }, + { + "epoch": 0.7057443609022557, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1985, + "step": 16733 + }, + { + "epoch": 0.7058045112781954, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2074, + "step": 16734 + }, + { + "epoch": 0.7058646616541353, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 16735 + }, + { + "epoch": 0.7059248120300752, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 16736 + }, + { + "epoch": 0.705984962406015, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 16737 + }, + { + "epoch": 0.7060451127819549, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.217, + "step": 16738 + }, + { + "epoch": 0.7061052631578948, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 16739 + }, + { + "epoch": 0.7061654135338346, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 16740 + }, + { + "epoch": 0.7062255639097744, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.283, + "step": 16741 + }, + { + "epoch": 0.7062857142857143, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 16742 + }, + { + "epoch": 0.7063458646616542, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 16743 + }, + { + "epoch": 0.706406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 16744 + }, + { + "epoch": 0.7064661654135338, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 16745 + }, + { + "epoch": 0.7065263157894737, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 16746 + }, + { + "epoch": 0.7065864661654135, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 16747 + }, + { + "epoch": 0.7066466165413534, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 16748 + }, + { + "epoch": 0.7067067669172933, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 16749 + }, + { + "epoch": 0.706766917293233, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 16750 + }, + { + "epoch": 0.7068270676691729, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 16751 + }, + { + "epoch": 0.7068872180451128, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 16752 + }, + { + "epoch": 0.7069473684210527, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 16753 + }, + { + "epoch": 0.7070075187969925, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 16754 + }, + { + "epoch": 0.7070676691729323, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 16755 + }, + { + "epoch": 0.7071278195488722, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1952, + "step": 16756 + }, + { + "epoch": 0.707187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 16757 + }, + { + "epoch": 0.7072481203007519, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 16758 + }, + { + "epoch": 0.7073082706766918, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2004, + "step": 16759 + }, + { + "epoch": 0.7073684210526315, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 16760 + }, + { + "epoch": 0.7074285714285714, + "grad_norm": 3.28125, + "learning_rate": 3e-05, + "loss": 1.2305, + "step": 16761 + }, + { + "epoch": 0.7074887218045113, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 16762 + }, + { + "epoch": 0.7075488721804511, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 16763 + }, + { + "epoch": 0.707609022556391, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 16764 + }, + { + "epoch": 0.7076691729323308, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 16765 + }, + { + "epoch": 0.7077293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 16766 + }, + { + "epoch": 0.7077894736842105, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 16767 + }, + { + "epoch": 0.7078496240601504, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 16768 + }, + { + "epoch": 0.7079097744360903, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 16769 + }, + { + "epoch": 0.70796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1326, + "step": 16770 + }, + { + "epoch": 0.7080300751879699, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 16771 + }, + { + "epoch": 0.7080902255639098, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 16772 + }, + { + "epoch": 0.7081503759398496, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 16773 + }, + { + "epoch": 0.7082105263157895, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 16774 + }, + { + "epoch": 0.7082706766917293, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 16775 + }, + { + "epoch": 0.7083308270676691, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 16776 + }, + { + "epoch": 0.708390977443609, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1928, + "step": 16777 + }, + { + "epoch": 0.7084511278195489, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 16778 + }, + { + "epoch": 0.7085112781954888, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 16779 + }, + { + "epoch": 0.7085714285714285, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 16780 + }, + { + "epoch": 0.7086315789473684, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 16781 + }, + { + "epoch": 0.7086917293233083, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 16782 + }, + { + "epoch": 0.7087518796992481, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 16783 + }, + { + "epoch": 0.708812030075188, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 16784 + }, + { + "epoch": 0.7088721804511278, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 16785 + }, + { + "epoch": 0.7089323308270676, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 16786 + }, + { + "epoch": 0.7089924812030075, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2052, + "step": 16787 + }, + { + "epoch": 0.7090526315789474, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 16788 + }, + { + "epoch": 0.7091127819548873, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 16789 + }, + { + "epoch": 0.7091729323308271, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2161, + "step": 16790 + }, + { + "epoch": 0.7092330827067669, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 16791 + }, + { + "epoch": 0.7092932330827068, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2082, + "step": 16792 + }, + { + "epoch": 0.7093533834586466, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 16793 + }, + { + "epoch": 0.7094135338345865, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 16794 + }, + { + "epoch": 0.7094736842105264, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 16795 + }, + { + "epoch": 0.7095338345864661, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 16796 + }, + { + "epoch": 0.709593984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 16797 + }, + { + "epoch": 0.7096541353383459, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 16798 + }, + { + "epoch": 0.7097142857142857, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 16799 + }, + { + "epoch": 0.7097744360902256, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2409, + "step": 16800 + }, + { + "epoch": 0.7098345864661654, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.2042, + "step": 16801 + }, + { + "epoch": 0.7098947368421052, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 16802 + }, + { + "epoch": 0.7099548872180451, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 16803 + }, + { + "epoch": 0.710015037593985, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2063, + "step": 16804 + }, + { + "epoch": 0.7100751879699249, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 16805 + }, + { + "epoch": 0.7101353383458646, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 16806 + }, + { + "epoch": 0.7101954887218045, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 16807 + }, + { + "epoch": 0.7102556390977444, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 16808 + }, + { + "epoch": 0.7103157894736842, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 16809 + }, + { + "epoch": 0.7103759398496241, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 16810 + }, + { + "epoch": 0.7104360902255639, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 16811 + }, + { + "epoch": 0.7104962406015037, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 16812 + }, + { + "epoch": 0.7105563909774436, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 16813 + }, + { + "epoch": 0.7106165413533835, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 16814 + }, + { + "epoch": 0.7106766917293234, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 16815 + }, + { + "epoch": 0.7107368421052631, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 16816 + }, + { + "epoch": 0.710796992481203, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 16817 + }, + { + "epoch": 0.7108571428571429, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 16818 + }, + { + "epoch": 0.7109172932330827, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 16819 + }, + { + "epoch": 0.7109774436090226, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 16820 + }, + { + "epoch": 0.7110375939849624, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1254, + "step": 16821 + }, + { + "epoch": 0.7110977443609022, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2154, + "step": 16822 + }, + { + "epoch": 0.7111578947368421, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 16823 + }, + { + "epoch": 0.711218045112782, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 16824 + }, + { + "epoch": 0.7112781954887218, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2177, + "step": 16825 + }, + { + "epoch": 0.7113383458646616, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 16826 + }, + { + "epoch": 0.7113984962406015, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 16827 + }, + { + "epoch": 0.7114586466165413, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 16828 + }, + { + "epoch": 0.7115187969924812, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 16829 + }, + { + "epoch": 0.7115789473684211, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 16830 + }, + { + "epoch": 0.7116390977443608, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 16831 + }, + { + "epoch": 0.7116992481203007, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 16832 + }, + { + "epoch": 0.7117593984962406, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 16833 + }, + { + "epoch": 0.7118195488721805, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 16834 + }, + { + "epoch": 0.7118796992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 16835 + }, + { + "epoch": 0.7119398496240601, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 16836 + }, + { + "epoch": 0.712, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 16837 + }, + { + "epoch": 0.7120601503759398, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 16838 + }, + { + "epoch": 0.7121203007518797, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 16839 + }, + { + "epoch": 0.7121804511278196, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 16840 + }, + { + "epoch": 0.7122406015037595, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 16841 + }, + { + "epoch": 0.7123007518796992, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 16842 + }, + { + "epoch": 0.7123609022556391, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 16843 + }, + { + "epoch": 0.712421052631579, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 16844 + }, + { + "epoch": 0.7124812030075188, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 16845 + }, + { + "epoch": 0.7125413533834587, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 16846 + }, + { + "epoch": 0.7126015037593985, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 16847 + }, + { + "epoch": 0.7126616541353383, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 16848 + }, + { + "epoch": 0.7127218045112782, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1766, + "step": 16849 + }, + { + "epoch": 0.7127819548872181, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 16850 + }, + { + "epoch": 0.7128421052631579, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 16851 + }, + { + "epoch": 0.7129022556390977, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2312, + "step": 16852 + }, + { + "epoch": 0.7129624060150376, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 16853 + }, + { + "epoch": 0.7130225563909774, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 16854 + }, + { + "epoch": 0.7130827067669173, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 16855 + }, + { + "epoch": 0.7131428571428572, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 16856 + }, + { + "epoch": 0.713203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 16857 + }, + { + "epoch": 0.7132631578947368, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 16858 + }, + { + "epoch": 0.7133233082706767, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 16859 + }, + { + "epoch": 0.7133834586466166, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 16860 + }, + { + "epoch": 0.7134436090225564, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 16861 + }, + { + "epoch": 0.7135037593984962, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2108, + "step": 16862 + }, + { + "epoch": 0.7135639097744361, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 16863 + }, + { + "epoch": 0.7136240601503759, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 16864 + }, + { + "epoch": 0.7136842105263158, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 16865 + }, + { + "epoch": 0.7137443609022557, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 16866 + }, + { + "epoch": 0.7138045112781954, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 16867 + }, + { + "epoch": 0.7138646616541353, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 16868 + }, + { + "epoch": 0.7139248120300752, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 16869 + }, + { + "epoch": 0.7139849624060151, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 16870 + }, + { + "epoch": 0.7140451127819549, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 16871 + }, + { + "epoch": 0.7141052631578947, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1839, + "step": 16872 + }, + { + "epoch": 0.7141654135338346, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 16873 + }, + { + "epoch": 0.7142255639097744, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 16874 + }, + { + "epoch": 0.7142857142857143, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 16875 + }, + { + "epoch": 0.7143458646616542, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 16876 + }, + { + "epoch": 0.7144060150375939, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 16877 + }, + { + "epoch": 0.7144661654135338, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 16878 + }, + { + "epoch": 0.7145263157894737, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1886, + "step": 16879 + }, + { + "epoch": 0.7145864661654135, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 16880 + }, + { + "epoch": 0.7146466165413534, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 16881 + }, + { + "epoch": 0.7147067669172932, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 16882 + }, + { + "epoch": 0.714766917293233, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 16883 + }, + { + "epoch": 0.7148270676691729, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 16884 + }, + { + "epoch": 0.7148872180451128, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 16885 + }, + { + "epoch": 0.7149473684210527, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 16886 + }, + { + "epoch": 0.7150075187969925, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 16887 + }, + { + "epoch": 0.7150676691729323, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 16888 + }, + { + "epoch": 0.7151278195488722, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 16889 + }, + { + "epoch": 0.715187969924812, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 16890 + }, + { + "epoch": 0.7152481203007519, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 16891 + }, + { + "epoch": 0.7153082706766918, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 16892 + }, + { + "epoch": 0.7153684210526315, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 16893 + }, + { + "epoch": 0.7154285714285714, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 16894 + }, + { + "epoch": 0.7154887218045113, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2161, + "step": 16895 + }, + { + "epoch": 0.7155488721804512, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 16896 + }, + { + "epoch": 0.715609022556391, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 16897 + }, + { + "epoch": 0.7156691729323308, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2625, + "step": 16898 + }, + { + "epoch": 0.7157293233082707, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 16899 + }, + { + "epoch": 0.7157894736842105, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 16900 + }, + { + "epoch": 0.7158496240601504, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 16901 + }, + { + "epoch": 0.7159097744360903, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 16902 + }, + { + "epoch": 0.71596992481203, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 16903 + }, + { + "epoch": 0.7160300751879699, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1726, + "step": 16904 + }, + { + "epoch": 0.7160902255639098, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 16905 + }, + { + "epoch": 0.7161503759398496, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2434, + "step": 16906 + }, + { + "epoch": 0.7162105263157895, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2004, + "step": 16907 + }, + { + "epoch": 0.7162706766917293, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 16908 + }, + { + "epoch": 0.7163308270676692, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 16909 + }, + { + "epoch": 0.716390977443609, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 16910 + }, + { + "epoch": 0.7164511278195489, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 16911 + }, + { + "epoch": 0.7165112781954888, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 16912 + }, + { + "epoch": 0.7165714285714285, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 16913 + }, + { + "epoch": 0.7166315789473684, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 16914 + }, + { + "epoch": 0.7166917293233083, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 16915 + }, + { + "epoch": 0.7167518796992481, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 16916 + }, + { + "epoch": 0.716812030075188, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2248, + "step": 16917 + }, + { + "epoch": 0.7168721804511278, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 16918 + }, + { + "epoch": 0.7169323308270676, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 16919 + }, + { + "epoch": 0.7169924812030075, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 16920 + }, + { + "epoch": 0.7170526315789474, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 16921 + }, + { + "epoch": 0.7171127819548873, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 16922 + }, + { + "epoch": 0.717172932330827, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1612, + "step": 16923 + }, + { + "epoch": 0.7172330827067669, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 16924 + }, + { + "epoch": 0.7172932330827068, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 16925 + }, + { + "epoch": 0.7173533834586466, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 16926 + }, + { + "epoch": 0.7174135338345865, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 16927 + }, + { + "epoch": 0.7174736842105263, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 16928 + }, + { + "epoch": 0.7175338345864661, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1998, + "step": 16929 + }, + { + "epoch": 0.717593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 16930 + }, + { + "epoch": 0.7176541353383459, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 16931 + }, + { + "epoch": 0.7177142857142857, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 16932 + }, + { + "epoch": 0.7177744360902255, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2169, + "step": 16933 + }, + { + "epoch": 0.7178345864661654, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 16934 + }, + { + "epoch": 0.7178947368421053, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 16935 + }, + { + "epoch": 0.7179548872180451, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 16936 + }, + { + "epoch": 0.718015037593985, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 16937 + }, + { + "epoch": 0.7180751879699249, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2134, + "step": 16938 + }, + { + "epoch": 0.7181353383458646, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 16939 + }, + { + "epoch": 0.7181954887218045, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 16940 + }, + { + "epoch": 0.7182556390977444, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 16941 + }, + { + "epoch": 0.7183157894736842, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 16942 + }, + { + "epoch": 0.7183759398496241, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 16943 + }, + { + "epoch": 0.7184360902255639, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 16944 + }, + { + "epoch": 0.7184962406015037, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1604, + "step": 16945 + }, + { + "epoch": 0.7185563909774436, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 16946 + }, + { + "epoch": 0.7186165413533835, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1357, + "step": 16947 + }, + { + "epoch": 0.7186766917293234, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 16948 + }, + { + "epoch": 0.7187368421052631, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1997, + "step": 16949 + }, + { + "epoch": 0.718796992481203, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 16950 + }, + { + "epoch": 0.7188571428571429, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 16951 + }, + { + "epoch": 0.7189172932330827, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 16952 + }, + { + "epoch": 0.7189774436090226, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 16953 + }, + { + "epoch": 0.7190375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 16954 + }, + { + "epoch": 0.7190977443609022, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 16955 + }, + { + "epoch": 0.7191578947368421, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 16956 + }, + { + "epoch": 0.719218045112782, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 16957 + }, + { + "epoch": 0.7192781954887218, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.2043, + "step": 16958 + }, + { + "epoch": 0.7193383458646616, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 16959 + }, + { + "epoch": 0.7193984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 16960 + }, + { + "epoch": 0.7194586466165414, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1727, + "step": 16961 + }, + { + "epoch": 0.7195187969924812, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 16962 + }, + { + "epoch": 0.7195789473684211, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 16963 + }, + { + "epoch": 0.7196390977443609, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 16964 + }, + { + "epoch": 0.7196992481203007, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 16965 + }, + { + "epoch": 0.7197593984962406, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 16966 + }, + { + "epoch": 0.7198195488721805, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 16967 + }, + { + "epoch": 0.7198796992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 16968 + }, + { + "epoch": 0.7199398496240601, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1073, + "step": 16969 + }, + { + "epoch": 0.72, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 16970 + }, + { + "epoch": 0.7200601503759398, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1653, + "step": 16971 + }, + { + "epoch": 0.7201203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 16972 + }, + { + "epoch": 0.7201804511278196, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 16973 + }, + { + "epoch": 0.7202406015037593, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 16974 + }, + { + "epoch": 0.7203007518796992, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 16975 + }, + { + "epoch": 0.7203609022556391, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 16976 + }, + { + "epoch": 0.720421052631579, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 16977 + }, + { + "epoch": 0.7204812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 16978 + }, + { + "epoch": 0.7205413533834586, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 16979 + }, + { + "epoch": 0.7206015037593985, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 16980 + }, + { + "epoch": 0.7206616541353383, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 16981 + }, + { + "epoch": 0.7207218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 16982 + }, + { + "epoch": 0.7207819548872181, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1714, + "step": 16983 + }, + { + "epoch": 0.7208421052631578, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 16984 + }, + { + "epoch": 0.7209022556390977, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 16985 + }, + { + "epoch": 0.7209624060150376, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 16986 + }, + { + "epoch": 0.7210225563909775, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 16987 + }, + { + "epoch": 0.7210827067669173, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 16988 + }, + { + "epoch": 0.7211428571428572, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.2406, + "step": 16989 + }, + { + "epoch": 0.721203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 16990 + }, + { + "epoch": 0.7212631578947368, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 16991 + }, + { + "epoch": 0.7213233082706767, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 16992 + }, + { + "epoch": 0.7213834586466166, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 16993 + }, + { + "epoch": 0.7214436090225564, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.2577, + "step": 16994 + }, + { + "epoch": 0.7215037593984962, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 16995 + }, + { + "epoch": 0.7215639097744361, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 16996 + }, + { + "epoch": 0.721624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 16997 + }, + { + "epoch": 0.7216842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 16998 + }, + { + "epoch": 0.7217443609022557, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 16999 + }, + { + "epoch": 0.7218045112781954, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 17000 + }, + { + "epoch": 0.7218646616541353, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 17001 + }, + { + "epoch": 0.7219248120300752, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 17002 + }, + { + "epoch": 0.7219849624060151, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2237, + "step": 17003 + }, + { + "epoch": 0.7220451127819549, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 17004 + }, + { + "epoch": 0.7221052631578947, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 17005 + }, + { + "epoch": 0.7221654135338346, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 17006 + }, + { + "epoch": 0.7222255639097744, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1918, + "step": 17007 + }, + { + "epoch": 0.7222857142857143, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 17008 + }, + { + "epoch": 0.7223458646616542, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 17009 + }, + { + "epoch": 0.7224060150375939, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 17010 + }, + { + "epoch": 0.7224661654135338, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 17011 + }, + { + "epoch": 0.7225263157894737, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 17012 + }, + { + "epoch": 0.7225864661654136, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 17013 + }, + { + "epoch": 0.7226466165413534, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2176, + "step": 17014 + }, + { + "epoch": 0.7227067669172932, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2417, + "step": 17015 + }, + { + "epoch": 0.7227669172932331, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 17016 + }, + { + "epoch": 0.7228270676691729, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1806, + "step": 17017 + }, + { + "epoch": 0.7228872180451128, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 17018 + }, + { + "epoch": 0.7229473684210527, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 17019 + }, + { + "epoch": 0.7230075187969924, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 17020 + }, + { + "epoch": 0.7230676691729323, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 17021 + }, + { + "epoch": 0.7231278195488722, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 17022 + }, + { + "epoch": 0.723187969924812, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 17023 + }, + { + "epoch": 0.7232481203007519, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1652, + "step": 17024 + }, + { + "epoch": 0.7233082706766917, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 17025 + }, + { + "epoch": 0.7233684210526315, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2126, + "step": 17026 + }, + { + "epoch": 0.7234285714285714, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 17027 + }, + { + "epoch": 0.7234887218045113, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 17028 + }, + { + "epoch": 0.7235488721804512, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 17029 + }, + { + "epoch": 0.7236090225563909, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 17030 + }, + { + "epoch": 0.7236691729323308, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 17031 + }, + { + "epoch": 0.7237293233082707, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2413, + "step": 17032 + }, + { + "epoch": 0.7237894736842105, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 17033 + }, + { + "epoch": 0.7238496240601504, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2148, + "step": 17034 + }, + { + "epoch": 0.7239097744360903, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2087, + "step": 17035 + }, + { + "epoch": 0.72396992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 17036 + }, + { + "epoch": 0.7240300751879699, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 17037 + }, + { + "epoch": 0.7240902255639098, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 17038 + }, + { + "epoch": 0.7241503759398497, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 17039 + }, + { + "epoch": 0.7242105263157895, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 17040 + }, + { + "epoch": 0.7242706766917293, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 17041 + }, + { + "epoch": 0.7243308270676692, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 17042 + }, + { + "epoch": 0.724390977443609, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2001, + "step": 17043 + }, + { + "epoch": 0.7244511278195489, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 17044 + }, + { + "epoch": 0.7245112781954888, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 17045 + }, + { + "epoch": 0.7245714285714285, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 17046 + }, + { + "epoch": 0.7246315789473684, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 17047 + }, + { + "epoch": 0.7246917293233083, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1713, + "step": 17048 + }, + { + "epoch": 0.7247518796992481, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 17049 + }, + { + "epoch": 0.724812030075188, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 17050 + }, + { + "epoch": 0.7248721804511278, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 17051 + }, + { + "epoch": 0.7249323308270677, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 17052 + }, + { + "epoch": 0.7249924812030075, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 17053 + }, + { + "epoch": 0.7250526315789474, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 17054 + }, + { + "epoch": 0.7251127819548873, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.3969, + "step": 17055 + }, + { + "epoch": 0.725172932330827, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.1808, + "step": 17056 + }, + { + "epoch": 0.7252330827067669, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 17057 + }, + { + "epoch": 0.7252932330827068, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 17058 + }, + { + "epoch": 0.7253533834586466, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 17059 + }, + { + "epoch": 0.7254135338345865, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 17060 + }, + { + "epoch": 0.7254736842105263, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 17061 + }, + { + "epoch": 0.7255338345864661, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 17062 + }, + { + "epoch": 0.725593984962406, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 17063 + }, + { + "epoch": 0.7256541353383459, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 17064 + }, + { + "epoch": 0.7257142857142858, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2895, + "step": 17065 + }, + { + "epoch": 0.7257744360902255, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.2221, + "step": 17066 + }, + { + "epoch": 0.7258345864661654, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 17067 + }, + { + "epoch": 0.7258947368421053, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 17068 + }, + { + "epoch": 0.7259548872180451, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 17069 + }, + { + "epoch": 0.726015037593985, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 17070 + }, + { + "epoch": 0.7260751879699248, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 17071 + }, + { + "epoch": 0.7261353383458646, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 17072 + }, + { + "epoch": 0.7261954887218045, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 17073 + }, + { + "epoch": 0.7262556390977444, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 17074 + }, + { + "epoch": 0.7263157894736842, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 17075 + }, + { + "epoch": 0.726375939849624, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 17076 + }, + { + "epoch": 0.7264360902255639, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 17077 + }, + { + "epoch": 0.7264962406015038, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 17078 + }, + { + "epoch": 0.7265563909774436, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 17079 + }, + { + "epoch": 0.7266165413533835, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 17080 + }, + { + "epoch": 0.7266766917293233, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 17081 + }, + { + "epoch": 0.7267368421052631, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 17082 + }, + { + "epoch": 0.726796992481203, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 17083 + }, + { + "epoch": 0.7268571428571429, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 17084 + }, + { + "epoch": 0.7269172932330827, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 17085 + }, + { + "epoch": 0.7269774436090226, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 17086 + }, + { + "epoch": 0.7270375939849624, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 17087 + }, + { + "epoch": 0.7270977443609022, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.3187, + "step": 17088 + }, + { + "epoch": 0.7271578947368421, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 17089 + }, + { + "epoch": 0.727218045112782, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 17090 + }, + { + "epoch": 0.7272781954887219, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 17091 + }, + { + "epoch": 0.7273383458646616, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 17092 + }, + { + "epoch": 0.7273984962406015, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 17093 + }, + { + "epoch": 0.7274586466165414, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 17094 + }, + { + "epoch": 0.7275187969924812, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1713, + "step": 17095 + }, + { + "epoch": 0.7275789473684211, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 17096 + }, + { + "epoch": 0.7276390977443609, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 17097 + }, + { + "epoch": 0.7276992481203007, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 17098 + }, + { + "epoch": 0.7277593984962406, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 17099 + }, + { + "epoch": 0.7278195488721805, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 17100 + }, + { + "epoch": 0.7278796992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 17101 + }, + { + "epoch": 0.7279398496240601, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 17102 + }, + { + "epoch": 0.728, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1938, + "step": 17103 + }, + { + "epoch": 0.7280601503759399, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 17104 + }, + { + "epoch": 0.7281203007518797, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 17105 + }, + { + "epoch": 0.7281804511278196, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 17106 + }, + { + "epoch": 0.7282406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 17107 + }, + { + "epoch": 0.7283007518796992, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 17108 + }, + { + "epoch": 0.7283609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 17109 + }, + { + "epoch": 0.728421052631579, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 17110 + }, + { + "epoch": 0.7284812030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 17111 + }, + { + "epoch": 0.7285413533834586, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 17112 + }, + { + "epoch": 0.7286015037593985, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 17113 + }, + { + "epoch": 0.7286616541353383, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2264, + "step": 17114 + }, + { + "epoch": 0.7287218045112782, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 17115 + }, + { + "epoch": 0.7287819548872181, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 17116 + }, + { + "epoch": 0.7288421052631578, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2967, + "step": 17117 + }, + { + "epoch": 0.7289022556390977, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 17118 + }, + { + "epoch": 0.7289624060150376, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 17119 + }, + { + "epoch": 0.7290225563909775, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 17120 + }, + { + "epoch": 0.7290827067669173, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 17121 + }, + { + "epoch": 0.7291428571428571, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.26, + "step": 17122 + }, + { + "epoch": 0.729203007518797, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 17123 + }, + { + "epoch": 0.7292631578947368, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 17124 + }, + { + "epoch": 0.7293233082706767, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 17125 + }, + { + "epoch": 0.7293834586466166, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 17126 + }, + { + "epoch": 0.7294436090225563, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 17127 + }, + { + "epoch": 0.7295037593984962, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 17128 + }, + { + "epoch": 0.7295639097744361, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 17129 + }, + { + "epoch": 0.729624060150376, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 17130 + }, + { + "epoch": 0.7296842105263158, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 17131 + }, + { + "epoch": 0.7297443609022556, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2194, + "step": 17132 + }, + { + "epoch": 0.7298045112781955, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 17133 + }, + { + "epoch": 0.7298646616541353, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 17134 + }, + { + "epoch": 0.7299248120300752, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 17135 + }, + { + "epoch": 0.7299849624060151, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 17136 + }, + { + "epoch": 0.7300451127819549, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 17137 + }, + { + "epoch": 0.7301052631578947, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 17138 + }, + { + "epoch": 0.7301654135338346, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 17139 + }, + { + "epoch": 0.7302255639097744, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 17140 + }, + { + "epoch": 0.7302857142857143, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 17141 + }, + { + "epoch": 0.7303458646616542, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 17142 + }, + { + "epoch": 0.730406015037594, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 17143 + }, + { + "epoch": 0.7304661654135338, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 17144 + }, + { + "epoch": 0.7305263157894737, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.3163, + "step": 17145 + }, + { + "epoch": 0.7305864661654136, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 17146 + }, + { + "epoch": 0.7306466165413534, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.289, + "step": 17147 + }, + { + "epoch": 0.7307067669172932, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 17148 + }, + { + "epoch": 0.7307669172932331, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2472, + "step": 17149 + }, + { + "epoch": 0.7308270676691729, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 17150 + }, + { + "epoch": 0.7308872180451128, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 17151 + }, + { + "epoch": 0.7309473684210527, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 17152 + }, + { + "epoch": 0.7310075187969924, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 17153 + }, + { + "epoch": 0.7310676691729323, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 17154 + }, + { + "epoch": 0.7311278195488722, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 17155 + }, + { + "epoch": 0.731187969924812, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 17156 + }, + { + "epoch": 0.7312481203007519, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 17157 + }, + { + "epoch": 0.7313082706766917, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 17158 + }, + { + "epoch": 0.7313684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 17159 + }, + { + "epoch": 0.7314285714285714, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 17160 + }, + { + "epoch": 0.7314887218045113, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2234, + "step": 17161 + }, + { + "epoch": 0.7315488721804512, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 17162 + }, + { + "epoch": 0.7316090225563909, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 17163 + }, + { + "epoch": 0.7316691729323308, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2153, + "step": 17164 + }, + { + "epoch": 0.7317293233082707, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 17165 + }, + { + "epoch": 0.7317894736842105, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 17166 + }, + { + "epoch": 0.7318496240601504, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 17167 + }, + { + "epoch": 0.7319097744360902, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1604, + "step": 17168 + }, + { + "epoch": 0.73196992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1832, + "step": 17169 + }, + { + "epoch": 0.7320300751879699, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 17170 + }, + { + "epoch": 0.7320902255639098, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 17171 + }, + { + "epoch": 0.7321503759398497, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 17172 + }, + { + "epoch": 0.7322105263157894, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 17173 + }, + { + "epoch": 0.7322706766917293, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 17174 + }, + { + "epoch": 0.7323308270676692, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 17175 + }, + { + "epoch": 0.732390977443609, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 17176 + }, + { + "epoch": 0.7324511278195489, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 17177 + }, + { + "epoch": 0.7325112781954887, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2129, + "step": 17178 + }, + { + "epoch": 0.7325714285714285, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 17179 + }, + { + "epoch": 0.7326315789473684, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 17180 + }, + { + "epoch": 0.7326917293233083, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 17181 + }, + { + "epoch": 0.7327518796992482, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 17182 + }, + { + "epoch": 0.7328120300751879, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 17183 + }, + { + "epoch": 0.7328721804511278, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 17184 + }, + { + "epoch": 0.7329323308270677, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 17185 + }, + { + "epoch": 0.7329924812030075, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 17186 + }, + { + "epoch": 0.7330526315789474, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 17187 + }, + { + "epoch": 0.7331127819548873, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 17188 + }, + { + "epoch": 0.733172932330827, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 17189 + }, + { + "epoch": 0.7332330827067669, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 17190 + }, + { + "epoch": 0.7332932330827068, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 17191 + }, + { + "epoch": 0.7333533834586466, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 17192 + }, + { + "epoch": 0.7334135338345865, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2172, + "step": 17193 + }, + { + "epoch": 0.7334736842105263, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 17194 + }, + { + "epoch": 0.7335338345864661, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 17195 + }, + { + "epoch": 0.733593984962406, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 17196 + }, + { + "epoch": 0.7336541353383459, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 17197 + }, + { + "epoch": 0.7337142857142858, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 17198 + }, + { + "epoch": 0.7337744360902255, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 17199 + }, + { + "epoch": 0.7338345864661654, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1925, + "step": 17200 + }, + { + "epoch": 0.7338947368421053, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 17201 + }, + { + "epoch": 0.7339548872180451, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 17202 + }, + { + "epoch": 0.734015037593985, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 17203 + }, + { + "epoch": 0.7340751879699248, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 17204 + }, + { + "epoch": 0.7341353383458646, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 17205 + }, + { + "epoch": 0.7341954887218045, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 17206 + }, + { + "epoch": 0.7342556390977444, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 17207 + }, + { + "epoch": 0.7343157894736843, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2036, + "step": 17208 + }, + { + "epoch": 0.734375939849624, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 17209 + }, + { + "epoch": 0.7344360902255639, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 17210 + }, + { + "epoch": 0.7344962406015038, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 17211 + }, + { + "epoch": 0.7345563909774436, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1339, + "step": 17212 + }, + { + "epoch": 0.7346165413533835, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 17213 + }, + { + "epoch": 0.7346766917293233, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2107, + "step": 17214 + }, + { + "epoch": 0.7347368421052631, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 17215 + }, + { + "epoch": 0.734796992481203, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 17216 + }, + { + "epoch": 0.7348571428571429, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1805, + "step": 17217 + }, + { + "epoch": 0.7349172932330827, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 17218 + }, + { + "epoch": 0.7349774436090225, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 17219 + }, + { + "epoch": 0.7350375939849624, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.241, + "step": 17220 + }, + { + "epoch": 0.7350977443609022, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 17221 + }, + { + "epoch": 0.7351578947368421, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 17222 + }, + { + "epoch": 0.735218045112782, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 17223 + }, + { + "epoch": 0.7352781954887218, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 17224 + }, + { + "epoch": 0.7353383458646616, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 17225 + }, + { + "epoch": 0.7353984962406015, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 17226 + }, + { + "epoch": 0.7354586466165414, + "grad_norm": 1.4140625, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 17227 + }, + { + "epoch": 0.7355187969924812, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 17228 + }, + { + "epoch": 0.735578947368421, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 17229 + }, + { + "epoch": 0.7356390977443609, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 17230 + }, + { + "epoch": 0.7356992481203007, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 17231 + }, + { + "epoch": 0.7357593984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 17232 + }, + { + "epoch": 0.7358195488721805, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 17233 + }, + { + "epoch": 0.7358796992481204, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 17234 + }, + { + "epoch": 0.7359398496240601, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 17235 + }, + { + "epoch": 0.736, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1526, + "step": 17236 + }, + { + "epoch": 0.7360601503759399, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 17237 + }, + { + "epoch": 0.7361203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 17238 + }, + { + "epoch": 0.7361804511278196, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 17239 + }, + { + "epoch": 0.7362406015037594, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 17240 + }, + { + "epoch": 0.7363007518796992, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 17241 + }, + { + "epoch": 0.7363609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 17242 + }, + { + "epoch": 0.736421052631579, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 17243 + }, + { + "epoch": 0.7364812030075188, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 17244 + }, + { + "epoch": 0.7365413533834586, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 17245 + }, + { + "epoch": 0.7366015037593985, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 17246 + }, + { + "epoch": 0.7366616541353384, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2118, + "step": 17247 + }, + { + "epoch": 0.7367218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 17248 + }, + { + "epoch": 0.7367819548872181, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 17249 + }, + { + "epoch": 0.7368421052631579, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2896, + "step": 17250 + }, + { + "epoch": 0.7369022556390977, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 17251 + }, + { + "epoch": 0.7369624060150376, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 17252 + }, + { + "epoch": 0.7370225563909775, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 17253 + }, + { + "epoch": 0.7370827067669173, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 17254 + }, + { + "epoch": 0.7371428571428571, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 17255 + }, + { + "epoch": 0.737203007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 17256 + }, + { + "epoch": 0.7372631578947368, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 17257 + }, + { + "epoch": 0.7373233082706767, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 17258 + }, + { + "epoch": 0.7373834586466166, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 17259 + }, + { + "epoch": 0.7374436090225563, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 17260 + }, + { + "epoch": 0.7375037593984962, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1264, + "step": 17261 + }, + { + "epoch": 0.7375639097744361, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 17262 + }, + { + "epoch": 0.737624060150376, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 17263 + }, + { + "epoch": 0.7376842105263158, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 17264 + }, + { + "epoch": 0.7377443609022556, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 17265 + }, + { + "epoch": 0.7378045112781955, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 17266 + }, + { + "epoch": 0.7378646616541353, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 17267 + }, + { + "epoch": 0.7379248120300752, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 17268 + }, + { + "epoch": 0.7379849624060151, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 17269 + }, + { + "epoch": 0.7380451127819548, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 17270 + }, + { + "epoch": 0.7381052631578947, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 17271 + }, + { + "epoch": 0.7381654135338346, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 17272 + }, + { + "epoch": 0.7382255639097745, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 17273 + }, + { + "epoch": 0.7382857142857143, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 17274 + }, + { + "epoch": 0.7383458646616541, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1845, + "step": 17275 + }, + { + "epoch": 0.738406015037594, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 17276 + }, + { + "epoch": 0.7384661654135338, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 17277 + }, + { + "epoch": 0.7385263157894737, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.3682, + "step": 17278 + }, + { + "epoch": 0.7385864661654136, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 17279 + }, + { + "epoch": 0.7386466165413533, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 17280 + }, + { + "epoch": 0.7387067669172932, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 17281 + }, + { + "epoch": 0.7387669172932331, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 17282 + }, + { + "epoch": 0.7388270676691729, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 17283 + }, + { + "epoch": 0.7388872180451128, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 17284 + }, + { + "epoch": 0.7389473684210527, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3365, + "step": 17285 + }, + { + "epoch": 0.7390075187969924, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 17286 + }, + { + "epoch": 0.7390676691729323, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 17287 + }, + { + "epoch": 0.7391278195488722, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 17288 + }, + { + "epoch": 0.7391879699248121, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2426, + "step": 17289 + }, + { + "epoch": 0.7392481203007519, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 17290 + }, + { + "epoch": 0.7393082706766917, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 17291 + }, + { + "epoch": 0.7393684210526316, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 17292 + }, + { + "epoch": 0.7394285714285714, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 17293 + }, + { + "epoch": 0.7394887218045113, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1955, + "step": 17294 + }, + { + "epoch": 0.7395488721804512, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 17295 + }, + { + "epoch": 0.7396090225563909, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 17296 + }, + { + "epoch": 0.7396691729323308, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 17297 + }, + { + "epoch": 0.7397293233082707, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 17298 + }, + { + "epoch": 0.7397894736842106, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 17299 + }, + { + "epoch": 0.7398496240601504, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 17300 + }, + { + "epoch": 0.7399097744360902, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 17301 + }, + { + "epoch": 0.73996992481203, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1702, + "step": 17302 + }, + { + "epoch": 0.7400300751879699, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 17303 + }, + { + "epoch": 0.7400902255639098, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 17304 + }, + { + "epoch": 0.7401503759398497, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 17305 + }, + { + "epoch": 0.7402105263157894, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1429, + "step": 17306 + }, + { + "epoch": 0.7402706766917293, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 17307 + }, + { + "epoch": 0.7403308270676692, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 17308 + }, + { + "epoch": 0.740390977443609, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 17309 + }, + { + "epoch": 0.7404511278195489, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 17310 + }, + { + "epoch": 0.7405112781954887, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2427, + "step": 17311 + }, + { + "epoch": 0.7405714285714285, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 17312 + }, + { + "epoch": 0.7406315789473684, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 17313 + }, + { + "epoch": 0.7406917293233083, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 17314 + }, + { + "epoch": 0.7407518796992482, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 17315 + }, + { + "epoch": 0.7408120300751879, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 17316 + }, + { + "epoch": 0.7408721804511278, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 17317 + }, + { + "epoch": 0.7409323308270677, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 17318 + }, + { + "epoch": 0.7409924812030075, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 17319 + }, + { + "epoch": 0.7410526315789474, + "grad_norm": 1.6953125, + "learning_rate": 3e-05, + "loss": 1.3676, + "step": 17320 + }, + { + "epoch": 0.7411127819548872, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 17321 + }, + { + "epoch": 0.741172932330827, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 17322 + }, + { + "epoch": 0.7412330827067669, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 17323 + }, + { + "epoch": 0.7412932330827068, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2197, + "step": 17324 + }, + { + "epoch": 0.7413533834586467, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1725, + "step": 17325 + }, + { + "epoch": 0.7414135338345864, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 17326 + }, + { + "epoch": 0.7414736842105263, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 17327 + }, + { + "epoch": 0.7415338345864662, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 17328 + }, + { + "epoch": 0.741593984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2316, + "step": 17329 + }, + { + "epoch": 0.7416541353383459, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2289, + "step": 17330 + }, + { + "epoch": 0.7417142857142857, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 17331 + }, + { + "epoch": 0.7417744360902255, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 17332 + }, + { + "epoch": 0.7418345864661654, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1677, + "step": 17333 + }, + { + "epoch": 0.7418947368421053, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.2224, + "step": 17334 + }, + { + "epoch": 0.7419548872180451, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1264, + "step": 17335 + }, + { + "epoch": 0.742015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.21, + "step": 17336 + }, + { + "epoch": 0.7420751879699248, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 17337 + }, + { + "epoch": 0.7421353383458646, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 17338 + }, + { + "epoch": 0.7421954887218045, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 17339 + }, + { + "epoch": 0.7422556390977444, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 17340 + }, + { + "epoch": 0.7423157894736843, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 17341 + }, + { + "epoch": 0.742375939849624, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 17342 + }, + { + "epoch": 0.7424360902255639, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 17343 + }, + { + "epoch": 0.7424962406015038, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 17344 + }, + { + "epoch": 0.7425563909774436, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2027, + "step": 17345 + }, + { + "epoch": 0.7426165413533835, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 17346 + }, + { + "epoch": 0.7426766917293233, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 17347 + }, + { + "epoch": 0.7427368421052631, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 17348 + }, + { + "epoch": 0.742796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 17349 + }, + { + "epoch": 0.7428571428571429, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 17350 + }, + { + "epoch": 0.7429172932330828, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1731, + "step": 17351 + }, + { + "epoch": 0.7429774436090225, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2829, + "step": 17352 + }, + { + "epoch": 0.7430375939849624, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 17353 + }, + { + "epoch": 0.7430977443609023, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 17354 + }, + { + "epoch": 0.7431578947368421, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 17355 + }, + { + "epoch": 0.743218045112782, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 17356 + }, + { + "epoch": 0.7432781954887218, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 17357 + }, + { + "epoch": 0.7433383458646616, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 17358 + }, + { + "epoch": 0.7433984962406015, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 17359 + }, + { + "epoch": 0.7434586466165414, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 17360 + }, + { + "epoch": 0.7435187969924812, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 17361 + }, + { + "epoch": 0.743578947368421, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 17362 + }, + { + "epoch": 0.7436390977443609, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 17363 + }, + { + "epoch": 0.7436992481203007, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 17364 + }, + { + "epoch": 0.7437593984962406, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.3739, + "step": 17365 + }, + { + "epoch": 0.7438195488721805, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 17366 + }, + { + "epoch": 0.7438796992481203, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 17367 + }, + { + "epoch": 0.7439398496240601, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 17368 + }, + { + "epoch": 0.744, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0994, + "step": 17369 + }, + { + "epoch": 0.7440601503759399, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 17370 + }, + { + "epoch": 0.7441203007518797, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 17371 + }, + { + "epoch": 0.7441804511278195, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 17372 + }, + { + "epoch": 0.7442406015037594, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 17373 + }, + { + "epoch": 0.7443007518796992, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2201, + "step": 17374 + }, + { + "epoch": 0.7443609022556391, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 17375 + }, + { + "epoch": 0.744421052631579, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 17376 + }, + { + "epoch": 0.7444812030075187, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 17377 + }, + { + "epoch": 0.7445413533834586, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 17378 + }, + { + "epoch": 0.7446015037593985, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 17379 + }, + { + "epoch": 0.7446616541353384, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 17380 + }, + { + "epoch": 0.7447218045112782, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 17381 + }, + { + "epoch": 0.7447819548872181, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 17382 + }, + { + "epoch": 0.7448421052631579, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1726, + "step": 17383 + }, + { + "epoch": 0.7449022556390977, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 17384 + }, + { + "epoch": 0.7449624060150376, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 17385 + }, + { + "epoch": 0.7450225563909775, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 17386 + }, + { + "epoch": 0.7450827067669173, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 17387 + }, + { + "epoch": 0.7451428571428571, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 17388 + }, + { + "epoch": 0.745203007518797, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 17389 + }, + { + "epoch": 0.7452631578947368, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.2117, + "step": 17390 + }, + { + "epoch": 0.7453233082706767, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 17391 + }, + { + "epoch": 0.7453834586466166, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 17392 + }, + { + "epoch": 0.7454436090225564, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 17393 + }, + { + "epoch": 0.7455037593984962, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 17394 + }, + { + "epoch": 0.7455639097744361, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 17395 + }, + { + "epoch": 0.745624060150376, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 17396 + }, + { + "epoch": 0.7456842105263158, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2465, + "step": 17397 + }, + { + "epoch": 0.7457443609022556, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1746, + "step": 17398 + }, + { + "epoch": 0.7458045112781955, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 17399 + }, + { + "epoch": 0.7458646616541353, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 17400 + }, + { + "epoch": 0.7459248120300752, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 17401 + }, + { + "epoch": 0.7459849624060151, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 17402 + }, + { + "epoch": 0.7460451127819548, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 17403 + }, + { + "epoch": 0.7461052631578947, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 17404 + }, + { + "epoch": 0.7461654135338346, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 17405 + }, + { + "epoch": 0.7462255639097745, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1496, + "step": 17406 + }, + { + "epoch": 0.7462857142857143, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 17407 + }, + { + "epoch": 0.7463458646616541, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 17408 + }, + { + "epoch": 0.746406015037594, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 17409 + }, + { + "epoch": 0.7464661654135338, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 17410 + }, + { + "epoch": 0.7465263157894737, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 17411 + }, + { + "epoch": 0.7465864661654136, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 17412 + }, + { + "epoch": 0.7466466165413533, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 17413 + }, + { + "epoch": 0.7467067669172932, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 17414 + }, + { + "epoch": 0.7467669172932331, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 17415 + }, + { + "epoch": 0.746827067669173, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 17416 + }, + { + "epoch": 0.7468872180451128, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 17417 + }, + { + "epoch": 0.7469473684210526, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 17418 + }, + { + "epoch": 0.7470075187969925, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 17419 + }, + { + "epoch": 0.7470676691729323, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 17420 + }, + { + "epoch": 0.7471278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 17421 + }, + { + "epoch": 0.7471879699248121, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 17422 + }, + { + "epoch": 0.7472481203007518, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 17423 + }, + { + "epoch": 0.7473082706766917, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1682, + "step": 17424 + }, + { + "epoch": 0.7473684210526316, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 17425 + }, + { + "epoch": 0.7474285714285714, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 17426 + }, + { + "epoch": 0.7474887218045113, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 17427 + }, + { + "epoch": 0.7475488721804511, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 17428 + }, + { + "epoch": 0.747609022556391, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 17429 + }, + { + "epoch": 0.7476691729323308, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 17430 + }, + { + "epoch": 0.7477293233082707, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 17431 + }, + { + "epoch": 0.7477894736842106, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2306, + "step": 17432 + }, + { + "epoch": 0.7478496240601504, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2976, + "step": 17433 + }, + { + "epoch": 0.7479097744360902, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 17434 + }, + { + "epoch": 0.7479699248120301, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 17435 + }, + { + "epoch": 0.7480300751879699, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 17436 + }, + { + "epoch": 0.7480902255639098, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 17437 + }, + { + "epoch": 0.7481503759398497, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 17438 + }, + { + "epoch": 0.7482105263157894, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1928, + "step": 17439 + }, + { + "epoch": 0.7482706766917293, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 17440 + }, + { + "epoch": 0.7483308270676692, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2072, + "step": 17441 + }, + { + "epoch": 0.748390977443609, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 17442 + }, + { + "epoch": 0.7484511278195489, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 17443 + }, + { + "epoch": 0.7485112781954887, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 17444 + }, + { + "epoch": 0.7485714285714286, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 17445 + }, + { + "epoch": 0.7486315789473684, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1671, + "step": 17446 + }, + { + "epoch": 0.7486917293233083, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 17447 + }, + { + "epoch": 0.7487518796992482, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 17448 + }, + { + "epoch": 0.7488120300751879, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 17449 + }, + { + "epoch": 0.7488721804511278, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 17450 + }, + { + "epoch": 0.7489323308270677, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 17451 + }, + { + "epoch": 0.7489924812030075, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 17452 + }, + { + "epoch": 0.7490526315789474, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 17453 + }, + { + "epoch": 0.7491127819548872, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 17454 + }, + { + "epoch": 0.749172932330827, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 17455 + }, + { + "epoch": 0.7492330827067669, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 17456 + }, + { + "epoch": 0.7492932330827068, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 17457 + }, + { + "epoch": 0.7493533834586467, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 17458 + }, + { + "epoch": 0.7494135338345864, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 17459 + }, + { + "epoch": 0.7494736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.3128, + "step": 17460 + }, + { + "epoch": 0.7495338345864662, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 17461 + }, + { + "epoch": 0.749593984962406, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 17462 + }, + { + "epoch": 0.7496541353383459, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 17463 + }, + { + "epoch": 0.7497142857142857, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 17464 + }, + { + "epoch": 0.7497744360902255, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2531, + "step": 17465 + }, + { + "epoch": 0.7498345864661654, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 17466 + }, + { + "epoch": 0.7498947368421053, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 17467 + }, + { + "epoch": 0.7499548872180452, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1957, + "step": 17468 + }, + { + "epoch": 0.7500150375939849, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 17469 + }, + { + "epoch": 0.7500751879699248, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 17470 + }, + { + "epoch": 0.7501353383458647, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 17471 + }, + { + "epoch": 0.7501954887218045, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 17472 + }, + { + "epoch": 0.7502556390977444, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 17473 + }, + { + "epoch": 0.7503157894736842, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1751, + "step": 17474 + }, + { + "epoch": 0.750375939849624, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 17475 + }, + { + "epoch": 0.7504360902255639, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 17476 + }, + { + "epoch": 0.7504962406015038, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 17477 + }, + { + "epoch": 0.7505563909774436, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 17478 + }, + { + "epoch": 0.7506165413533834, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 17479 + }, + { + "epoch": 0.7506766917293233, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 17480 + }, + { + "epoch": 0.7507368421052631, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 17481 + }, + { + "epoch": 0.750796992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 17482 + }, + { + "epoch": 0.7508571428571429, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 17483 + }, + { + "epoch": 0.7509172932330828, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 17484 + }, + { + "epoch": 0.7509774436090225, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 17485 + }, + { + "epoch": 0.7510375939849624, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 17486 + }, + { + "epoch": 0.7510977443609023, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 17487 + }, + { + "epoch": 0.7511578947368421, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 17488 + }, + { + "epoch": 0.751218045112782, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 17489 + }, + { + "epoch": 0.7512781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1418, + "step": 17490 + }, + { + "epoch": 0.7513383458646616, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 17491 + }, + { + "epoch": 0.7513984962406015, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 17492 + }, + { + "epoch": 0.7514586466165414, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 17493 + }, + { + "epoch": 0.7515187969924813, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 17494 + }, + { + "epoch": 0.751578947368421, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 17495 + }, + { + "epoch": 0.7516390977443609, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 17496 + }, + { + "epoch": 0.7516992481203008, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 17497 + }, + { + "epoch": 0.7517593984962406, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 17498 + }, + { + "epoch": 0.7518195488721805, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 17499 + }, + { + "epoch": 0.7518796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 17500 + }, + { + "epoch": 0.7519398496240601, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 17501 + }, + { + "epoch": 0.752, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 17502 + }, + { + "epoch": 0.7520601503759399, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 17503 + }, + { + "epoch": 0.7521203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 17504 + }, + { + "epoch": 0.7521804511278195, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 17505 + }, + { + "epoch": 0.7522406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 17506 + }, + { + "epoch": 0.7523007518796992, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 17507 + }, + { + "epoch": 0.7523609022556391, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 17508 + }, + { + "epoch": 0.752421052631579, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 17509 + }, + { + "epoch": 0.7524812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1326, + "step": 17510 + }, + { + "epoch": 0.7525413533834586, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 17511 + }, + { + "epoch": 0.7526015037593985, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1962, + "step": 17512 + }, + { + "epoch": 0.7526616541353384, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 17513 + }, + { + "epoch": 0.7527218045112782, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 17514 + }, + { + "epoch": 0.752781954887218, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 17515 + }, + { + "epoch": 0.7528421052631579, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 17516 + }, + { + "epoch": 0.7529022556390977, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 17517 + }, + { + "epoch": 0.7529624060150376, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 17518 + }, + { + "epoch": 0.7530225563909775, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 17519 + }, + { + "epoch": 0.7530827067669172, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 17520 + }, + { + "epoch": 0.7531428571428571, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 17521 + }, + { + "epoch": 0.753203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 17522 + }, + { + "epoch": 0.7532631578947369, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 17523 + }, + { + "epoch": 0.7533233082706767, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 17524 + }, + { + "epoch": 0.7533834586466165, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2154, + "step": 17525 + }, + { + "epoch": 0.7534436090225564, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 17526 + }, + { + "epoch": 0.7535037593984962, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 17527 + }, + { + "epoch": 0.7535639097744361, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 17528 + }, + { + "epoch": 0.753624060150376, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2829, + "step": 17529 + }, + { + "epoch": 0.7536842105263157, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 17530 + }, + { + "epoch": 0.7537443609022556, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 17531 + }, + { + "epoch": 0.7538045112781955, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 17532 + }, + { + "epoch": 0.7538646616541353, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 17533 + }, + { + "epoch": 0.7539248120300752, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 17534 + }, + { + "epoch": 0.7539849624060151, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2965, + "step": 17535 + }, + { + "epoch": 0.7540451127819549, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.2188, + "step": 17536 + }, + { + "epoch": 0.7541052631578947, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 17537 + }, + { + "epoch": 0.7541654135338346, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 17538 + }, + { + "epoch": 0.7542255639097745, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 17539 + }, + { + "epoch": 0.7542857142857143, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1986, + "step": 17540 + }, + { + "epoch": 0.7543458646616541, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 17541 + }, + { + "epoch": 0.754406015037594, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 17542 + }, + { + "epoch": 0.7544661654135338, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 17543 + }, + { + "epoch": 0.7545263157894737, + "grad_norm": 1.5390625, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 17544 + }, + { + "epoch": 0.7545864661654136, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2908, + "step": 17545 + }, + { + "epoch": 0.7546466165413533, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2546, + "step": 17546 + }, + { + "epoch": 0.7547067669172932, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 17547 + }, + { + "epoch": 0.7547669172932331, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 17548 + }, + { + "epoch": 0.754827067669173, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1722, + "step": 17549 + }, + { + "epoch": 0.7548872180451128, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 17550 + }, + { + "epoch": 0.7549473684210526, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 17551 + }, + { + "epoch": 0.7550075187969925, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 17552 + }, + { + "epoch": 0.7550676691729323, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 17553 + }, + { + "epoch": 0.7551278195488722, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 17554 + }, + { + "epoch": 0.7551879699248121, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 17555 + }, + { + "epoch": 0.7552481203007518, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 17556 + }, + { + "epoch": 0.7553082706766917, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 17557 + }, + { + "epoch": 0.7553684210526316, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 17558 + }, + { + "epoch": 0.7554285714285714, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 17559 + }, + { + "epoch": 0.7554887218045113, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2638, + "step": 17560 + }, + { + "epoch": 0.7555488721804511, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 17561 + }, + { + "epoch": 0.755609022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 17562 + }, + { + "epoch": 0.7556691729323308, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1677, + "step": 17563 + }, + { + "epoch": 0.7557293233082707, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 17564 + }, + { + "epoch": 0.7557894736842106, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 17565 + }, + { + "epoch": 0.7558496240601503, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 17566 + }, + { + "epoch": 0.7559097744360902, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 17567 + }, + { + "epoch": 0.7559699248120301, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 17568 + }, + { + "epoch": 0.7560300751879699, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 17569 + }, + { + "epoch": 0.7560902255639098, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 17570 + }, + { + "epoch": 0.7561503759398496, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1814, + "step": 17571 + }, + { + "epoch": 0.7562105263157894, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 17572 + }, + { + "epoch": 0.7562706766917293, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 17573 + }, + { + "epoch": 0.7563308270676692, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 17574 + }, + { + "epoch": 0.7563909774436091, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 17575 + }, + { + "epoch": 0.7564511278195488, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 17576 + }, + { + "epoch": 0.7565112781954887, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 17577 + }, + { + "epoch": 0.7565714285714286, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 17578 + }, + { + "epoch": 0.7566315789473684, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 17579 + }, + { + "epoch": 0.7566917293233083, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 17580 + }, + { + "epoch": 0.7567518796992482, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2442, + "step": 17581 + }, + { + "epoch": 0.7568120300751879, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 17582 + }, + { + "epoch": 0.7568721804511278, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 17583 + }, + { + "epoch": 0.7569323308270677, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 17584 + }, + { + "epoch": 0.7569924812030076, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2646, + "step": 17585 + }, + { + "epoch": 0.7570526315789474, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 17586 + }, + { + "epoch": 0.7571127819548872, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 17587 + }, + { + "epoch": 0.757172932330827, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 17588 + }, + { + "epoch": 0.7572330827067669, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1332, + "step": 17589 + }, + { + "epoch": 0.7572932330827068, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 17590 + }, + { + "epoch": 0.7573533834586467, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 17591 + }, + { + "epoch": 0.7574135338345864, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 17592 + }, + { + "epoch": 0.7574736842105263, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1692, + "step": 17593 + }, + { + "epoch": 0.7575338345864662, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1904, + "step": 17594 + }, + { + "epoch": 0.757593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 17595 + }, + { + "epoch": 0.7576541353383459, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 17596 + }, + { + "epoch": 0.7577142857142857, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 17597 + }, + { + "epoch": 0.7577744360902255, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 17598 + }, + { + "epoch": 0.7578345864661654, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 17599 + }, + { + "epoch": 0.7578947368421053, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 17600 + }, + { + "epoch": 0.7579548872180452, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 17601 + }, + { + "epoch": 0.7580150375939849, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 17602 + }, + { + "epoch": 0.7580751879699248, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 17603 + }, + { + "epoch": 0.7581353383458647, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 17604 + }, + { + "epoch": 0.7581954887218045, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 17605 + }, + { + "epoch": 0.7582556390977444, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 17606 + }, + { + "epoch": 0.7583157894736842, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 17607 + }, + { + "epoch": 0.758375939849624, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 17608 + }, + { + "epoch": 0.7584360902255639, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1868, + "step": 17609 + }, + { + "epoch": 0.7584962406015038, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 17610 + }, + { + "epoch": 0.7585563909774437, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 17611 + }, + { + "epoch": 0.7586165413533834, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 17612 + }, + { + "epoch": 0.7586766917293233, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 17613 + }, + { + "epoch": 0.7587368421052632, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 17614 + }, + { + "epoch": 0.758796992481203, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 17615 + }, + { + "epoch": 0.7588571428571429, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 17616 + }, + { + "epoch": 0.7589172932330827, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 17617 + }, + { + "epoch": 0.7589774436090225, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 17618 + }, + { + "epoch": 0.7590375939849624, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 17619 + }, + { + "epoch": 0.7590977443609023, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 17620 + }, + { + "epoch": 0.7591578947368421, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 17621 + }, + { + "epoch": 0.7592180451127819, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 17622 + }, + { + "epoch": 0.7592781954887218, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 17623 + }, + { + "epoch": 0.7593383458646616, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 17624 + }, + { + "epoch": 0.7593984962406015, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2645, + "step": 17625 + }, + { + "epoch": 0.7594586466165414, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1894, + "step": 17626 + }, + { + "epoch": 0.7595187969924811, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 17627 + }, + { + "epoch": 0.759578947368421, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 17628 + }, + { + "epoch": 0.7596390977443609, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 17629 + }, + { + "epoch": 0.7596992481203008, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 17630 + }, + { + "epoch": 0.7597593984962406, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 17631 + }, + { + "epoch": 0.7598195488721805, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 17632 + }, + { + "epoch": 0.7598796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 17633 + }, + { + "epoch": 0.7599398496240601, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 17634 + }, + { + "epoch": 0.76, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 17635 + }, + { + "epoch": 0.7600601503759399, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 17636 + }, + { + "epoch": 0.7601203007518798, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 17637 + }, + { + "epoch": 0.7601804511278195, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 17638 + }, + { + "epoch": 0.7602406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 17639 + }, + { + "epoch": 0.7603007518796993, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 17640 + }, + { + "epoch": 0.7603609022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 17641 + }, + { + "epoch": 0.760421052631579, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 17642 + }, + { + "epoch": 0.7604812030075188, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 17643 + }, + { + "epoch": 0.7605413533834586, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1909, + "step": 17644 + }, + { + "epoch": 0.7606015037593985, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 17645 + }, + { + "epoch": 0.7606616541353384, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 17646 + }, + { + "epoch": 0.7607218045112782, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 17647 + }, + { + "epoch": 0.760781954887218, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2493, + "step": 17648 + }, + { + "epoch": 0.7608421052631579, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 17649 + }, + { + "epoch": 0.7609022556390977, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 17650 + }, + { + "epoch": 0.7609624060150376, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 17651 + }, + { + "epoch": 0.7610225563909775, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 17652 + }, + { + "epoch": 0.7610827067669172, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 17653 + }, + { + "epoch": 0.7611428571428571, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 17654 + }, + { + "epoch": 0.761203007518797, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 17655 + }, + { + "epoch": 0.7612631578947369, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 17656 + }, + { + "epoch": 0.7613233082706767, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 17657 + }, + { + "epoch": 0.7613834586466165, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 17658 + }, + { + "epoch": 0.7614436090225564, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 17659 + }, + { + "epoch": 0.7615037593984962, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 17660 + }, + { + "epoch": 0.7615639097744361, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 17661 + }, + { + "epoch": 0.761624060150376, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.213, + "step": 17662 + }, + { + "epoch": 0.7616842105263157, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 17663 + }, + { + "epoch": 0.7617443609022556, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 17664 + }, + { + "epoch": 0.7618045112781955, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 17665 + }, + { + "epoch": 0.7618646616541354, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 17666 + }, + { + "epoch": 0.7619248120300752, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 17667 + }, + { + "epoch": 0.761984962406015, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 17668 + }, + { + "epoch": 0.7620451127819549, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 17669 + }, + { + "epoch": 0.7621052631578947, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 17670 + }, + { + "epoch": 0.7621654135338346, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 17671 + }, + { + "epoch": 0.7622255639097745, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 17672 + }, + { + "epoch": 0.7622857142857142, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1333, + "step": 17673 + }, + { + "epoch": 0.7623458646616541, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1721, + "step": 17674 + }, + { + "epoch": 0.762406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 17675 + }, + { + "epoch": 0.7624661654135338, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 17676 + }, + { + "epoch": 0.7625263157894737, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 17677 + }, + { + "epoch": 0.7625864661654135, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 17678 + }, + { + "epoch": 0.7626466165413534, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 17679 + }, + { + "epoch": 0.7627067669172932, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 17680 + }, + { + "epoch": 0.7627669172932331, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 17681 + }, + { + "epoch": 0.762827067669173, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 17682 + }, + { + "epoch": 0.7628872180451128, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 17683 + }, + { + "epoch": 0.7629473684210526, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 17684 + }, + { + "epoch": 0.7630075187969925, + "grad_norm": 1.8203125, + "learning_rate": 3e-05, + "loss": 1.2192, + "step": 17685 + }, + { + "epoch": 0.7630676691729323, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 17686 + }, + { + "epoch": 0.7631278195488722, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 17687 + }, + { + "epoch": 0.7631879699248121, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 17688 + }, + { + "epoch": 0.7632481203007518, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 17689 + }, + { + "epoch": 0.7633082706766917, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 17690 + }, + { + "epoch": 0.7633684210526316, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 17691 + }, + { + "epoch": 0.7634285714285715, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.148, + "step": 17692 + }, + { + "epoch": 0.7634887218045113, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 17693 + }, + { + "epoch": 0.7635488721804511, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 17694 + }, + { + "epoch": 0.763609022556391, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 17695 + }, + { + "epoch": 0.7636691729323308, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 17696 + }, + { + "epoch": 0.7637293233082707, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 17697 + }, + { + "epoch": 0.7637894736842106, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 17698 + }, + { + "epoch": 0.7638496240601503, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 17699 + }, + { + "epoch": 0.7639097744360902, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 17700 + }, + { + "epoch": 0.7639699248120301, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 17701 + }, + { + "epoch": 0.76403007518797, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 17702 + }, + { + "epoch": 0.7640902255639098, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 17703 + }, + { + "epoch": 0.7641503759398496, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 17704 + }, + { + "epoch": 0.7642105263157895, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 17705 + }, + { + "epoch": 0.7642706766917293, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 17706 + }, + { + "epoch": 0.7643308270676692, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 17707 + }, + { + "epoch": 0.7643909774436091, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 17708 + }, + { + "epoch": 0.7644511278195488, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 17709 + }, + { + "epoch": 0.7645112781954887, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2898, + "step": 17710 + }, + { + "epoch": 0.7645714285714286, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 17711 + }, + { + "epoch": 0.7646315789473684, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2489, + "step": 17712 + }, + { + "epoch": 0.7646917293233083, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 17713 + }, + { + "epoch": 0.7647518796992481, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 17714 + }, + { + "epoch": 0.7648120300751879, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 17715 + }, + { + "epoch": 0.7648721804511278, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 17716 + }, + { + "epoch": 0.7649323308270677, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 17717 + }, + { + "epoch": 0.7649924812030076, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 17718 + }, + { + "epoch": 0.7650526315789473, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 17719 + }, + { + "epoch": 0.7651127819548872, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 17720 + }, + { + "epoch": 0.7651729323308271, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 17721 + }, + { + "epoch": 0.7652330827067669, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 17722 + }, + { + "epoch": 0.7652932330827068, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 17723 + }, + { + "epoch": 0.7653533834586466, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 17724 + }, + { + "epoch": 0.7654135338345864, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.3026, + "step": 17725 + }, + { + "epoch": 0.7654736842105263, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 17726 + }, + { + "epoch": 0.7655338345864662, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 17727 + }, + { + "epoch": 0.765593984962406, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 17728 + }, + { + "epoch": 0.7656541353383458, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1936, + "step": 17729 + }, + { + "epoch": 0.7657142857142857, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 17730 + }, + { + "epoch": 0.7657744360902256, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 17731 + }, + { + "epoch": 0.7658345864661654, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 17732 + }, + { + "epoch": 0.7658947368421053, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 17733 + }, + { + "epoch": 0.7659548872180452, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 17734 + }, + { + "epoch": 0.7660150375939849, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 17735 + }, + { + "epoch": 0.7660751879699248, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 17736 + }, + { + "epoch": 0.7661353383458647, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 17737 + }, + { + "epoch": 0.7661954887218045, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 17738 + }, + { + "epoch": 0.7662556390977444, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 17739 + }, + { + "epoch": 0.7663157894736842, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 17740 + }, + { + "epoch": 0.766375939849624, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 17741 + }, + { + "epoch": 0.7664360902255639, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1925, + "step": 17742 + }, + { + "epoch": 0.7664962406015038, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 17743 + }, + { + "epoch": 0.7665563909774437, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 17744 + }, + { + "epoch": 0.7666165413533834, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 17745 + }, + { + "epoch": 0.7666766917293233, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 17746 + }, + { + "epoch": 0.7667368421052632, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 17747 + }, + { + "epoch": 0.766796992481203, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1689, + "step": 17748 + }, + { + "epoch": 0.7668571428571429, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 17749 + }, + { + "epoch": 0.7669172932330827, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 17750 + }, + { + "epoch": 0.7669774436090225, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 17751 + }, + { + "epoch": 0.7670375939849624, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 17752 + }, + { + "epoch": 0.7670977443609023, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 17753 + }, + { + "epoch": 0.7671578947368421, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 17754 + }, + { + "epoch": 0.7672180451127819, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 17755 + }, + { + "epoch": 0.7672781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 17756 + }, + { + "epoch": 0.7673383458646617, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 17757 + }, + { + "epoch": 0.7673984962406015, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 17758 + }, + { + "epoch": 0.7674586466165414, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 17759 + }, + { + "epoch": 0.7675187969924812, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 17760 + }, + { + "epoch": 0.767578947368421, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 17761 + }, + { + "epoch": 0.7676390977443609, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 17762 + }, + { + "epoch": 0.7676992481203008, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2069, + "step": 17763 + }, + { + "epoch": 0.7677593984962406, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 17764 + }, + { + "epoch": 0.7678195488721804, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2235, + "step": 17765 + }, + { + "epoch": 0.7678796992481203, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 17766 + }, + { + "epoch": 0.7679398496240601, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1707, + "step": 17767 + }, + { + "epoch": 0.768, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 17768 + }, + { + "epoch": 0.7680601503759399, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 17769 + }, + { + "epoch": 0.7681203007518796, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 17770 + }, + { + "epoch": 0.7681804511278195, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 17771 + }, + { + "epoch": 0.7682406015037594, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 17772 + }, + { + "epoch": 0.7683007518796993, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 17773 + }, + { + "epoch": 0.7683609022556391, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 17774 + }, + { + "epoch": 0.7684210526315789, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 17775 + }, + { + "epoch": 0.7684812030075188, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 17776 + }, + { + "epoch": 0.7685413533834586, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 17777 + }, + { + "epoch": 0.7686015037593985, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 17778 + }, + { + "epoch": 0.7686616541353384, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 17779 + }, + { + "epoch": 0.7687218045112783, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 17780 + }, + { + "epoch": 0.768781954887218, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 17781 + }, + { + "epoch": 0.7688421052631579, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 17782 + }, + { + "epoch": 0.7689022556390978, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 17783 + }, + { + "epoch": 0.7689624060150376, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 17784 + }, + { + "epoch": 0.7690225563909775, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 17785 + }, + { + "epoch": 0.7690827067669173, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 17786 + }, + { + "epoch": 0.7691428571428571, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 17787 + }, + { + "epoch": 0.769203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 17788 + }, + { + "epoch": 0.7692631578947369, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 17789 + }, + { + "epoch": 0.7693233082706767, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 17790 + }, + { + "epoch": 0.7693834586466165, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 17791 + }, + { + "epoch": 0.7694436090225564, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2163, + "step": 17792 + }, + { + "epoch": 0.7695037593984962, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 17793 + }, + { + "epoch": 0.7695639097744361, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 17794 + }, + { + "epoch": 0.769624060150376, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 17795 + }, + { + "epoch": 0.7696842105263157, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 17796 + }, + { + "epoch": 0.7697443609022556, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 17797 + }, + { + "epoch": 0.7698045112781955, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 17798 + }, + { + "epoch": 0.7698646616541354, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 17799 + }, + { + "epoch": 0.7699248120300752, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.163, + "step": 17800 + }, + { + "epoch": 0.769984962406015, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 17801 + }, + { + "epoch": 0.7700451127819549, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 17802 + }, + { + "epoch": 0.7701052631578947, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 17803 + }, + { + "epoch": 0.7701654135338346, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 17804 + }, + { + "epoch": 0.7702255639097745, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 17805 + }, + { + "epoch": 0.7702857142857142, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.3964, + "step": 17806 + }, + { + "epoch": 0.7703458646616541, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 17807 + }, + { + "epoch": 0.770406015037594, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 17808 + }, + { + "epoch": 0.7704661654135339, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 17809 + }, + { + "epoch": 0.7705263157894737, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 17810 + }, + { + "epoch": 0.7705864661654135, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 17811 + }, + { + "epoch": 0.7706466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 17812 + }, + { + "epoch": 0.7707067669172932, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 17813 + }, + { + "epoch": 0.7707669172932331, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2571, + "step": 17814 + }, + { + "epoch": 0.770827067669173, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 17815 + }, + { + "epoch": 0.7708872180451127, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 17816 + }, + { + "epoch": 0.7709473684210526, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 17817 + }, + { + "epoch": 0.7710075187969925, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 17818 + }, + { + "epoch": 0.7710676691729323, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 17819 + }, + { + "epoch": 0.7711278195488722, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 17820 + }, + { + "epoch": 0.771187969924812, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 17821 + }, + { + "epoch": 0.7712481203007518, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 17822 + }, + { + "epoch": 0.7713082706766917, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 17823 + }, + { + "epoch": 0.7713684210526316, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 17824 + }, + { + "epoch": 0.7714285714285715, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 17825 + }, + { + "epoch": 0.7714887218045112, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2097, + "step": 17826 + }, + { + "epoch": 0.7715488721804511, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 17827 + }, + { + "epoch": 0.771609022556391, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 17828 + }, + { + "epoch": 0.7716691729323308, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 17829 + }, + { + "epoch": 0.7717293233082707, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 17830 + }, + { + "epoch": 0.7717894736842106, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 17831 + }, + { + "epoch": 0.7718496240601503, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 17832 + }, + { + "epoch": 0.7719097744360902, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 17833 + }, + { + "epoch": 0.7719699248120301, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 17834 + }, + { + "epoch": 0.77203007518797, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 17835 + }, + { + "epoch": 0.7720902255639098, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 17836 + }, + { + "epoch": 0.7721503759398496, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 17837 + }, + { + "epoch": 0.7722105263157895, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 17838 + }, + { + "epoch": 0.7722706766917293, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 17839 + }, + { + "epoch": 0.7723308270676692, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 17840 + }, + { + "epoch": 0.7723909774436091, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2359, + "step": 17841 + }, + { + "epoch": 0.7724511278195488, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 17842 + }, + { + "epoch": 0.7725112781954887, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 17843 + }, + { + "epoch": 0.7725714285714286, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 17844 + }, + { + "epoch": 0.7726315789473684, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 17845 + }, + { + "epoch": 0.7726917293233083, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.2512, + "step": 17846 + }, + { + "epoch": 0.7727518796992481, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 17847 + }, + { + "epoch": 0.772812030075188, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 17848 + }, + { + "epoch": 0.7728721804511278, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 17849 + }, + { + "epoch": 0.7729323308270677, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 17850 + }, + { + "epoch": 0.7729924812030076, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1805, + "step": 17851 + }, + { + "epoch": 0.7730526315789473, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 17852 + }, + { + "epoch": 0.7731127819548872, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 17853 + }, + { + "epoch": 0.7731729323308271, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 17854 + }, + { + "epoch": 0.7732330827067669, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 17855 + }, + { + "epoch": 0.7732932330827068, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 17856 + }, + { + "epoch": 0.7733533834586466, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 17857 + }, + { + "epoch": 0.7734135338345864, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 17858 + }, + { + "epoch": 0.7734736842105263, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 17859 + }, + { + "epoch": 0.7735338345864662, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 17860 + }, + { + "epoch": 0.7735939849624061, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 17861 + }, + { + "epoch": 0.7736541353383458, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 17862 + }, + { + "epoch": 0.7737142857142857, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 17863 + }, + { + "epoch": 0.7737744360902256, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 17864 + }, + { + "epoch": 0.7738345864661654, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 17865 + }, + { + "epoch": 0.7738947368421053, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 17866 + }, + { + "epoch": 0.7739548872180451, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 17867 + }, + { + "epoch": 0.7740150375939849, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 17868 + }, + { + "epoch": 0.7740751879699248, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 17869 + }, + { + "epoch": 0.7741353383458647, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 17870 + }, + { + "epoch": 0.7741954887218045, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 17871 + }, + { + "epoch": 0.7742556390977443, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 17872 + }, + { + "epoch": 0.7743157894736842, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 17873 + }, + { + "epoch": 0.774375939849624, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 17874 + }, + { + "epoch": 0.7744360902255639, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 17875 + }, + { + "epoch": 0.7744962406015038, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 17876 + }, + { + "epoch": 0.7745563909774436, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2, + "step": 17877 + }, + { + "epoch": 0.7746165413533834, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 17878 + }, + { + "epoch": 0.7746766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 17879 + }, + { + "epoch": 0.7747368421052632, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 17880 + }, + { + "epoch": 0.774796992481203, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 17881 + }, + { + "epoch": 0.7748571428571429, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 17882 + }, + { + "epoch": 0.7749172932330827, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 17883 + }, + { + "epoch": 0.7749774436090225, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2838, + "step": 17884 + }, + { + "epoch": 0.7750375939849624, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 17885 + }, + { + "epoch": 0.7750977443609023, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 17886 + }, + { + "epoch": 0.7751578947368422, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 17887 + }, + { + "epoch": 0.7752180451127819, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 17888 + }, + { + "epoch": 0.7752781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 17889 + }, + { + "epoch": 0.7753383458646617, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1446, + "step": 17890 + }, + { + "epoch": 0.7753984962406015, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.2296, + "step": 17891 + }, + { + "epoch": 0.7754586466165414, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 17892 + }, + { + "epoch": 0.7755187969924812, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 17893 + }, + { + "epoch": 0.775578947368421, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2627, + "step": 17894 + }, + { + "epoch": 0.7756390977443609, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 17895 + }, + { + "epoch": 0.7756992481203008, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 17896 + }, + { + "epoch": 0.7757593984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 17897 + }, + { + "epoch": 0.7758195488721804, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 17898 + }, + { + "epoch": 0.7758796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2034, + "step": 17899 + }, + { + "epoch": 0.7759398496240602, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 17900 + }, + { + "epoch": 0.776, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2045, + "step": 17901 + }, + { + "epoch": 0.7760601503759399, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 17902 + }, + { + "epoch": 0.7761203007518797, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 17903 + }, + { + "epoch": 0.7761804511278195, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 17904 + }, + { + "epoch": 0.7762406015037594, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 17905 + }, + { + "epoch": 0.7763007518796993, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 17906 + }, + { + "epoch": 0.7763609022556391, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 17907 + }, + { + "epoch": 0.7764210526315789, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2453, + "step": 17908 + }, + { + "epoch": 0.7764812030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 17909 + }, + { + "epoch": 0.7765413533834586, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 17910 + }, + { + "epoch": 0.7766015037593985, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 17911 + }, + { + "epoch": 0.7766616541353384, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 17912 + }, + { + "epoch": 0.7767218045112781, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 17913 + }, + { + "epoch": 0.776781954887218, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 17914 + }, + { + "epoch": 0.7768421052631579, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 17915 + }, + { + "epoch": 0.7769022556390978, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 17916 + }, + { + "epoch": 0.7769624060150376, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2698, + "step": 17917 + }, + { + "epoch": 0.7770225563909774, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 17918 + }, + { + "epoch": 0.7770827067669173, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 17919 + }, + { + "epoch": 0.7771428571428571, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 17920 + }, + { + "epoch": 0.777203007518797, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 17921 + }, + { + "epoch": 0.7772631578947369, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 17922 + }, + { + "epoch": 0.7773233082706766, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 17923 + }, + { + "epoch": 0.7773834586466165, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 17924 + }, + { + "epoch": 0.7774436090225564, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 17925 + }, + { + "epoch": 0.7775037593984963, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2774, + "step": 17926 + }, + { + "epoch": 0.7775639097744361, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 17927 + }, + { + "epoch": 0.777624060150376, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 17928 + }, + { + "epoch": 0.7776842105263158, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 17929 + }, + { + "epoch": 0.7777443609022556, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 17930 + }, + { + "epoch": 0.7778045112781955, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1333, + "step": 17931 + }, + { + "epoch": 0.7778646616541354, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 17932 + }, + { + "epoch": 0.7779248120300752, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 17933 + }, + { + "epoch": 0.777984962406015, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 17934 + }, + { + "epoch": 0.7780451127819549, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.188, + "step": 17935 + }, + { + "epoch": 0.7781052631578947, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 17936 + }, + { + "epoch": 0.7781654135338346, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 17937 + }, + { + "epoch": 0.7782255639097745, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2633, + "step": 17938 + }, + { + "epoch": 0.7782857142857142, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 17939 + }, + { + "epoch": 0.7783458646616541, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2156, + "step": 17940 + }, + { + "epoch": 0.778406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 17941 + }, + { + "epoch": 0.7784661654135339, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.2866, + "step": 17942 + }, + { + "epoch": 0.7785263157894737, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 17943 + }, + { + "epoch": 0.7785864661654135, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 17944 + }, + { + "epoch": 0.7786466165413534, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1839, + "step": 17945 + }, + { + "epoch": 0.7787067669172932, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2619, + "step": 17946 + }, + { + "epoch": 0.7787669172932331, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.2089, + "step": 17947 + }, + { + "epoch": 0.778827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.185, + "step": 17948 + }, + { + "epoch": 0.7788872180451127, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 17949 + }, + { + "epoch": 0.7789473684210526, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 17950 + }, + { + "epoch": 0.7790075187969925, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 17951 + }, + { + "epoch": 0.7790676691729324, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 17952 + }, + { + "epoch": 0.7791278195488722, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 17953 + }, + { + "epoch": 0.779187969924812, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 17954 + }, + { + "epoch": 0.7792481203007519, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 17955 + }, + { + "epoch": 0.7793082706766917, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 17956 + }, + { + "epoch": 0.7793684210526316, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 17957 + }, + { + "epoch": 0.7794285714285715, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 17958 + }, + { + "epoch": 0.7794887218045112, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 17959 + }, + { + "epoch": 0.7795488721804511, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 17960 + }, + { + "epoch": 0.779609022556391, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 17961 + }, + { + "epoch": 0.7796691729323308, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 17962 + }, + { + "epoch": 0.7797293233082707, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 17963 + }, + { + "epoch": 0.7797894736842105, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 17964 + }, + { + "epoch": 0.7798496240601503, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 17965 + }, + { + "epoch": 0.7799097744360902, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 17966 + }, + { + "epoch": 0.7799699248120301, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 17967 + }, + { + "epoch": 0.78003007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 17968 + }, + { + "epoch": 0.7800902255639097, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 17969 + }, + { + "epoch": 0.7801503759398496, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 17970 + }, + { + "epoch": 0.7802105263157895, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 17971 + }, + { + "epoch": 0.7802706766917293, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 17972 + }, + { + "epoch": 0.7803308270676692, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 17973 + }, + { + "epoch": 0.780390977443609, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 17974 + }, + { + "epoch": 0.7804511278195488, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 17975 + }, + { + "epoch": 0.7805112781954887, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 17976 + }, + { + "epoch": 0.7805714285714286, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 17977 + }, + { + "epoch": 0.7806315789473685, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 17978 + }, + { + "epoch": 0.7806917293233083, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 17979 + }, + { + "epoch": 0.7807518796992481, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 17980 + }, + { + "epoch": 0.780812030075188, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 17981 + }, + { + "epoch": 0.7808721804511278, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.3307, + "step": 17982 + }, + { + "epoch": 0.7809323308270677, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 17983 + }, + { + "epoch": 0.7809924812030076, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 17984 + }, + { + "epoch": 0.7810526315789473, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 17985 + }, + { + "epoch": 0.7811127819548872, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 17986 + }, + { + "epoch": 0.7811729323308271, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 17987 + }, + { + "epoch": 0.781233082706767, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 17988 + }, + { + "epoch": 0.7812932330827068, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 17989 + }, + { + "epoch": 0.7813533834586466, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 17990 + }, + { + "epoch": 0.7814135338345864, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2065, + "step": 17991 + }, + { + "epoch": 0.7814736842105263, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 17992 + }, + { + "epoch": 0.7815338345864662, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 17993 + }, + { + "epoch": 0.7815939849624061, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 17994 + }, + { + "epoch": 0.7816541353383458, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 17995 + }, + { + "epoch": 0.7817142857142857, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 17996 + }, + { + "epoch": 0.7817744360902256, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 17997 + }, + { + "epoch": 0.7818345864661654, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 17998 + }, + { + "epoch": 0.7818947368421053, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 17999 + }, + { + "epoch": 0.7819548872180451, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 18000 + }, + { + "epoch": 0.7820150375939849, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 18001 + }, + { + "epoch": 0.7820751879699248, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.248, + "step": 18002 + }, + { + "epoch": 0.7821353383458647, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2171, + "step": 18003 + }, + { + "epoch": 0.7821954887218046, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 18004 + }, + { + "epoch": 0.7822556390977443, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 18005 + }, + { + "epoch": 0.7823157894736842, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 18006 + }, + { + "epoch": 0.7823759398496241, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 18007 + }, + { + "epoch": 0.7824360902255639, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 18008 + }, + { + "epoch": 0.7824962406015038, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 18009 + }, + { + "epoch": 0.7825563909774436, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 18010 + }, + { + "epoch": 0.7826165413533834, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 18011 + }, + { + "epoch": 0.7826766917293233, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 18012 + }, + { + "epoch": 0.7827368421052632, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 18013 + }, + { + "epoch": 0.782796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 18014 + }, + { + "epoch": 0.7828571428571428, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 18015 + }, + { + "epoch": 0.7829172932330827, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 18016 + }, + { + "epoch": 0.7829774436090225, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 18017 + }, + { + "epoch": 0.7830375939849624, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 18018 + }, + { + "epoch": 0.7830977443609023, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 18019 + }, + { + "epoch": 0.783157894736842, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 18020 + }, + { + "epoch": 0.7832180451127819, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 18021 + }, + { + "epoch": 0.7832781954887218, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 18022 + }, + { + "epoch": 0.7833383458646617, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2588, + "step": 18023 + }, + { + "epoch": 0.7833984962406015, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 18024 + }, + { + "epoch": 0.7834586466165413, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 18025 + }, + { + "epoch": 0.7835187969924812, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 18026 + }, + { + "epoch": 0.783578947368421, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 18027 + }, + { + "epoch": 0.7836390977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 18028 + }, + { + "epoch": 0.7836992481203008, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 18029 + }, + { + "epoch": 0.7837593984962407, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 18030 + }, + { + "epoch": 0.7838195488721804, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.146, + "step": 18031 + }, + { + "epoch": 0.7838796992481203, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 18032 + }, + { + "epoch": 0.7839398496240602, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1774, + "step": 18033 + }, + { + "epoch": 0.784, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 18034 + }, + { + "epoch": 0.7840601503759399, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 18035 + }, + { + "epoch": 0.7841203007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 18036 + }, + { + "epoch": 0.7841804511278195, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 18037 + }, + { + "epoch": 0.7842406015037594, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 18038 + }, + { + "epoch": 0.7843007518796993, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 18039 + }, + { + "epoch": 0.7843609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.3359, + "step": 18040 + }, + { + "epoch": 0.7844210526315789, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 18041 + }, + { + "epoch": 0.7844812030075188, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 18042 + }, + { + "epoch": 0.7845413533834587, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 18043 + }, + { + "epoch": 0.7846015037593985, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 18044 + }, + { + "epoch": 0.7846616541353384, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 18045 + }, + { + "epoch": 0.7847218045112782, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 18046 + }, + { + "epoch": 0.784781954887218, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 18047 + }, + { + "epoch": 0.7848421052631579, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 18048 + }, + { + "epoch": 0.7849022556390978, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1861, + "step": 18049 + }, + { + "epoch": 0.7849624060150376, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 18050 + }, + { + "epoch": 0.7850225563909774, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 18051 + }, + { + "epoch": 0.7850827067669173, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 18052 + }, + { + "epoch": 0.7851428571428571, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 18053 + }, + { + "epoch": 0.785203007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 18054 + }, + { + "epoch": 0.7852631578947369, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 18055 + }, + { + "epoch": 0.7853233082706766, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 18056 + }, + { + "epoch": 0.7853834586466165, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 18057 + }, + { + "epoch": 0.7854436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.23, + "step": 18058 + }, + { + "epoch": 0.7855037593984963, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 18059 + }, + { + "epoch": 0.7855639097744361, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2091, + "step": 18060 + }, + { + "epoch": 0.7856240601503759, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 18061 + }, + { + "epoch": 0.7856842105263158, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 18062 + }, + { + "epoch": 0.7857443609022556, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 18063 + }, + { + "epoch": 0.7858045112781955, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 18064 + }, + { + "epoch": 0.7858646616541354, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 18065 + }, + { + "epoch": 0.7859248120300751, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 18066 + }, + { + "epoch": 0.785984962406015, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 18067 + }, + { + "epoch": 0.7860451127819549, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 18068 + }, + { + "epoch": 0.7861052631578948, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 18069 + }, + { + "epoch": 0.7861654135338346, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 18070 + }, + { + "epoch": 0.7862255639097744, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 18071 + }, + { + "epoch": 0.7862857142857143, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 18072 + }, + { + "epoch": 0.7863458646616541, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1467, + "step": 18073 + }, + { + "epoch": 0.786406015037594, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 18074 + }, + { + "epoch": 0.7864661654135339, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 18075 + }, + { + "epoch": 0.7865263157894736, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 18076 + }, + { + "epoch": 0.7865864661654135, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 18077 + }, + { + "epoch": 0.7866466165413534, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 18078 + }, + { + "epoch": 0.7867067669172932, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 18079 + }, + { + "epoch": 0.7867669172932331, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 18080 + }, + { + "epoch": 0.786827067669173, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 18081 + }, + { + "epoch": 0.7868872180451127, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 18082 + }, + { + "epoch": 0.7869473684210526, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 18083 + }, + { + "epoch": 0.7870075187969925, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 18084 + }, + { + "epoch": 0.7870676691729324, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 18085 + }, + { + "epoch": 0.7871278195488722, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 18086 + }, + { + "epoch": 0.787187969924812, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 18087 + }, + { + "epoch": 0.7872481203007519, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 18088 + }, + { + "epoch": 0.7873082706766917, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 18089 + }, + { + "epoch": 0.7873684210526316, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2308, + "step": 18090 + }, + { + "epoch": 0.7874285714285715, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 18091 + }, + { + "epoch": 0.7874887218045112, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 18092 + }, + { + "epoch": 0.7875488721804511, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 18093 + }, + { + "epoch": 0.787609022556391, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 18094 + }, + { + "epoch": 0.7876691729323309, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 18095 + }, + { + "epoch": 0.7877293233082707, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 18096 + }, + { + "epoch": 0.7877894736842105, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.3164, + "step": 18097 + }, + { + "epoch": 0.7878496240601504, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.2363, + "step": 18098 + }, + { + "epoch": 0.7879097744360902, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 18099 + }, + { + "epoch": 0.7879699248120301, + "grad_norm": 1.4921875, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 18100 + }, + { + "epoch": 0.78803007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 18101 + }, + { + "epoch": 0.7880902255639097, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 18102 + }, + { + "epoch": 0.7881503759398496, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 18103 + }, + { + "epoch": 0.7882105263157895, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 18104 + }, + { + "epoch": 0.7882706766917293, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 18105 + }, + { + "epoch": 0.7883308270676692, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 18106 + }, + { + "epoch": 0.788390977443609, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 18107 + }, + { + "epoch": 0.7884511278195488, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 18108 + }, + { + "epoch": 0.7885112781954887, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 18109 + }, + { + "epoch": 0.7885714285714286, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 18110 + }, + { + "epoch": 0.7886315789473685, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 18111 + }, + { + "epoch": 0.7886917293233082, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 18112 + }, + { + "epoch": 0.7887518796992481, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 18113 + }, + { + "epoch": 0.788812030075188, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 18114 + }, + { + "epoch": 0.7888721804511278, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 18115 + }, + { + "epoch": 0.7889323308270677, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 18116 + }, + { + "epoch": 0.7889924812030075, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 18117 + }, + { + "epoch": 0.7890526315789473, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 18118 + }, + { + "epoch": 0.7891127819548872, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 18119 + }, + { + "epoch": 0.7891729323308271, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 18120 + }, + { + "epoch": 0.789233082706767, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 18121 + }, + { + "epoch": 0.7892932330827067, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 18122 + }, + { + "epoch": 0.7893533834586466, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 18123 + }, + { + "epoch": 0.7894135338345865, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1429, + "step": 18124 + }, + { + "epoch": 0.7894736842105263, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 18125 + }, + { + "epoch": 0.7895338345864662, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 18126 + }, + { + "epoch": 0.7895939849624061, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2244, + "step": 18127 + }, + { + "epoch": 0.7896541353383458, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 18128 + }, + { + "epoch": 0.7897142857142857, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 18129 + }, + { + "epoch": 0.7897744360902256, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 18130 + }, + { + "epoch": 0.7898345864661654, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 18131 + }, + { + "epoch": 0.7898947368421053, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 18132 + }, + { + "epoch": 0.7899548872180451, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 18133 + }, + { + "epoch": 0.790015037593985, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 18134 + }, + { + "epoch": 0.7900751879699248, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 18135 + }, + { + "epoch": 0.7901353383458647, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2757, + "step": 18136 + }, + { + "epoch": 0.7901954887218046, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1994, + "step": 18137 + }, + { + "epoch": 0.7902556390977443, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 18138 + }, + { + "epoch": 0.7903157894736842, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 18139 + }, + { + "epoch": 0.7903759398496241, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 18140 + }, + { + "epoch": 0.7904360902255639, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 18141 + }, + { + "epoch": 0.7904962406015038, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.3312, + "step": 18142 + }, + { + "epoch": 0.7905563909774436, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 18143 + }, + { + "epoch": 0.7906165413533834, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 18144 + }, + { + "epoch": 0.7906766917293233, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 18145 + }, + { + "epoch": 0.7907368421052632, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 18146 + }, + { + "epoch": 0.790796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 18147 + }, + { + "epoch": 0.7908571428571428, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 18148 + }, + { + "epoch": 0.7909172932330827, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 18149 + }, + { + "epoch": 0.7909774436090226, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1922, + "step": 18150 + }, + { + "epoch": 0.7910375939849624, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 18151 + }, + { + "epoch": 0.7910977443609023, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 18152 + }, + { + "epoch": 0.7911578947368421, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 18153 + }, + { + "epoch": 0.7912180451127819, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 18154 + }, + { + "epoch": 0.7912781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 18155 + }, + { + "epoch": 0.7913383458646617, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 18156 + }, + { + "epoch": 0.7913984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 18157 + }, + { + "epoch": 0.7914586466165413, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 18158 + }, + { + "epoch": 0.7915187969924812, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 18159 + }, + { + "epoch": 0.791578947368421, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 18160 + }, + { + "epoch": 0.7916390977443609, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 18161 + }, + { + "epoch": 0.7916992481203008, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 18162 + }, + { + "epoch": 0.7917593984962406, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 18163 + }, + { + "epoch": 0.7918195488721804, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 18164 + }, + { + "epoch": 0.7918796992481203, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 18165 + }, + { + "epoch": 0.7919398496240602, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 18166 + }, + { + "epoch": 0.792, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 18167 + }, + { + "epoch": 0.7920601503759398, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 18168 + }, + { + "epoch": 0.7921203007518797, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 18169 + }, + { + "epoch": 0.7921804511278195, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 18170 + }, + { + "epoch": 0.7922406015037594, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 18171 + }, + { + "epoch": 0.7923007518796993, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 18172 + }, + { + "epoch": 0.792360902255639, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 18173 + }, + { + "epoch": 0.7924210526315789, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 18174 + }, + { + "epoch": 0.7924812030075188, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 18175 + }, + { + "epoch": 0.7925413533834587, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 18176 + }, + { + "epoch": 0.7926015037593985, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 18177 + }, + { + "epoch": 0.7926616541353384, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 18178 + }, + { + "epoch": 0.7927218045112782, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 18179 + }, + { + "epoch": 0.792781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 18180 + }, + { + "epoch": 0.7928421052631579, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 18181 + }, + { + "epoch": 0.7929022556390978, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 18182 + }, + { + "epoch": 0.7929624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 18183 + }, + { + "epoch": 0.7930225563909774, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 18184 + }, + { + "epoch": 0.7930827067669173, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 18185 + }, + { + "epoch": 0.7931428571428571, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1727, + "step": 18186 + }, + { + "epoch": 0.793203007518797, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2178, + "step": 18187 + }, + { + "epoch": 0.7932631578947369, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 18188 + }, + { + "epoch": 0.7933233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 18189 + }, + { + "epoch": 0.7933834586466165, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 18190 + }, + { + "epoch": 0.7934436090225564, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 18191 + }, + { + "epoch": 0.7935037593984963, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 18192 + }, + { + "epoch": 0.7935639097744361, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 18193 + }, + { + "epoch": 0.7936240601503759, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 18194 + }, + { + "epoch": 0.7936842105263158, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 18195 + }, + { + "epoch": 0.7937443609022556, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 18196 + }, + { + "epoch": 0.7938045112781955, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 18197 + }, + { + "epoch": 0.7938646616541354, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 18198 + }, + { + "epoch": 0.7939248120300751, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 18199 + }, + { + "epoch": 0.793984962406015, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 18200 + }, + { + "epoch": 0.7940451127819549, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 18201 + }, + { + "epoch": 0.7941052631578948, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 18202 + }, + { + "epoch": 0.7941654135338346, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 18203 + }, + { + "epoch": 0.7942255639097744, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 18204 + }, + { + "epoch": 0.7942857142857143, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 18205 + }, + { + "epoch": 0.7943458646616541, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 18206 + }, + { + "epoch": 0.794406015037594, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 18207 + }, + { + "epoch": 0.7944661654135339, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 18208 + }, + { + "epoch": 0.7945263157894736, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 18209 + }, + { + "epoch": 0.7945864661654135, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 18210 + }, + { + "epoch": 0.7946466165413534, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 18211 + }, + { + "epoch": 0.7947067669172933, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 18212 + }, + { + "epoch": 0.7947669172932331, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2923, + "step": 18213 + }, + { + "epoch": 0.7948270676691729, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 18214 + }, + { + "epoch": 0.7948872180451128, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 18215 + }, + { + "epoch": 0.7949473684210526, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2018, + "step": 18216 + }, + { + "epoch": 0.7950075187969925, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 18217 + }, + { + "epoch": 0.7950676691729324, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 18218 + }, + { + "epoch": 0.7951278195488721, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 18219 + }, + { + "epoch": 0.795187969924812, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 18220 + }, + { + "epoch": 0.7952481203007519, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 18221 + }, + { + "epoch": 0.7953082706766917, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 18222 + }, + { + "epoch": 0.7953684210526316, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 18223 + }, + { + "epoch": 0.7954285714285714, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 18224 + }, + { + "epoch": 0.7954887218045112, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 18225 + }, + { + "epoch": 0.7955488721804511, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 18226 + }, + { + "epoch": 0.795609022556391, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 18227 + }, + { + "epoch": 0.7956691729323309, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 18228 + }, + { + "epoch": 0.7957293233082707, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 18229 + }, + { + "epoch": 0.7957894736842105, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 18230 + }, + { + "epoch": 0.7958496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 18231 + }, + { + "epoch": 0.7959097744360902, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 18232 + }, + { + "epoch": 0.7959699248120301, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 18233 + }, + { + "epoch": 0.79603007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 18234 + }, + { + "epoch": 0.7960902255639097, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 18235 + }, + { + "epoch": 0.7961503759398496, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2328, + "step": 18236 + }, + { + "epoch": 0.7962105263157895, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 18237 + }, + { + "epoch": 0.7962706766917294, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1992, + "step": 18238 + }, + { + "epoch": 0.7963308270676692, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 18239 + }, + { + "epoch": 0.796390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 18240 + }, + { + "epoch": 0.7964511278195489, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.233, + "step": 18241 + }, + { + "epoch": 0.7965112781954887, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 18242 + }, + { + "epoch": 0.7965714285714286, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 18243 + }, + { + "epoch": 0.7966315789473685, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 18244 + }, + { + "epoch": 0.7966917293233082, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 18245 + }, + { + "epoch": 0.7967518796992481, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 18246 + }, + { + "epoch": 0.796812030075188, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 18247 + }, + { + "epoch": 0.7968721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 18248 + }, + { + "epoch": 0.7969323308270677, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 18249 + }, + { + "epoch": 0.7969924812030075, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 18250 + }, + { + "epoch": 0.7970526315789473, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 18251 + }, + { + "epoch": 0.7971127819548872, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2665, + "step": 18252 + }, + { + "epoch": 0.7971729323308271, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 18253 + }, + { + "epoch": 0.797233082706767, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 18254 + }, + { + "epoch": 0.7972932330827067, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 18255 + }, + { + "epoch": 0.7973533834586466, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2013, + "step": 18256 + }, + { + "epoch": 0.7974135338345865, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 18257 + }, + { + "epoch": 0.7974736842105263, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 18258 + }, + { + "epoch": 0.7975338345864662, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 18259 + }, + { + "epoch": 0.797593984962406, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 18260 + }, + { + "epoch": 0.7976541353383458, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 18261 + }, + { + "epoch": 0.7977142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 18262 + }, + { + "epoch": 0.7977744360902256, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 18263 + }, + { + "epoch": 0.7978345864661655, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 18264 + }, + { + "epoch": 0.7978947368421052, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 18265 + }, + { + "epoch": 0.7979548872180451, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 18266 + }, + { + "epoch": 0.798015037593985, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 18267 + }, + { + "epoch": 0.7980751879699248, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 18268 + }, + { + "epoch": 0.7981353383458647, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 18269 + }, + { + "epoch": 0.7981954887218045, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 18270 + }, + { + "epoch": 0.7982556390977443, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1718, + "step": 18271 + }, + { + "epoch": 0.7983157894736842, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 18272 + }, + { + "epoch": 0.7983759398496241, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 18273 + }, + { + "epoch": 0.7984360902255639, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 18274 + }, + { + "epoch": 0.7984962406015037, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 18275 + }, + { + "epoch": 0.7985563909774436, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 18276 + }, + { + "epoch": 0.7986165413533834, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 18277 + }, + { + "epoch": 0.7986766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 18278 + }, + { + "epoch": 0.7987368421052632, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 18279 + }, + { + "epoch": 0.7987969924812031, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.3518, + "step": 18280 + }, + { + "epoch": 0.7988571428571428, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1552, + "step": 18281 + }, + { + "epoch": 0.7989172932330827, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 18282 + }, + { + "epoch": 0.7989774436090226, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 18283 + }, + { + "epoch": 0.7990375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 18284 + }, + { + "epoch": 0.7990977443609023, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 18285 + }, + { + "epoch": 0.7991578947368421, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 18286 + }, + { + "epoch": 0.7992180451127819, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 18287 + }, + { + "epoch": 0.7992781954887218, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 18288 + }, + { + "epoch": 0.7993383458646617, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 18289 + }, + { + "epoch": 0.7993984962406016, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 18290 + }, + { + "epoch": 0.7994586466165413, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1731, + "step": 18291 + }, + { + "epoch": 0.7995187969924812, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 18292 + }, + { + "epoch": 0.7995789473684211, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 18293 + }, + { + "epoch": 0.7996390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 18294 + }, + { + "epoch": 0.7996992481203008, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 18295 + }, + { + "epoch": 0.7997593984962406, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 18296 + }, + { + "epoch": 0.7998195488721804, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 18297 + }, + { + "epoch": 0.7998796992481203, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 18298 + }, + { + "epoch": 0.7999398496240602, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.155, + "step": 18299 + }, + { + "epoch": 0.8, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 18300 + }, + { + "epoch": 0.8000601503759398, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 18301 + }, + { + "epoch": 0.8001203007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 18302 + }, + { + "epoch": 0.8001804511278195, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1713, + "step": 18303 + }, + { + "epoch": 0.8002406015037594, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 18304 + }, + { + "epoch": 0.8003007518796993, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 18305 + }, + { + "epoch": 0.800360902255639, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 18306 + }, + { + "epoch": 0.8004210526315789, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2738, + "step": 18307 + }, + { + "epoch": 0.8004812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 18308 + }, + { + "epoch": 0.8005413533834587, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 18309 + }, + { + "epoch": 0.8006015037593985, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 18310 + }, + { + "epoch": 0.8006616541353383, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 18311 + }, + { + "epoch": 0.8007218045112782, + "grad_norm": 1.6484375, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 18312 + }, + { + "epoch": 0.800781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 18313 + }, + { + "epoch": 0.8008421052631579, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 18314 + }, + { + "epoch": 0.8009022556390978, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 18315 + }, + { + "epoch": 0.8009624060150375, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 18316 + }, + { + "epoch": 0.8010225563909774, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.211, + "step": 18317 + }, + { + "epoch": 0.8010827067669173, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 18318 + }, + { + "epoch": 0.8011428571428572, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 18319 + }, + { + "epoch": 0.801203007518797, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 18320 + }, + { + "epoch": 0.8012631578947368, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 18321 + }, + { + "epoch": 0.8013233082706767, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 18322 + }, + { + "epoch": 0.8013834586466165, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1767, + "step": 18323 + }, + { + "epoch": 0.8014436090225564, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 18324 + }, + { + "epoch": 0.8015037593984963, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 18325 + }, + { + "epoch": 0.8015639097744361, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 18326 + }, + { + "epoch": 0.8016240601503759, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 18327 + }, + { + "epoch": 0.8016842105263158, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1891, + "step": 18328 + }, + { + "epoch": 0.8017443609022556, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 18329 + }, + { + "epoch": 0.8018045112781955, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 18330 + }, + { + "epoch": 0.8018646616541354, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 18331 + }, + { + "epoch": 0.8019248120300752, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 18332 + }, + { + "epoch": 0.801984962406015, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 18333 + }, + { + "epoch": 0.8020451127819549, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 18334 + }, + { + "epoch": 0.8021052631578948, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 18335 + }, + { + "epoch": 0.8021654135338346, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 18336 + }, + { + "epoch": 0.8022255639097744, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 18337 + }, + { + "epoch": 0.8022857142857143, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 18338 + }, + { + "epoch": 0.8023458646616541, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 18339 + }, + { + "epoch": 0.802406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 18340 + }, + { + "epoch": 0.8024661654135339, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 18341 + }, + { + "epoch": 0.8025263157894736, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 18342 + }, + { + "epoch": 0.8025864661654135, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 18343 + }, + { + "epoch": 0.8026466165413534, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 18344 + }, + { + "epoch": 0.8027067669172933, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 18345 + }, + { + "epoch": 0.8027669172932331, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 18346 + }, + { + "epoch": 0.8028270676691729, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 18347 + }, + { + "epoch": 0.8028872180451128, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 18348 + }, + { + "epoch": 0.8029473684210526, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 18349 + }, + { + "epoch": 0.8030075187969925, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 18350 + }, + { + "epoch": 0.8030676691729324, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 18351 + }, + { + "epoch": 0.8031278195488721, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 18352 + }, + { + "epoch": 0.803187969924812, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 18353 + }, + { + "epoch": 0.8032481203007519, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 18354 + }, + { + "epoch": 0.8033082706766917, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 18355 + }, + { + "epoch": 0.8033684210526316, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 18356 + }, + { + "epoch": 0.8034285714285714, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 18357 + }, + { + "epoch": 0.8034887218045113, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 18358 + }, + { + "epoch": 0.8035488721804511, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 18359 + }, + { + "epoch": 0.803609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 18360 + }, + { + "epoch": 0.8036691729323309, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 18361 + }, + { + "epoch": 0.8037293233082706, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 18362 + }, + { + "epoch": 0.8037894736842105, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 18363 + }, + { + "epoch": 0.8038496240601504, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 18364 + }, + { + "epoch": 0.8039097744360902, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 18365 + }, + { + "epoch": 0.8039699248120301, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 18366 + }, + { + "epoch": 0.8040300751879699, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 18367 + }, + { + "epoch": 0.8040902255639097, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 18368 + }, + { + "epoch": 0.8041503759398496, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 18369 + }, + { + "epoch": 0.8042105263157895, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 18370 + }, + { + "epoch": 0.8042706766917294, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 18371 + }, + { + "epoch": 0.8043308270676691, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1707, + "step": 18372 + }, + { + "epoch": 0.804390977443609, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 18373 + }, + { + "epoch": 0.8044511278195489, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 18374 + }, + { + "epoch": 0.8045112781954887, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 18375 + }, + { + "epoch": 0.8045714285714286, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 18376 + }, + { + "epoch": 0.8046315789473685, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2022, + "step": 18377 + }, + { + "epoch": 0.8046917293233082, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 18378 + }, + { + "epoch": 0.8047518796992481, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 18379 + }, + { + "epoch": 0.804812030075188, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 18380 + }, + { + "epoch": 0.8048721804511278, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 18381 + }, + { + "epoch": 0.8049323308270677, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2268, + "step": 18382 + }, + { + "epoch": 0.8049924812030075, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 18383 + }, + { + "epoch": 0.8050526315789474, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 18384 + }, + { + "epoch": 0.8051127819548872, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 18385 + }, + { + "epoch": 0.8051729323308271, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 18386 + }, + { + "epoch": 0.805233082706767, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 18387 + }, + { + "epoch": 0.8052932330827067, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1947, + "step": 18388 + }, + { + "epoch": 0.8053533834586466, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 18389 + }, + { + "epoch": 0.8054135338345865, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 18390 + }, + { + "epoch": 0.8054736842105263, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1444, + "step": 18391 + }, + { + "epoch": 0.8055338345864662, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 18392 + }, + { + "epoch": 0.805593984962406, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 18393 + }, + { + "epoch": 0.8056541353383458, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 18394 + }, + { + "epoch": 0.8057142857142857, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2129, + "step": 18395 + }, + { + "epoch": 0.8057744360902256, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 18396 + }, + { + "epoch": 0.8058345864661655, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 18397 + }, + { + "epoch": 0.8058947368421052, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 18398 + }, + { + "epoch": 0.8059548872180451, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 18399 + }, + { + "epoch": 0.806015037593985, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 18400 + }, + { + "epoch": 0.8060751879699248, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 18401 + }, + { + "epoch": 0.8061353383458647, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 18402 + }, + { + "epoch": 0.8061954887218045, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 18403 + }, + { + "epoch": 0.8062556390977443, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 18404 + }, + { + "epoch": 0.8063157894736842, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 18405 + }, + { + "epoch": 0.8063759398496241, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 18406 + }, + { + "epoch": 0.806436090225564, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 18407 + }, + { + "epoch": 0.8064962406015037, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 18408 + }, + { + "epoch": 0.8065563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1864, + "step": 18409 + }, + { + "epoch": 0.8066165413533835, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 18410 + }, + { + "epoch": 0.8066766917293233, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 18411 + }, + { + "epoch": 0.8067368421052632, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1271, + "step": 18412 + }, + { + "epoch": 0.806796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 18413 + }, + { + "epoch": 0.8068571428571428, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 18414 + }, + { + "epoch": 0.8069172932330827, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 18415 + }, + { + "epoch": 0.8069774436090226, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1429, + "step": 18416 + }, + { + "epoch": 0.8070375939849624, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2145, + "step": 18417 + }, + { + "epoch": 0.8070977443609022, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 18418 + }, + { + "epoch": 0.8071578947368421, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 18419 + }, + { + "epoch": 0.807218045112782, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 18420 + }, + { + "epoch": 0.8072781954887218, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 18421 + }, + { + "epoch": 0.8073383458646617, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 18422 + }, + { + "epoch": 0.8073984962406014, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 18423 + }, + { + "epoch": 0.8074586466165413, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1567, + "step": 18424 + }, + { + "epoch": 0.8075187969924812, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 18425 + }, + { + "epoch": 0.8075789473684211, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 18426 + }, + { + "epoch": 0.8076390977443609, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 18427 + }, + { + "epoch": 0.8076992481203008, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 18428 + }, + { + "epoch": 0.8077593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 18429 + }, + { + "epoch": 0.8078195488721804, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 18430 + }, + { + "epoch": 0.8078796992481203, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 18431 + }, + { + "epoch": 0.8079398496240602, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 18432 + }, + { + "epoch": 0.808, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2401, + "step": 18433 + }, + { + "epoch": 0.8080601503759398, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 18434 + }, + { + "epoch": 0.8081203007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 18435 + }, + { + "epoch": 0.8081804511278196, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 18436 + }, + { + "epoch": 0.8082406015037594, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.146, + "step": 18437 + }, + { + "epoch": 0.8083007518796993, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 18438 + }, + { + "epoch": 0.8083609022556391, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 18439 + }, + { + "epoch": 0.8084210526315789, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 18440 + }, + { + "epoch": 0.8084812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 18441 + }, + { + "epoch": 0.8085413533834587, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 18442 + }, + { + "epoch": 0.8086015037593985, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 18443 + }, + { + "epoch": 0.8086616541353383, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 18444 + }, + { + "epoch": 0.8087218045112782, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 18445 + }, + { + "epoch": 0.808781954887218, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2743, + "step": 18446 + }, + { + "epoch": 0.8088421052631579, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 18447 + }, + { + "epoch": 0.8089022556390978, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 18448 + }, + { + "epoch": 0.8089624060150375, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 18449 + }, + { + "epoch": 0.8090225563909774, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 18450 + }, + { + "epoch": 0.8090827067669173, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.2126, + "step": 18451 + }, + { + "epoch": 0.8091428571428572, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 18452 + }, + { + "epoch": 0.809203007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 18453 + }, + { + "epoch": 0.8092631578947368, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 18454 + }, + { + "epoch": 0.8093233082706767, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 18455 + }, + { + "epoch": 0.8093834586466165, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 18456 + }, + { + "epoch": 0.8094436090225564, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 18457 + }, + { + "epoch": 0.8095037593984963, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2084, + "step": 18458 + }, + { + "epoch": 0.809563909774436, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 18459 + }, + { + "epoch": 0.8096240601503759, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 18460 + }, + { + "epoch": 0.8096842105263158, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 18461 + }, + { + "epoch": 0.8097443609022557, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 18462 + }, + { + "epoch": 0.8098045112781955, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 18463 + }, + { + "epoch": 0.8098646616541353, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 18464 + }, + { + "epoch": 0.8099248120300752, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 18465 + }, + { + "epoch": 0.809984962406015, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 18466 + }, + { + "epoch": 0.8100451127819549, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 18467 + }, + { + "epoch": 0.8101052631578948, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 18468 + }, + { + "epoch": 0.8101654135338345, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1271, + "step": 18469 + }, + { + "epoch": 0.8102255639097744, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 18470 + }, + { + "epoch": 0.8102857142857143, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 18471 + }, + { + "epoch": 0.8103458646616541, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 18472 + }, + { + "epoch": 0.810406015037594, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 18473 + }, + { + "epoch": 0.8104661654135339, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 18474 + }, + { + "epoch": 0.8105263157894737, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 18475 + }, + { + "epoch": 0.8105864661654135, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 18476 + }, + { + "epoch": 0.8106466165413534, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1801, + "step": 18477 + }, + { + "epoch": 0.8107067669172933, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 18478 + }, + { + "epoch": 0.8107669172932331, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 18479 + }, + { + "epoch": 0.8108270676691729, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1512, + "step": 18480 + }, + { + "epoch": 0.8108872180451128, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 18481 + }, + { + "epoch": 0.8109473684210526, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 18482 + }, + { + "epoch": 0.8110075187969925, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 18483 + }, + { + "epoch": 0.8110676691729324, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 18484 + }, + { + "epoch": 0.8111278195488721, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 18485 + }, + { + "epoch": 0.811187969924812, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 18486 + }, + { + "epoch": 0.8112481203007519, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 18487 + }, + { + "epoch": 0.8113082706766918, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 18488 + }, + { + "epoch": 0.8113684210526316, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 18489 + }, + { + "epoch": 0.8114285714285714, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 18490 + }, + { + "epoch": 0.8114887218045113, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 18491 + }, + { + "epoch": 0.8115488721804511, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 18492 + }, + { + "epoch": 0.811609022556391, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 18493 + }, + { + "epoch": 0.8116691729323309, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 18494 + }, + { + "epoch": 0.8117293233082706, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 18495 + }, + { + "epoch": 0.8117894736842105, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 18496 + }, + { + "epoch": 0.8118496240601504, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 18497 + }, + { + "epoch": 0.8119097744360902, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 18498 + }, + { + "epoch": 0.8119699248120301, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1929, + "step": 18499 + }, + { + "epoch": 0.8120300751879699, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.223, + "step": 18500 + }, + { + "epoch": 0.8120902255639098, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 18501 + }, + { + "epoch": 0.8121503759398496, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.188, + "step": 18502 + }, + { + "epoch": 0.8122105263157895, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 18503 + }, + { + "epoch": 0.8122706766917294, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 18504 + }, + { + "epoch": 0.8123308270676691, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 18505 + }, + { + "epoch": 0.812390977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 18506 + }, + { + "epoch": 0.8124511278195489, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 18507 + }, + { + "epoch": 0.8125112781954887, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 18508 + }, + { + "epoch": 0.8125714285714286, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 18509 + }, + { + "epoch": 0.8126315789473684, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 18510 + }, + { + "epoch": 0.8126917293233082, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 18511 + }, + { + "epoch": 0.8127518796992481, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 18512 + }, + { + "epoch": 0.812812030075188, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 18513 + }, + { + "epoch": 0.8128721804511279, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 18514 + }, + { + "epoch": 0.8129323308270676, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 18515 + }, + { + "epoch": 0.8129924812030075, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 18516 + }, + { + "epoch": 0.8130526315789474, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 18517 + }, + { + "epoch": 0.8131127819548872, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 18518 + }, + { + "epoch": 0.8131729323308271, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 18519 + }, + { + "epoch": 0.8132330827067669, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 18520 + }, + { + "epoch": 0.8132932330827067, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 18521 + }, + { + "epoch": 0.8133533834586466, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 18522 + }, + { + "epoch": 0.8134135338345865, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 18523 + }, + { + "epoch": 0.8134736842105263, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2021, + "step": 18524 + }, + { + "epoch": 0.8135338345864662, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2102, + "step": 18525 + }, + { + "epoch": 0.813593984962406, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 18526 + }, + { + "epoch": 0.8136541353383459, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 18527 + }, + { + "epoch": 0.8137142857142857, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 18528 + }, + { + "epoch": 0.8137744360902256, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 18529 + }, + { + "epoch": 0.8138345864661655, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 18530 + }, + { + "epoch": 0.8138947368421052, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 18531 + }, + { + "epoch": 0.8139548872180451, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1512, + "step": 18532 + }, + { + "epoch": 0.814015037593985, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 18533 + }, + { + "epoch": 0.8140751879699248, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 18534 + }, + { + "epoch": 0.8141353383458647, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 18535 + }, + { + "epoch": 0.8141954887218045, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 18536 + }, + { + "epoch": 0.8142556390977443, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 18537 + }, + { + "epoch": 0.8143157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 18538 + }, + { + "epoch": 0.8143759398496241, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 18539 + }, + { + "epoch": 0.814436090225564, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 18540 + }, + { + "epoch": 0.8144962406015037, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 18541 + }, + { + "epoch": 0.8145563909774436, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 18542 + }, + { + "epoch": 0.8146165413533835, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 18543 + }, + { + "epoch": 0.8146766917293233, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 18544 + }, + { + "epoch": 0.8147368421052632, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 18545 + }, + { + "epoch": 0.814796992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 18546 + }, + { + "epoch": 0.8148571428571428, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 18547 + }, + { + "epoch": 0.8149172932330827, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 18548 + }, + { + "epoch": 0.8149774436090226, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 18549 + }, + { + "epoch": 0.8150375939849624, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 18550 + }, + { + "epoch": 0.8150977443609022, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 18551 + }, + { + "epoch": 0.8151578947368421, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 18552 + }, + { + "epoch": 0.815218045112782, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 18553 + }, + { + "epoch": 0.8152781954887218, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 18554 + }, + { + "epoch": 0.8153383458646617, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 18555 + }, + { + "epoch": 0.8153984962406015, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 18556 + }, + { + "epoch": 0.8154586466165413, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 18557 + }, + { + "epoch": 0.8155187969924812, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 18558 + }, + { + "epoch": 0.8155789473684211, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1718, + "step": 18559 + }, + { + "epoch": 0.8156390977443609, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 18560 + }, + { + "epoch": 0.8156992481203007, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 18561 + }, + { + "epoch": 0.8157593984962406, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 18562 + }, + { + "epoch": 0.8158195488721804, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 18563 + }, + { + "epoch": 0.8158796992481203, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 18564 + }, + { + "epoch": 0.8159398496240602, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 18565 + }, + { + "epoch": 0.816, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 18566 + }, + { + "epoch": 0.8160601503759398, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 18567 + }, + { + "epoch": 0.8161203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 18568 + }, + { + "epoch": 0.8161804511278196, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 18569 + }, + { + "epoch": 0.8162406015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 18570 + }, + { + "epoch": 0.8163007518796992, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 18571 + }, + { + "epoch": 0.8163609022556391, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 18572 + }, + { + "epoch": 0.8164210526315789, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2164, + "step": 18573 + }, + { + "epoch": 0.8164812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 18574 + }, + { + "epoch": 0.8165413533834587, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 18575 + }, + { + "epoch": 0.8166015037593986, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 18576 + }, + { + "epoch": 0.8166616541353383, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2177, + "step": 18577 + }, + { + "epoch": 0.8167218045112782, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1995, + "step": 18578 + }, + { + "epoch": 0.816781954887218, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 18579 + }, + { + "epoch": 0.8168421052631579, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.1124, + "step": 18580 + }, + { + "epoch": 0.8169022556390978, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 18581 + }, + { + "epoch": 0.8169624060150376, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 18582 + }, + { + "epoch": 0.8170225563909774, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 18583 + }, + { + "epoch": 0.8170827067669173, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1904, + "step": 18584 + }, + { + "epoch": 0.8171428571428572, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2565, + "step": 18585 + }, + { + "epoch": 0.817203007518797, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1518, + "step": 18586 + }, + { + "epoch": 0.8172631578947368, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 18587 + }, + { + "epoch": 0.8173233082706767, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 18588 + }, + { + "epoch": 0.8173834586466165, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 18589 + }, + { + "epoch": 0.8174436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 18590 + }, + { + "epoch": 0.8175037593984963, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 18591 + }, + { + "epoch": 0.817563909774436, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1444, + "step": 18592 + }, + { + "epoch": 0.8176240601503759, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 18593 + }, + { + "epoch": 0.8176842105263158, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 18594 + }, + { + "epoch": 0.8177443609022557, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 18595 + }, + { + "epoch": 0.8178045112781955, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.2619, + "step": 18596 + }, + { + "epoch": 0.8178646616541353, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.189, + "step": 18597 + }, + { + "epoch": 0.8179248120300752, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 18598 + }, + { + "epoch": 0.817984962406015, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 18599 + }, + { + "epoch": 0.8180451127819549, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 18600 + }, + { + "epoch": 0.8181052631578948, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 18601 + }, + { + "epoch": 0.8181654135338345, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 18602 + }, + { + "epoch": 0.8182255639097744, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 18603 + }, + { + "epoch": 0.8182857142857143, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 18604 + }, + { + "epoch": 0.8183458646616542, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1073, + "step": 18605 + }, + { + "epoch": 0.818406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 18606 + }, + { + "epoch": 0.8184661654135338, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 18607 + }, + { + "epoch": 0.8185263157894737, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 18608 + }, + { + "epoch": 0.8185864661654135, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 18609 + }, + { + "epoch": 0.8186466165413534, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 18610 + }, + { + "epoch": 0.8187067669172933, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 18611 + }, + { + "epoch": 0.818766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 18612 + }, + { + "epoch": 0.8188270676691729, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 18613 + }, + { + "epoch": 0.8188872180451128, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 18614 + }, + { + "epoch": 0.8189473684210526, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2132, + "step": 18615 + }, + { + "epoch": 0.8190075187969925, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 18616 + }, + { + "epoch": 0.8190676691729323, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 18617 + }, + { + "epoch": 0.8191278195488721, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 18618 + }, + { + "epoch": 0.819187969924812, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 18619 + }, + { + "epoch": 0.8192481203007519, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 18620 + }, + { + "epoch": 0.8193082706766918, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 18621 + }, + { + "epoch": 0.8193684210526315, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 18622 + }, + { + "epoch": 0.8194285714285714, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 18623 + }, + { + "epoch": 0.8194887218045113, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 18624 + }, + { + "epoch": 0.8195488721804511, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 18625 + }, + { + "epoch": 0.819609022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 18626 + }, + { + "epoch": 0.8196691729323309, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 18627 + }, + { + "epoch": 0.8197293233082706, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 18628 + }, + { + "epoch": 0.8197894736842105, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 18629 + }, + { + "epoch": 0.8198496240601504, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2034, + "step": 18630 + }, + { + "epoch": 0.8199097744360903, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 18631 + }, + { + "epoch": 0.8199699248120301, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 18632 + }, + { + "epoch": 0.8200300751879699, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 18633 + }, + { + "epoch": 0.8200902255639098, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 18634 + }, + { + "epoch": 0.8201503759398496, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 18635 + }, + { + "epoch": 0.8202105263157895, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 18636 + }, + { + "epoch": 0.8202706766917294, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 18637 + }, + { + "epoch": 0.8203308270676691, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 18638 + }, + { + "epoch": 0.820390977443609, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 18639 + }, + { + "epoch": 0.8204511278195489, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 18640 + }, + { + "epoch": 0.8205112781954887, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 18641 + }, + { + "epoch": 0.8205714285714286, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 18642 + }, + { + "epoch": 0.8206315789473684, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 18643 + }, + { + "epoch": 0.8206917293233082, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2047, + "step": 18644 + }, + { + "epoch": 0.8207518796992481, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1658, + "step": 18645 + }, + { + "epoch": 0.820812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 18646 + }, + { + "epoch": 0.8208721804511279, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 18647 + }, + { + "epoch": 0.8209323308270676, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2927, + "step": 18648 + }, + { + "epoch": 0.8209924812030075, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1864, + "step": 18649 + }, + { + "epoch": 0.8210526315789474, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 18650 + }, + { + "epoch": 0.8211127819548872, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 18651 + }, + { + "epoch": 0.8211729323308271, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 18652 + }, + { + "epoch": 0.8212330827067669, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 18653 + }, + { + "epoch": 0.8212932330827067, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 18654 + }, + { + "epoch": 0.8213533834586466, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.413, + "step": 18655 + }, + { + "epoch": 0.8214135338345865, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 18656 + }, + { + "epoch": 0.8214736842105264, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 18657 + }, + { + "epoch": 0.8215338345864661, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 18658 + }, + { + "epoch": 0.821593984962406, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2246, + "step": 18659 + }, + { + "epoch": 0.8216541353383459, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 18660 + }, + { + "epoch": 0.8217142857142857, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1972, + "step": 18661 + }, + { + "epoch": 0.8217744360902256, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 18662 + }, + { + "epoch": 0.8218345864661654, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 18663 + }, + { + "epoch": 0.8218947368421052, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 18664 + }, + { + "epoch": 0.8219548872180451, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.146, + "step": 18665 + }, + { + "epoch": 0.822015037593985, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 18666 + }, + { + "epoch": 0.8220751879699248, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 18667 + }, + { + "epoch": 0.8221353383458646, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 18668 + }, + { + "epoch": 0.8221954887218045, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 18669 + }, + { + "epoch": 0.8222556390977444, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 18670 + }, + { + "epoch": 0.8223157894736842, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 18671 + }, + { + "epoch": 0.8223759398496241, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2111, + "step": 18672 + }, + { + "epoch": 0.822436090225564, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 18673 + }, + { + "epoch": 0.8224962406015037, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 18674 + }, + { + "epoch": 0.8225563909774436, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 18675 + }, + { + "epoch": 0.8226165413533835, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 18676 + }, + { + "epoch": 0.8226766917293233, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 18677 + }, + { + "epoch": 0.8227368421052632, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 18678 + }, + { + "epoch": 0.822796992481203, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 18679 + }, + { + "epoch": 0.8228571428571428, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 18680 + }, + { + "epoch": 0.8229172932330827, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 18681 + }, + { + "epoch": 0.8229774436090226, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 18682 + }, + { + "epoch": 0.8230375939849625, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 18683 + }, + { + "epoch": 0.8230977443609022, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2188, + "step": 18684 + }, + { + "epoch": 0.8231578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 18685 + }, + { + "epoch": 0.823218045112782, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 18686 + }, + { + "epoch": 0.8232781954887218, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 18687 + }, + { + "epoch": 0.8233383458646617, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 18688 + }, + { + "epoch": 0.8233984962406015, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 18689 + }, + { + "epoch": 0.8234586466165413, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 18690 + }, + { + "epoch": 0.8235187969924812, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 18691 + }, + { + "epoch": 0.8235789473684211, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 18692 + }, + { + "epoch": 0.823639097744361, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 18693 + }, + { + "epoch": 0.8236992481203007, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 18694 + }, + { + "epoch": 0.8237593984962406, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 18695 + }, + { + "epoch": 0.8238195488721805, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 18696 + }, + { + "epoch": 0.8238796992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 18697 + }, + { + "epoch": 0.8239398496240602, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 18698 + }, + { + "epoch": 0.824, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.4162, + "step": 18699 + }, + { + "epoch": 0.8240601503759398, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 18700 + }, + { + "epoch": 0.8241203007518797, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.2627, + "step": 18701 + }, + { + "epoch": 0.8241804511278196, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 18702 + }, + { + "epoch": 0.8242406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 18703 + }, + { + "epoch": 0.8243007518796992, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 18704 + }, + { + "epoch": 0.8243609022556391, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 18705 + }, + { + "epoch": 0.8244210526315789, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 18706 + }, + { + "epoch": 0.8244812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 18707 + }, + { + "epoch": 0.8245413533834587, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2123, + "step": 18708 + }, + { + "epoch": 0.8246015037593984, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 18709 + }, + { + "epoch": 0.8246616541353383, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 18710 + }, + { + "epoch": 0.8247218045112782, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 18711 + }, + { + "epoch": 0.8247819548872181, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 18712 + }, + { + "epoch": 0.8248421052631579, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 18713 + }, + { + "epoch": 0.8249022556390977, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 18714 + }, + { + "epoch": 0.8249624060150376, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1597, + "step": 18715 + }, + { + "epoch": 0.8250225563909774, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 18716 + }, + { + "epoch": 0.8250827067669173, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 18717 + }, + { + "epoch": 0.8251428571428572, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 18718 + }, + { + "epoch": 0.8252030075187969, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 18719 + }, + { + "epoch": 0.8252631578947368, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 18720 + }, + { + "epoch": 0.8253233082706767, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 18721 + }, + { + "epoch": 0.8253834586466166, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 18722 + }, + { + "epoch": 0.8254436090225564, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 18723 + }, + { + "epoch": 0.8255037593984963, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 18724 + }, + { + "epoch": 0.825563909774436, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 18725 + }, + { + "epoch": 0.8256240601503759, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 18726 + }, + { + "epoch": 0.8256842105263158, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 18727 + }, + { + "epoch": 0.8257443609022557, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 18728 + }, + { + "epoch": 0.8258045112781955, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 18729 + }, + { + "epoch": 0.8258646616541353, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 18730 + }, + { + "epoch": 0.8259248120300752, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 18731 + }, + { + "epoch": 0.825984962406015, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 18732 + }, + { + "epoch": 0.8260451127819549, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2399, + "step": 18733 + }, + { + "epoch": 0.8261052631578948, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 18734 + }, + { + "epoch": 0.8261654135338345, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 18735 + }, + { + "epoch": 0.8262255639097744, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 18736 + }, + { + "epoch": 0.8262857142857143, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 18737 + }, + { + "epoch": 0.8263458646616542, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 18738 + }, + { + "epoch": 0.826406015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 18739 + }, + { + "epoch": 0.8264661654135338, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 18740 + }, + { + "epoch": 0.8265263157894737, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 18741 + }, + { + "epoch": 0.8265864661654135, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 18742 + }, + { + "epoch": 0.8266466165413534, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 18743 + }, + { + "epoch": 0.8267067669172933, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 18744 + }, + { + "epoch": 0.826766917293233, + "grad_norm": 2.390625, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 18745 + }, + { + "epoch": 0.8268270676691729, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 18746 + }, + { + "epoch": 0.8268872180451128, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 18747 + }, + { + "epoch": 0.8269473684210527, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.2189, + "step": 18748 + }, + { + "epoch": 0.8270075187969925, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 18749 + }, + { + "epoch": 0.8270676691729323, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1702, + "step": 18750 + }, + { + "epoch": 0.8271278195488722, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.2802, + "step": 18751 + }, + { + "epoch": 0.827187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 18752 + }, + { + "epoch": 0.8272481203007519, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 18753 + }, + { + "epoch": 0.8273082706766918, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 18754 + }, + { + "epoch": 0.8273684210526315, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 18755 + }, + { + "epoch": 0.8274285714285714, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 18756 + }, + { + "epoch": 0.8274887218045113, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 18757 + }, + { + "epoch": 0.8275488721804511, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 18758 + }, + { + "epoch": 0.827609022556391, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 18759 + }, + { + "epoch": 0.8276691729323308, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 18760 + }, + { + "epoch": 0.8277293233082706, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 18761 + }, + { + "epoch": 0.8277894736842105, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 18762 + }, + { + "epoch": 0.8278496240601504, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1357, + "step": 18763 + }, + { + "epoch": 0.8279097744360903, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 18764 + }, + { + "epoch": 0.82796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 18765 + }, + { + "epoch": 0.8280300751879699, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 18766 + }, + { + "epoch": 0.8280902255639098, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2808, + "step": 18767 + }, + { + "epoch": 0.8281503759398496, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 18768 + }, + { + "epoch": 0.8282105263157895, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 18769 + }, + { + "epoch": 0.8282706766917293, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 18770 + }, + { + "epoch": 0.8283308270676691, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 18771 + }, + { + "epoch": 0.828390977443609, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 18772 + }, + { + "epoch": 0.8284511278195489, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 18773 + }, + { + "epoch": 0.8285112781954888, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 18774 + }, + { + "epoch": 0.8285714285714286, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 18775 + }, + { + "epoch": 0.8286315789473684, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 18776 + }, + { + "epoch": 0.8286917293233083, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 18777 + }, + { + "epoch": 0.8287518796992481, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 18778 + }, + { + "epoch": 0.828812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 18779 + }, + { + "epoch": 0.8288721804511279, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2094, + "step": 18780 + }, + { + "epoch": 0.8289323308270676, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 18781 + }, + { + "epoch": 0.8289924812030075, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 18782 + }, + { + "epoch": 0.8290526315789474, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 18783 + }, + { + "epoch": 0.8291127819548872, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 18784 + }, + { + "epoch": 0.8291729323308271, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 18785 + }, + { + "epoch": 0.8292330827067669, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 18786 + }, + { + "epoch": 0.8292932330827067, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 18787 + }, + { + "epoch": 0.8293533834586466, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 18788 + }, + { + "epoch": 0.8294135338345865, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1956, + "step": 18789 + }, + { + "epoch": 0.8294736842105264, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 18790 + }, + { + "epoch": 0.8295338345864661, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2234, + "step": 18791 + }, + { + "epoch": 0.829593984962406, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.944, + "step": 18792 + }, + { + "epoch": 0.8296541353383459, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 18793 + }, + { + "epoch": 0.8297142857142857, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 18794 + }, + { + "epoch": 0.8297744360902256, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 18795 + }, + { + "epoch": 0.8298345864661654, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 18796 + }, + { + "epoch": 0.8298947368421052, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 18797 + }, + { + "epoch": 0.8299548872180451, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 18798 + }, + { + "epoch": 0.830015037593985, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.3181, + "step": 18799 + }, + { + "epoch": 0.8300751879699249, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 18800 + }, + { + "epoch": 0.8301353383458646, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 18801 + }, + { + "epoch": 0.8301954887218045, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 18802 + }, + { + "epoch": 0.8302556390977444, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 18803 + }, + { + "epoch": 0.8303157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 18804 + }, + { + "epoch": 0.8303759398496241, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2395, + "step": 18805 + }, + { + "epoch": 0.8304360902255639, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 18806 + }, + { + "epoch": 0.8304962406015037, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 18807 + }, + { + "epoch": 0.8305563909774436, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 18808 + }, + { + "epoch": 0.8306165413533835, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 18809 + }, + { + "epoch": 0.8306766917293233, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 18810 + }, + { + "epoch": 0.8307368421052631, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 18811 + }, + { + "epoch": 0.830796992481203, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 18812 + }, + { + "epoch": 0.8308571428571428, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2252, + "step": 18813 + }, + { + "epoch": 0.8309172932330827, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 18814 + }, + { + "epoch": 0.8309774436090226, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 18815 + }, + { + "epoch": 0.8310375939849624, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 18816 + }, + { + "epoch": 0.8310977443609022, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 18817 + }, + { + "epoch": 0.8311578947368421, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 18818 + }, + { + "epoch": 0.831218045112782, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 18819 + }, + { + "epoch": 0.8312781954887218, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 18820 + }, + { + "epoch": 0.8313383458646616, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 18821 + }, + { + "epoch": 0.8313984962406015, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 18822 + }, + { + "epoch": 0.8314586466165413, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 18823 + }, + { + "epoch": 0.8315187969924812, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 18824 + }, + { + "epoch": 0.8315789473684211, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 18825 + }, + { + "epoch": 0.831639097744361, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1494, + "step": 18826 + }, + { + "epoch": 0.8316992481203007, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 18827 + }, + { + "epoch": 0.8317593984962406, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 18828 + }, + { + "epoch": 0.8318195488721805, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 18829 + }, + { + "epoch": 0.8318796992481203, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 18830 + }, + { + "epoch": 0.8319398496240602, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 18831 + }, + { + "epoch": 0.832, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 18832 + }, + { + "epoch": 0.8320601503759398, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 18833 + }, + { + "epoch": 0.8321203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 18834 + }, + { + "epoch": 0.8321804511278196, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 18835 + }, + { + "epoch": 0.8322406015037594, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 18836 + }, + { + "epoch": 0.8323007518796992, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 18837 + }, + { + "epoch": 0.8323609022556391, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 18838 + }, + { + "epoch": 0.832421052631579, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 18839 + }, + { + "epoch": 0.8324812030075188, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 18840 + }, + { + "epoch": 0.8325413533834587, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 18841 + }, + { + "epoch": 0.8326015037593985, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.2607, + "step": 18842 + }, + { + "epoch": 0.8326616541353383, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 18843 + }, + { + "epoch": 0.8327218045112782, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1897, + "step": 18844 + }, + { + "epoch": 0.8327819548872181, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 18845 + }, + { + "epoch": 0.8328421052631579, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 18846 + }, + { + "epoch": 0.8329022556390977, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 18847 + }, + { + "epoch": 0.8329624060150376, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 18848 + }, + { + "epoch": 0.8330225563909774, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 18849 + }, + { + "epoch": 0.8330827067669173, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 18850 + }, + { + "epoch": 0.8331428571428572, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 18851 + }, + { + "epoch": 0.833203007518797, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 18852 + }, + { + "epoch": 0.8332631578947368, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 18853 + }, + { + "epoch": 0.8333233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 18854 + }, + { + "epoch": 0.8333834586466166, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 18855 + }, + { + "epoch": 0.8334436090225564, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 18856 + }, + { + "epoch": 0.8335037593984962, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 18857 + }, + { + "epoch": 0.8335639097744361, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 18858 + }, + { + "epoch": 0.8336240601503759, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 18859 + }, + { + "epoch": 0.8336842105263158, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 18860 + }, + { + "epoch": 0.8337443609022557, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1986, + "step": 18861 + }, + { + "epoch": 0.8338045112781954, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 18862 + }, + { + "epoch": 0.8338646616541353, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 18863 + }, + { + "epoch": 0.8339248120300752, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 18864 + }, + { + "epoch": 0.833984962406015, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 18865 + }, + { + "epoch": 0.8340451127819549, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 18866 + }, + { + "epoch": 0.8341052631578947, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.3011, + "step": 18867 + }, + { + "epoch": 0.8341654135338346, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 18868 + }, + { + "epoch": 0.8342255639097744, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 18869 + }, + { + "epoch": 0.8342857142857143, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 18870 + }, + { + "epoch": 0.8343458646616542, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 18871 + }, + { + "epoch": 0.834406015037594, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 18872 + }, + { + "epoch": 0.8344661654135338, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 18873 + }, + { + "epoch": 0.8345263157894737, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 18874 + }, + { + "epoch": 0.8345864661654135, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 18875 + }, + { + "epoch": 0.8346466165413534, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 18876 + }, + { + "epoch": 0.8347067669172933, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 18877 + }, + { + "epoch": 0.834766917293233, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 18878 + }, + { + "epoch": 0.8348270676691729, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 18879 + }, + { + "epoch": 0.8348872180451128, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 18880 + }, + { + "epoch": 0.8349473684210527, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.2409, + "step": 18881 + }, + { + "epoch": 0.8350075187969925, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 18882 + }, + { + "epoch": 0.8350676691729323, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.348, + "step": 18883 + }, + { + "epoch": 0.8351278195488722, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 18884 + }, + { + "epoch": 0.835187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 18885 + }, + { + "epoch": 0.8352481203007519, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 18886 + }, + { + "epoch": 0.8353082706766918, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 18887 + }, + { + "epoch": 0.8353684210526315, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 18888 + }, + { + "epoch": 0.8354285714285714, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 18889 + }, + { + "epoch": 0.8354887218045113, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 18890 + }, + { + "epoch": 0.8355488721804512, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 18891 + }, + { + "epoch": 0.835609022556391, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 18892 + }, + { + "epoch": 0.8356691729323308, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.197, + "step": 18893 + }, + { + "epoch": 0.8357293233082707, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 18894 + }, + { + "epoch": 0.8357894736842105, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 18895 + }, + { + "epoch": 0.8358496240601504, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 18896 + }, + { + "epoch": 0.8359097744360903, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 18897 + }, + { + "epoch": 0.83596992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 18898 + }, + { + "epoch": 0.8360300751879699, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 18899 + }, + { + "epoch": 0.8360902255639098, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1167, + "step": 18900 + }, + { + "epoch": 0.8361503759398496, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 18901 + }, + { + "epoch": 0.8362105263157895, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 18902 + }, + { + "epoch": 0.8362706766917293, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 18903 + }, + { + "epoch": 0.8363308270676691, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 18904 + }, + { + "epoch": 0.836390977443609, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 18905 + }, + { + "epoch": 0.8364511278195489, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 18906 + }, + { + "epoch": 0.8365112781954888, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 18907 + }, + { + "epoch": 0.8365714285714285, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 18908 + }, + { + "epoch": 0.8366315789473684, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 18909 + }, + { + "epoch": 0.8366917293233083, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1913, + "step": 18910 + }, + { + "epoch": 0.8367518796992481, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 18911 + }, + { + "epoch": 0.836812030075188, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 18912 + }, + { + "epoch": 0.8368721804511278, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 18913 + }, + { + "epoch": 0.8369323308270676, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 18914 + }, + { + "epoch": 0.8369924812030075, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 18915 + }, + { + "epoch": 0.8370526315789474, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 18916 + }, + { + "epoch": 0.8371127819548873, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 18917 + }, + { + "epoch": 0.837172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 18918 + }, + { + "epoch": 0.8372330827067669, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 18919 + }, + { + "epoch": 0.8372932330827068, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 18920 + }, + { + "epoch": 0.8373533834586466, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 18921 + }, + { + "epoch": 0.8374135338345865, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 18922 + }, + { + "epoch": 0.8374736842105264, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 18923 + }, + { + "epoch": 0.8375338345864661, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 18924 + }, + { + "epoch": 0.837593984962406, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2986, + "step": 18925 + }, + { + "epoch": 0.8376541353383459, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 18926 + }, + { + "epoch": 0.8377142857142857, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 18927 + }, + { + "epoch": 0.8377744360902256, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2254, + "step": 18928 + }, + { + "epoch": 0.8378345864661654, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2154, + "step": 18929 + }, + { + "epoch": 0.8378947368421052, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 18930 + }, + { + "epoch": 0.8379548872180451, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 18931 + }, + { + "epoch": 0.838015037593985, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 18932 + }, + { + "epoch": 0.8380751879699249, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 18933 + }, + { + "epoch": 0.8381353383458646, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 18934 + }, + { + "epoch": 0.8381954887218045, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 18935 + }, + { + "epoch": 0.8382556390977444, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 18936 + }, + { + "epoch": 0.8383157894736842, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 18937 + }, + { + "epoch": 0.8383759398496241, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 18938 + }, + { + "epoch": 0.8384360902255639, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 18939 + }, + { + "epoch": 0.8384962406015037, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 18940 + }, + { + "epoch": 0.8385563909774436, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 18941 + }, + { + "epoch": 0.8386165413533835, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 18942 + }, + { + "epoch": 0.8386766917293234, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 18943 + }, + { + "epoch": 0.8387368421052631, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 18944 + }, + { + "epoch": 0.838796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 18945 + }, + { + "epoch": 0.8388571428571429, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 18946 + }, + { + "epoch": 0.8389172932330827, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 18947 + }, + { + "epoch": 0.8389774436090226, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 18948 + }, + { + "epoch": 0.8390375939849624, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 18949 + }, + { + "epoch": 0.8390977443609022, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 18950 + }, + { + "epoch": 0.8391578947368421, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 18951 + }, + { + "epoch": 0.839218045112782, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1496, + "step": 18952 + }, + { + "epoch": 0.8392781954887218, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 18953 + }, + { + "epoch": 0.8393383458646616, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 18954 + }, + { + "epoch": 0.8393984962406015, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 18955 + }, + { + "epoch": 0.8394586466165413, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 18956 + }, + { + "epoch": 0.8395187969924812, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 18957 + }, + { + "epoch": 0.8395789473684211, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 18958 + }, + { + "epoch": 0.8396390977443609, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2061, + "step": 18959 + }, + { + "epoch": 0.8396992481203007, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 18960 + }, + { + "epoch": 0.8397593984962406, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 18961 + }, + { + "epoch": 0.8398195488721805, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 18962 + }, + { + "epoch": 0.8398796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 18963 + }, + { + "epoch": 0.8399398496240601, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 18964 + }, + { + "epoch": 0.84, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 18965 + }, + { + "epoch": 0.8400601503759398, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1073, + "step": 18966 + }, + { + "epoch": 0.8401203007518797, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 18967 + }, + { + "epoch": 0.8401804511278196, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 18968 + }, + { + "epoch": 0.8402406015037593, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 18969 + }, + { + "epoch": 0.8403007518796992, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 18970 + }, + { + "epoch": 0.8403609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 18971 + }, + { + "epoch": 0.840421052631579, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 18972 + }, + { + "epoch": 0.8404812030075188, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 18973 + }, + { + "epoch": 0.8405413533834587, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1651, + "step": 18974 + }, + { + "epoch": 0.8406015037593985, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 18975 + }, + { + "epoch": 0.8406616541353383, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 18976 + }, + { + "epoch": 0.8407218045112782, + "grad_norm": 1.6171875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 18977 + }, + { + "epoch": 0.8407819548872181, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 18978 + }, + { + "epoch": 0.840842105263158, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 18979 + }, + { + "epoch": 0.8409022556390977, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 18980 + }, + { + "epoch": 0.8409624060150376, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 18981 + }, + { + "epoch": 0.8410225563909774, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 18982 + }, + { + "epoch": 0.8410827067669173, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1879, + "step": 18983 + }, + { + "epoch": 0.8411428571428572, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 18984 + }, + { + "epoch": 0.841203007518797, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.2718, + "step": 18985 + }, + { + "epoch": 0.8412631578947368, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 18986 + }, + { + "epoch": 0.8413233082706767, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1879, + "step": 18987 + }, + { + "epoch": 0.8413834586466166, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2022, + "step": 18988 + }, + { + "epoch": 0.8414436090225564, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.2681, + "step": 18989 + }, + { + "epoch": 0.8415037593984962, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 18990 + }, + { + "epoch": 0.8415639097744361, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1768, + "step": 18991 + }, + { + "epoch": 0.8416240601503759, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 18992 + }, + { + "epoch": 0.8416842105263158, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 18993 + }, + { + "epoch": 0.8417443609022557, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 18994 + }, + { + "epoch": 0.8418045112781954, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 18995 + }, + { + "epoch": 0.8418646616541353, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2541, + "step": 18996 + }, + { + "epoch": 0.8419248120300752, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 18997 + }, + { + "epoch": 0.8419849624060151, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 18998 + }, + { + "epoch": 0.8420451127819549, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 18999 + }, + { + "epoch": 0.8421052631578947, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 19000 + }, + { + "epoch": 0.8421654135338346, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 19001 + }, + { + "epoch": 0.8422255639097744, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 19002 + }, + { + "epoch": 0.8422857142857143, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1884, + "step": 19003 + }, + { + "epoch": 0.8423458646616542, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 19004 + }, + { + "epoch": 0.8424060150375939, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 19005 + }, + { + "epoch": 0.8424661654135338, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 19006 + }, + { + "epoch": 0.8425263157894737, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 19007 + }, + { + "epoch": 0.8425864661654136, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 19008 + }, + { + "epoch": 0.8426466165413534, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 19009 + }, + { + "epoch": 0.8427067669172932, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 19010 + }, + { + "epoch": 0.842766917293233, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 19011 + }, + { + "epoch": 0.8428270676691729, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 19012 + }, + { + "epoch": 0.8428872180451128, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1626, + "step": 19013 + }, + { + "epoch": 0.8429473684210527, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 19014 + }, + { + "epoch": 0.8430075187969924, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 19015 + }, + { + "epoch": 0.8430676691729323, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 19016 + }, + { + "epoch": 0.8431278195488722, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 19017 + }, + { + "epoch": 0.843187969924812, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 19018 + }, + { + "epoch": 0.8432481203007519, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 19019 + }, + { + "epoch": 0.8433082706766918, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 19020 + }, + { + "epoch": 0.8433684210526315, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 19021 + }, + { + "epoch": 0.8434285714285714, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 19022 + }, + { + "epoch": 0.8434887218045113, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 19023 + }, + { + "epoch": 0.8435488721804512, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 19024 + }, + { + "epoch": 0.843609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 19025 + }, + { + "epoch": 0.8436691729323308, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 19026 + }, + { + "epoch": 0.8437293233082707, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2079, + "step": 19027 + }, + { + "epoch": 0.8437894736842105, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 19028 + }, + { + "epoch": 0.8438496240601504, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 19029 + }, + { + "epoch": 0.8439097744360903, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 19030 + }, + { + "epoch": 0.84396992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 19031 + }, + { + "epoch": 0.8440300751879699, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 19032 + }, + { + "epoch": 0.8440902255639098, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 19033 + }, + { + "epoch": 0.8441503759398497, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 19034 + }, + { + "epoch": 0.8442105263157895, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 19035 + }, + { + "epoch": 0.8442706766917293, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 19036 + }, + { + "epoch": 0.8443308270676692, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 19037 + }, + { + "epoch": 0.844390977443609, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1696, + "step": 19038 + }, + { + "epoch": 0.8444511278195489, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 19039 + }, + { + "epoch": 0.8445112781954888, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 19040 + }, + { + "epoch": 0.8445714285714285, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 19041 + }, + { + "epoch": 0.8446315789473684, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 19042 + }, + { + "epoch": 0.8446917293233083, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 19043 + }, + { + "epoch": 0.8447518796992481, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 19044 + }, + { + "epoch": 0.844812030075188, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 19045 + }, + { + "epoch": 0.8448721804511278, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 19046 + }, + { + "epoch": 0.8449323308270676, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 19047 + }, + { + "epoch": 0.8449924812030075, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 19048 + }, + { + "epoch": 0.8450526315789474, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 19049 + }, + { + "epoch": 0.8451127819548873, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 19050 + }, + { + "epoch": 0.845172932330827, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 19051 + }, + { + "epoch": 0.8452330827067669, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 19052 + }, + { + "epoch": 0.8452932330827068, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 19053 + }, + { + "epoch": 0.8453533834586466, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 19054 + }, + { + "epoch": 0.8454135338345865, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 19055 + }, + { + "epoch": 0.8454736842105263, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1732, + "step": 19056 + }, + { + "epoch": 0.8455338345864661, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 19057 + }, + { + "epoch": 0.845593984962406, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 19058 + }, + { + "epoch": 0.8456541353383459, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 19059 + }, + { + "epoch": 0.8457142857142858, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 19060 + }, + { + "epoch": 0.8457744360902255, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 19061 + }, + { + "epoch": 0.8458345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 19062 + }, + { + "epoch": 0.8458947368421053, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.4322, + "step": 19063 + }, + { + "epoch": 0.8459548872180451, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 19064 + }, + { + "epoch": 0.846015037593985, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.216, + "step": 19065 + }, + { + "epoch": 0.8460751879699248, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 19066 + }, + { + "epoch": 0.8461353383458646, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 19067 + }, + { + "epoch": 0.8461954887218045, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 19068 + }, + { + "epoch": 0.8462556390977444, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 19069 + }, + { + "epoch": 0.8463157894736842, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1552, + "step": 19070 + }, + { + "epoch": 0.8463759398496241, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 19071 + }, + { + "epoch": 0.8464360902255639, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 19072 + }, + { + "epoch": 0.8464962406015037, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 19073 + }, + { + "epoch": 0.8465563909774436, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 19074 + }, + { + "epoch": 0.8466165413533835, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 19075 + }, + { + "epoch": 0.8466766917293234, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 19076 + }, + { + "epoch": 0.8467368421052631, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 19077 + }, + { + "epoch": 0.846796992481203, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 19078 + }, + { + "epoch": 0.8468571428571429, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 19079 + }, + { + "epoch": 0.8469172932330827, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.1911, + "step": 19080 + }, + { + "epoch": 0.8469774436090226, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 19081 + }, + { + "epoch": 0.8470375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 19082 + }, + { + "epoch": 0.8470977443609022, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 19083 + }, + { + "epoch": 0.8471578947368421, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 19084 + }, + { + "epoch": 0.847218045112782, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 19085 + }, + { + "epoch": 0.8472781954887219, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 19086 + }, + { + "epoch": 0.8473383458646616, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 19087 + }, + { + "epoch": 0.8473984962406015, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 19088 + }, + { + "epoch": 0.8474586466165414, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 19089 + }, + { + "epoch": 0.8475187969924812, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 19090 + }, + { + "epoch": 0.8475789473684211, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 19091 + }, + { + "epoch": 0.8476390977443609, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 19092 + }, + { + "epoch": 0.8476992481203007, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 19093 + }, + { + "epoch": 0.8477593984962406, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 19094 + }, + { + "epoch": 0.8478195488721805, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 19095 + }, + { + "epoch": 0.8478796992481203, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 19096 + }, + { + "epoch": 0.8479398496240601, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 19097 + }, + { + "epoch": 0.848, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1928, + "step": 19098 + }, + { + "epoch": 0.8480601503759398, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 19099 + }, + { + "epoch": 0.8481203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 19100 + }, + { + "epoch": 0.8481804511278196, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 19101 + }, + { + "epoch": 0.8482406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 19102 + }, + { + "epoch": 0.8483007518796992, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 19103 + }, + { + "epoch": 0.8483609022556391, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 19104 + }, + { + "epoch": 0.848421052631579, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 19105 + }, + { + "epoch": 0.8484812030075188, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 19106 + }, + { + "epoch": 0.8485413533834586, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 19107 + }, + { + "epoch": 0.8486015037593985, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2354, + "step": 19108 + }, + { + "epoch": 0.8486616541353383, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 19109 + }, + { + "epoch": 0.8487218045112782, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 19110 + }, + { + "epoch": 0.8487819548872181, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 19111 + }, + { + "epoch": 0.8488421052631578, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1723, + "step": 19112 + }, + { + "epoch": 0.8489022556390977, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1585, + "step": 19113 + }, + { + "epoch": 0.8489624060150376, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 19114 + }, + { + "epoch": 0.8490225563909775, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 19115 + }, + { + "epoch": 0.8490827067669173, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 19116 + }, + { + "epoch": 0.8491428571428571, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 19117 + }, + { + "epoch": 0.849203007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 19118 + }, + { + "epoch": 0.8492631578947368, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 19119 + }, + { + "epoch": 0.8493233082706767, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 19120 + }, + { + "epoch": 0.8493834586466166, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 19121 + }, + { + "epoch": 0.8494436090225564, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 19122 + }, + { + "epoch": 0.8495037593984962, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 19123 + }, + { + "epoch": 0.8495639097744361, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 19124 + }, + { + "epoch": 0.849624060150376, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 19125 + }, + { + "epoch": 0.8496842105263158, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 19126 + }, + { + "epoch": 0.8497443609022557, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 19127 + }, + { + "epoch": 0.8498045112781955, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 19128 + }, + { + "epoch": 0.8498646616541353, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.22, + "step": 19129 + }, + { + "epoch": 0.8499248120300752, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 19130 + }, + { + "epoch": 0.8499849624060151, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 19131 + }, + { + "epoch": 0.8500451127819549, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 19132 + }, + { + "epoch": 0.8501052631578947, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 19133 + }, + { + "epoch": 0.8501654135338346, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 19134 + }, + { + "epoch": 0.8502255639097744, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 19135 + }, + { + "epoch": 0.8502857142857143, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 19136 + }, + { + "epoch": 0.8503458646616542, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.163, + "step": 19137 + }, + { + "epoch": 0.8504060150375939, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2075, + "step": 19138 + }, + { + "epoch": 0.8504661654135338, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 19139 + }, + { + "epoch": 0.8505263157894737, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 19140 + }, + { + "epoch": 0.8505864661654136, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 19141 + }, + { + "epoch": 0.8506466165413534, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1637, + "step": 19142 + }, + { + "epoch": 0.8507067669172932, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 19143 + }, + { + "epoch": 0.8507669172932331, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 19144 + }, + { + "epoch": 0.8508270676691729, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 19145 + }, + { + "epoch": 0.8508872180451128, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 19146 + }, + { + "epoch": 0.8509473684210527, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 19147 + }, + { + "epoch": 0.8510075187969924, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 19148 + }, + { + "epoch": 0.8510676691729323, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 19149 + }, + { + "epoch": 0.8511278195488722, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 19150 + }, + { + "epoch": 0.851187969924812, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 19151 + }, + { + "epoch": 0.8512481203007519, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 19152 + }, + { + "epoch": 0.8513082706766917, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 19153 + }, + { + "epoch": 0.8513684210526316, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 19154 + }, + { + "epoch": 0.8514285714285714, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 19155 + }, + { + "epoch": 0.8514887218045113, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 19156 + }, + { + "epoch": 0.8515488721804512, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 19157 + }, + { + "epoch": 0.8516090225563909, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 19158 + }, + { + "epoch": 0.8516691729323308, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 19159 + }, + { + "epoch": 0.8517293233082707, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 19160 + }, + { + "epoch": 0.8517894736842105, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 19161 + }, + { + "epoch": 0.8518496240601504, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 19162 + }, + { + "epoch": 0.8519097744360902, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 19163 + }, + { + "epoch": 0.85196992481203, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 19164 + }, + { + "epoch": 0.8520300751879699, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 19165 + }, + { + "epoch": 0.8520902255639098, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 19166 + }, + { + "epoch": 0.8521503759398497, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 19167 + }, + { + "epoch": 0.8522105263157894, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 19168 + }, + { + "epoch": 0.8522706766917293, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 19169 + }, + { + "epoch": 0.8523308270676692, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 19170 + }, + { + "epoch": 0.852390977443609, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 19171 + }, + { + "epoch": 0.8524511278195489, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 19172 + }, + { + "epoch": 0.8525112781954888, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 19173 + }, + { + "epoch": 0.8525714285714285, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 19174 + }, + { + "epoch": 0.8526315789473684, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 19175 + }, + { + "epoch": 0.8526917293233083, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 19176 + }, + { + "epoch": 0.8527518796992481, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 19177 + }, + { + "epoch": 0.852812030075188, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 19178 + }, + { + "epoch": 0.8528721804511278, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 19179 + }, + { + "epoch": 0.8529323308270677, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 19180 + }, + { + "epoch": 0.8529924812030075, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 19181 + }, + { + "epoch": 0.8530526315789474, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 19182 + }, + { + "epoch": 0.8531127819548873, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 19183 + }, + { + "epoch": 0.853172932330827, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 19184 + }, + { + "epoch": 0.8532330827067669, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 19185 + }, + { + "epoch": 0.8532932330827068, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 19186 + }, + { + "epoch": 0.8533533834586466, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 19187 + }, + { + "epoch": 0.8534135338345865, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1716, + "step": 19188 + }, + { + "epoch": 0.8534736842105263, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 19189 + }, + { + "epoch": 0.8535338345864661, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 19190 + }, + { + "epoch": 0.853593984962406, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 19191 + }, + { + "epoch": 0.8536541353383459, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 19192 + }, + { + "epoch": 0.8537142857142858, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 19193 + }, + { + "epoch": 0.8537744360902255, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 19194 + }, + { + "epoch": 0.8538345864661654, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 19195 + }, + { + "epoch": 0.8538947368421053, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 19196 + }, + { + "epoch": 0.8539548872180451, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 19197 + }, + { + "epoch": 0.854015037593985, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1732, + "step": 19198 + }, + { + "epoch": 0.8540751879699248, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 19199 + }, + { + "epoch": 0.8541353383458646, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 19200 + }, + { + "epoch": 0.8541954887218045, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 19201 + }, + { + "epoch": 0.8542556390977444, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 19202 + }, + { + "epoch": 0.8543157894736843, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 19203 + }, + { + "epoch": 0.854375939849624, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 19204 + }, + { + "epoch": 0.8544360902255639, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 19205 + }, + { + "epoch": 0.8544962406015038, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 19206 + }, + { + "epoch": 0.8545563909774436, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 19207 + }, + { + "epoch": 0.8546165413533835, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 19208 + }, + { + "epoch": 0.8546766917293233, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.2443, + "step": 19209 + }, + { + "epoch": 0.8547368421052631, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 19210 + }, + { + "epoch": 0.854796992481203, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 19211 + }, + { + "epoch": 0.8548571428571429, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 19212 + }, + { + "epoch": 0.8549172932330827, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 19213 + }, + { + "epoch": 0.8549774436090225, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 19214 + }, + { + "epoch": 0.8550375939849624, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 19215 + }, + { + "epoch": 0.8550977443609022, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 19216 + }, + { + "epoch": 0.8551578947368421, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2028, + "step": 19217 + }, + { + "epoch": 0.855218045112782, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 19218 + }, + { + "epoch": 0.8552781954887219, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 19219 + }, + { + "epoch": 0.8553383458646616, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 19220 + }, + { + "epoch": 0.8553984962406015, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 19221 + }, + { + "epoch": 0.8554586466165414, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 19222 + }, + { + "epoch": 0.8555187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 19223 + }, + { + "epoch": 0.8555789473684211, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 19224 + }, + { + "epoch": 0.8556390977443609, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 19225 + }, + { + "epoch": 0.8556992481203007, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 19226 + }, + { + "epoch": 0.8557593984962406, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 19227 + }, + { + "epoch": 0.8558195488721805, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 19228 + }, + { + "epoch": 0.8558796992481204, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 19229 + }, + { + "epoch": 0.8559398496240601, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.192, + "step": 19230 + }, + { + "epoch": 0.856, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 19231 + }, + { + "epoch": 0.8560601503759399, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 19232 + }, + { + "epoch": 0.8561203007518797, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 19233 + }, + { + "epoch": 0.8561804511278196, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.3748, + "step": 19234 + }, + { + "epoch": 0.8562406015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.3379, + "step": 19235 + }, + { + "epoch": 0.8563007518796992, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 19236 + }, + { + "epoch": 0.8563609022556391, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 19237 + }, + { + "epoch": 0.856421052631579, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 19238 + }, + { + "epoch": 0.8564812030075188, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 19239 + }, + { + "epoch": 0.8565413533834586, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 19240 + }, + { + "epoch": 0.8566015037593985, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 19241 + }, + { + "epoch": 0.8566616541353383, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 19242 + }, + { + "epoch": 0.8567218045112782, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2157, + "step": 19243 + }, + { + "epoch": 0.8567819548872181, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 19244 + }, + { + "epoch": 0.8568421052631578, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 19245 + }, + { + "epoch": 0.8569022556390977, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 19246 + }, + { + "epoch": 0.8569624060150376, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 19247 + }, + { + "epoch": 0.8570225563909775, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 19248 + }, + { + "epoch": 0.8570827067669173, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 19249 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 19250 + }, + { + "epoch": 0.857203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 19251 + }, + { + "epoch": 0.8572631578947368, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2306, + "step": 19252 + }, + { + "epoch": 0.8573233082706767, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 19253 + }, + { + "epoch": 0.8573834586466166, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 19254 + }, + { + "epoch": 0.8574436090225563, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 19255 + }, + { + "epoch": 0.8575037593984962, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 19256 + }, + { + "epoch": 0.8575639097744361, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 19257 + }, + { + "epoch": 0.857624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 19258 + }, + { + "epoch": 0.8576842105263158, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 19259 + }, + { + "epoch": 0.8577443609022556, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 19260 + }, + { + "epoch": 0.8578045112781955, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 19261 + }, + { + "epoch": 0.8578646616541353, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 19262 + }, + { + "epoch": 0.8579248120300752, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.2025, + "step": 19263 + }, + { + "epoch": 0.8579849624060151, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 19264 + }, + { + "epoch": 0.8580451127819548, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 19265 + }, + { + "epoch": 0.8581052631578947, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 19266 + }, + { + "epoch": 0.8581654135338346, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 19267 + }, + { + "epoch": 0.8582255639097744, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 19268 + }, + { + "epoch": 0.8582857142857143, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 19269 + }, + { + "epoch": 0.8583458646616542, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 19270 + }, + { + "epoch": 0.858406015037594, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 19271 + }, + { + "epoch": 0.8584661654135338, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 19272 + }, + { + "epoch": 0.8585263157894737, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 19273 + }, + { + "epoch": 0.8585864661654136, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 19274 + }, + { + "epoch": 0.8586466165413534, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 19275 + }, + { + "epoch": 0.8587067669172932, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 19276 + }, + { + "epoch": 0.8587669172932331, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 19277 + }, + { + "epoch": 0.8588270676691729, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 19278 + }, + { + "epoch": 0.8588872180451128, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 19279 + }, + { + "epoch": 0.8589473684210527, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 19280 + }, + { + "epoch": 0.8590075187969924, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 19281 + }, + { + "epoch": 0.8590676691729323, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 19282 + }, + { + "epoch": 0.8591278195488722, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 19283 + }, + { + "epoch": 0.8591879699248121, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 19284 + }, + { + "epoch": 0.8592481203007519, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 19285 + }, + { + "epoch": 0.8593082706766917, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 19286 + }, + { + "epoch": 0.8593684210526316, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 19287 + }, + { + "epoch": 0.8594285714285714, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 19288 + }, + { + "epoch": 0.8594887218045113, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 19289 + }, + { + "epoch": 0.8595488721804512, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 19290 + }, + { + "epoch": 0.8596090225563909, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 19291 + }, + { + "epoch": 0.8596691729323308, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 19292 + }, + { + "epoch": 0.8597293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 19293 + }, + { + "epoch": 0.8597894736842105, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 19294 + }, + { + "epoch": 0.8598496240601504, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 19295 + }, + { + "epoch": 0.8599097744360902, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 19296 + }, + { + "epoch": 0.85996992481203, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 19297 + }, + { + "epoch": 0.8600300751879699, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2014, + "step": 19298 + }, + { + "epoch": 0.8600902255639098, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 19299 + }, + { + "epoch": 0.8601503759398497, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 19300 + }, + { + "epoch": 0.8602105263157894, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 19301 + }, + { + "epoch": 0.8602706766917293, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 19302 + }, + { + "epoch": 0.8603308270676692, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 19303 + }, + { + "epoch": 0.860390977443609, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 19304 + }, + { + "epoch": 0.8604511278195489, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1333, + "step": 19305 + }, + { + "epoch": 0.8605112781954887, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 19306 + }, + { + "epoch": 0.8605714285714285, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 19307 + }, + { + "epoch": 0.8606315789473684, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 19308 + }, + { + "epoch": 0.8606917293233083, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 19309 + }, + { + "epoch": 0.8607518796992482, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 19310 + }, + { + "epoch": 0.8608120300751879, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 19311 + }, + { + "epoch": 0.8608721804511278, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 19312 + }, + { + "epoch": 0.8609323308270677, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 19313 + }, + { + "epoch": 0.8609924812030075, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 19314 + }, + { + "epoch": 0.8610526315789474, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 19315 + }, + { + "epoch": 0.8611127819548872, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 19316 + }, + { + "epoch": 0.861172932330827, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 19317 + }, + { + "epoch": 0.8612330827067669, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 19318 + }, + { + "epoch": 0.8612932330827068, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 19319 + }, + { + "epoch": 0.8613533834586466, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 19320 + }, + { + "epoch": 0.8614135338345865, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 19321 + }, + { + "epoch": 0.8614736842105263, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1692, + "step": 19322 + }, + { + "epoch": 0.8615338345864662, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 19323 + }, + { + "epoch": 0.861593984962406, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 19324 + }, + { + "epoch": 0.8616541353383459, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 19325 + }, + { + "epoch": 0.8617142857142858, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 19326 + }, + { + "epoch": 0.8617744360902255, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 19327 + }, + { + "epoch": 0.8618345864661654, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 19328 + }, + { + "epoch": 0.8618947368421053, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 19329 + }, + { + "epoch": 0.8619548872180451, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 19330 + }, + { + "epoch": 0.862015037593985, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 19331 + }, + { + "epoch": 0.8620751879699248, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 19332 + }, + { + "epoch": 0.8621353383458646, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 19333 + }, + { + "epoch": 0.8621954887218045, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 19334 + }, + { + "epoch": 0.8622556390977444, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2114, + "step": 19335 + }, + { + "epoch": 0.8623157894736843, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 19336 + }, + { + "epoch": 0.862375939849624, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 19337 + }, + { + "epoch": 0.8624360902255639, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 19338 + }, + { + "epoch": 0.8624962406015038, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 19339 + }, + { + "epoch": 0.8625563909774436, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 19340 + }, + { + "epoch": 0.8626165413533835, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 19341 + }, + { + "epoch": 0.8626766917293233, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 19342 + }, + { + "epoch": 0.8627368421052631, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 19343 + }, + { + "epoch": 0.862796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 19344 + }, + { + "epoch": 0.8628571428571429, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 19345 + }, + { + "epoch": 0.8629172932330827, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 19346 + }, + { + "epoch": 0.8629774436090225, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 19347 + }, + { + "epoch": 0.8630375939849624, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 19348 + }, + { + "epoch": 0.8630977443609023, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 19349 + }, + { + "epoch": 0.8631578947368421, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 19350 + }, + { + "epoch": 0.863218045112782, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 19351 + }, + { + "epoch": 0.8632781954887218, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1956, + "step": 19352 + }, + { + "epoch": 0.8633383458646616, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 19353 + }, + { + "epoch": 0.8633984962406015, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 19354 + }, + { + "epoch": 0.8634586466165414, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 19355 + }, + { + "epoch": 0.8635187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 19356 + }, + { + "epoch": 0.863578947368421, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 19357 + }, + { + "epoch": 0.8636390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 19358 + }, + { + "epoch": 0.8636992481203007, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 19359 + }, + { + "epoch": 0.8637593984962406, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2816, + "step": 19360 + }, + { + "epoch": 0.8638195488721805, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 19361 + }, + { + "epoch": 0.8638796992481202, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 19362 + }, + { + "epoch": 0.8639398496240601, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 19363 + }, + { + "epoch": 0.864, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 19364 + }, + { + "epoch": 0.8640601503759399, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.3123, + "step": 19365 + }, + { + "epoch": 0.8641203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 19366 + }, + { + "epoch": 0.8641804511278196, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 19367 + }, + { + "epoch": 0.8642406015037594, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 19368 + }, + { + "epoch": 0.8643007518796992, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 19369 + }, + { + "epoch": 0.8643609022556391, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2048, + "step": 19370 + }, + { + "epoch": 0.864421052631579, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 19371 + }, + { + "epoch": 0.8644812030075189, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 19372 + }, + { + "epoch": 0.8645413533834586, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 19373 + }, + { + "epoch": 0.8646015037593985, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 19374 + }, + { + "epoch": 0.8646616541353384, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 19375 + }, + { + "epoch": 0.8647218045112782, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 19376 + }, + { + "epoch": 0.8647819548872181, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 19377 + }, + { + "epoch": 0.8648421052631579, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 19378 + }, + { + "epoch": 0.8649022556390977, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 19379 + }, + { + "epoch": 0.8649624060150376, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1712, + "step": 19380 + }, + { + "epoch": 0.8650225563909775, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 19381 + }, + { + "epoch": 0.8650827067669173, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 19382 + }, + { + "epoch": 0.8651428571428571, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 19383 + }, + { + "epoch": 0.865203007518797, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 19384 + }, + { + "epoch": 0.8652631578947368, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 19385 + }, + { + "epoch": 0.8653233082706767, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 19386 + }, + { + "epoch": 0.8653834586466166, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 19387 + }, + { + "epoch": 0.8654436090225563, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 19388 + }, + { + "epoch": 0.8655037593984962, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 19389 + }, + { + "epoch": 0.8655639097744361, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 19390 + }, + { + "epoch": 0.865624060150376, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 19391 + }, + { + "epoch": 0.8656842105263158, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 19392 + }, + { + "epoch": 0.8657443609022556, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1766, + "step": 19393 + }, + { + "epoch": 0.8658045112781955, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 19394 + }, + { + "epoch": 0.8658646616541353, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 19395 + }, + { + "epoch": 0.8659248120300752, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 19396 + }, + { + "epoch": 0.8659849624060151, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 19397 + }, + { + "epoch": 0.8660451127819548, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 19398 + }, + { + "epoch": 0.8661052631578947, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 19399 + }, + { + "epoch": 0.8661654135338346, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 19400 + }, + { + "epoch": 0.8662255639097745, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 19401 + }, + { + "epoch": 0.8662857142857143, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 19402 + }, + { + "epoch": 0.8663458646616541, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 19403 + }, + { + "epoch": 0.866406015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 19404 + }, + { + "epoch": 0.8664661654135338, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 19405 + }, + { + "epoch": 0.8665263157894737, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 19406 + }, + { + "epoch": 0.8665864661654136, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 19407 + }, + { + "epoch": 0.8666466165413533, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 19408 + }, + { + "epoch": 0.8667067669172932, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 19409 + }, + { + "epoch": 0.8667669172932331, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 19410 + }, + { + "epoch": 0.866827067669173, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 19411 + }, + { + "epoch": 0.8668872180451128, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 19412 + }, + { + "epoch": 0.8669473684210526, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 19413 + }, + { + "epoch": 0.8670075187969924, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 19414 + }, + { + "epoch": 0.8670676691729323, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 19415 + }, + { + "epoch": 0.8671278195488722, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 19416 + }, + { + "epoch": 0.8671879699248121, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 19417 + }, + { + "epoch": 0.8672481203007519, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 19418 + }, + { + "epoch": 0.8673082706766917, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 19419 + }, + { + "epoch": 0.8673684210526316, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 19420 + }, + { + "epoch": 0.8674285714285714, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 19421 + }, + { + "epoch": 0.8674887218045113, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 19422 + }, + { + "epoch": 0.8675488721804512, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 19423 + }, + { + "epoch": 0.8676090225563909, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 19424 + }, + { + "epoch": 0.8676691729323308, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 19425 + }, + { + "epoch": 0.8677293233082707, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 19426 + }, + { + "epoch": 0.8677894736842106, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 19427 + }, + { + "epoch": 0.8678496240601504, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 19428 + }, + { + "epoch": 0.8679097744360902, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 19429 + }, + { + "epoch": 0.8679699248120301, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 19430 + }, + { + "epoch": 0.8680300751879699, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 19431 + }, + { + "epoch": 0.8680902255639098, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 19432 + }, + { + "epoch": 0.8681503759398497, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 19433 + }, + { + "epoch": 0.8682105263157894, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 19434 + }, + { + "epoch": 0.8682706766917293, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 19435 + }, + { + "epoch": 0.8683308270676692, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 19436 + }, + { + "epoch": 0.868390977443609, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 19437 + }, + { + "epoch": 0.8684511278195489, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 19438 + }, + { + "epoch": 0.8685112781954887, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 19439 + }, + { + "epoch": 0.8685714285714285, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 19440 + }, + { + "epoch": 0.8686315789473684, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 19441 + }, + { + "epoch": 0.8686917293233083, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 19442 + }, + { + "epoch": 0.8687518796992482, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 19443 + }, + { + "epoch": 0.8688120300751879, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 19444 + }, + { + "epoch": 0.8688721804511278, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 19445 + }, + { + "epoch": 0.8689323308270677, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 19446 + }, + { + "epoch": 0.8689924812030075, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 19447 + }, + { + "epoch": 0.8690526315789474, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 19448 + }, + { + "epoch": 0.8691127819548872, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 19449 + }, + { + "epoch": 0.869172932330827, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 19450 + }, + { + "epoch": 0.8692330827067669, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 19451 + }, + { + "epoch": 0.8692932330827068, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 19452 + }, + { + "epoch": 0.8693533834586467, + "grad_norm": 1.75, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 19453 + }, + { + "epoch": 0.8694135338345864, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 19454 + }, + { + "epoch": 0.8694736842105263, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 19455 + }, + { + "epoch": 0.8695338345864662, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 19456 + }, + { + "epoch": 0.869593984962406, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 19457 + }, + { + "epoch": 0.8696541353383459, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 19458 + }, + { + "epoch": 0.8697142857142857, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 19459 + }, + { + "epoch": 0.8697744360902255, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 19460 + }, + { + "epoch": 0.8698345864661654, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 19461 + }, + { + "epoch": 0.8698947368421053, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 19462 + }, + { + "epoch": 0.8699548872180451, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 19463 + }, + { + "epoch": 0.8700150375939849, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 19464 + }, + { + "epoch": 0.8700751879699248, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.355, + "step": 19465 + }, + { + "epoch": 0.8701353383458647, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 19466 + }, + { + "epoch": 0.8701954887218045, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1177, + "step": 19467 + }, + { + "epoch": 0.8702556390977444, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 19468 + }, + { + "epoch": 0.8703157894736843, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 19469 + }, + { + "epoch": 0.870375939849624, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 19470 + }, + { + "epoch": 0.8704360902255639, + "grad_norm": 1.9921875, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 19471 + }, + { + "epoch": 0.8704962406015038, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 19472 + }, + { + "epoch": 0.8705563909774436, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 19473 + }, + { + "epoch": 0.8706165413533835, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 19474 + }, + { + "epoch": 0.8706766917293233, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 19475 + }, + { + "epoch": 0.8707368421052631, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 19476 + }, + { + "epoch": 0.870796992481203, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 19477 + }, + { + "epoch": 0.8708571428571429, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 19478 + }, + { + "epoch": 0.8709172932330828, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 19479 + }, + { + "epoch": 0.8709774436090225, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 19480 + }, + { + "epoch": 0.8710375939849624, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.242, + "step": 19481 + }, + { + "epoch": 0.8710977443609023, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 19482 + }, + { + "epoch": 0.8711578947368421, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 19483 + }, + { + "epoch": 0.871218045112782, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 19484 + }, + { + "epoch": 0.8712781954887218, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 19485 + }, + { + "epoch": 0.8713383458646616, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 19486 + }, + { + "epoch": 0.8713984962406015, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 19487 + }, + { + "epoch": 0.8714586466165414, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 19488 + }, + { + "epoch": 0.8715187969924812, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 19489 + }, + { + "epoch": 0.871578947368421, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 19490 + }, + { + "epoch": 0.8716390977443609, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 19491 + }, + { + "epoch": 0.8716992481203008, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 19492 + }, + { + "epoch": 0.8717593984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 19493 + }, + { + "epoch": 0.8718195488721805, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 19494 + }, + { + "epoch": 0.8718796992481203, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.4236, + "step": 19495 + }, + { + "epoch": 0.8719398496240601, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 19496 + }, + { + "epoch": 0.872, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 19497 + }, + { + "epoch": 0.8720601503759399, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 19498 + }, + { + "epoch": 0.8721203007518797, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 19499 + }, + { + "epoch": 0.8721804511278195, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 19500 + }, + { + "epoch": 0.8722406015037594, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1917, + "step": 19501 + }, + { + "epoch": 0.8723007518796992, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 19502 + }, + { + "epoch": 0.8723609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 19503 + }, + { + "epoch": 0.872421052631579, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 19504 + }, + { + "epoch": 0.8724812030075187, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 19505 + }, + { + "epoch": 0.8725413533834586, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2566, + "step": 19506 + }, + { + "epoch": 0.8726015037593985, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 19507 + }, + { + "epoch": 0.8726616541353384, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 19508 + }, + { + "epoch": 0.8727218045112782, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 19509 + }, + { + "epoch": 0.872781954887218, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 19510 + }, + { + "epoch": 0.8728421052631579, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 19511 + }, + { + "epoch": 0.8729022556390977, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 19512 + }, + { + "epoch": 0.8729624060150376, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 19513 + }, + { + "epoch": 0.8730225563909775, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 19514 + }, + { + "epoch": 0.8730827067669172, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1177, + "step": 19515 + }, + { + "epoch": 0.8731428571428571, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 19516 + }, + { + "epoch": 0.873203007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1567, + "step": 19517 + }, + { + "epoch": 0.8732631578947369, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 19518 + }, + { + "epoch": 0.8733233082706767, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 19519 + }, + { + "epoch": 0.8733834586466166, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 19520 + }, + { + "epoch": 0.8734436090225564, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 19521 + }, + { + "epoch": 0.8735037593984962, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2215, + "step": 19522 + }, + { + "epoch": 0.8735639097744361, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 19523 + }, + { + "epoch": 0.873624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 19524 + }, + { + "epoch": 0.8736842105263158, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 19525 + }, + { + "epoch": 0.8737443609022556, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 19526 + }, + { + "epoch": 0.8738045112781955, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 19527 + }, + { + "epoch": 0.8738646616541353, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 19528 + }, + { + "epoch": 0.8739248120300752, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 19529 + }, + { + "epoch": 0.8739849624060151, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 19530 + }, + { + "epoch": 0.8740451127819548, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 19531 + }, + { + "epoch": 0.8741052631578947, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 19532 + }, + { + "epoch": 0.8741654135338346, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 19533 + }, + { + "epoch": 0.8742255639097745, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 19534 + }, + { + "epoch": 0.8742857142857143, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 19535 + }, + { + "epoch": 0.8743458646616541, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 19536 + }, + { + "epoch": 0.874406015037594, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 19537 + }, + { + "epoch": 0.8744661654135338, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 19538 + }, + { + "epoch": 0.8745263157894737, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 19539 + }, + { + "epoch": 0.8745864661654136, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 19540 + }, + { + "epoch": 0.8746466165413533, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 19541 + }, + { + "epoch": 0.8747067669172932, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 19542 + }, + { + "epoch": 0.8747669172932331, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 19543 + }, + { + "epoch": 0.874827067669173, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 19544 + }, + { + "epoch": 0.8748872180451128, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 19545 + }, + { + "epoch": 0.8749473684210526, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 19546 + }, + { + "epoch": 0.8750075187969925, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 19547 + }, + { + "epoch": 0.8750676691729323, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 19548 + }, + { + "epoch": 0.8751278195488722, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 19549 + }, + { + "epoch": 0.8751879699248121, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 19550 + }, + { + "epoch": 0.8752481203007518, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 19551 + }, + { + "epoch": 0.8753082706766917, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 19552 + }, + { + "epoch": 0.8753684210526316, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 19553 + }, + { + "epoch": 0.8754285714285714, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 19554 + }, + { + "epoch": 0.8754887218045113, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.2066, + "step": 19555 + }, + { + "epoch": 0.8755488721804511, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 19556 + }, + { + "epoch": 0.875609022556391, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 19557 + }, + { + "epoch": 0.8756691729323308, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 19558 + }, + { + "epoch": 0.8757293233082707, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 19559 + }, + { + "epoch": 0.8757894736842106, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 19560 + }, + { + "epoch": 0.8758496240601503, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 19561 + }, + { + "epoch": 0.8759097744360902, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 19562 + }, + { + "epoch": 0.8759699248120301, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 19563 + }, + { + "epoch": 0.8760300751879699, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 19564 + }, + { + "epoch": 0.8760902255639098, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1543, + "step": 19565 + }, + { + "epoch": 0.8761503759398497, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 19566 + }, + { + "epoch": 0.8762105263157894, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 19567 + }, + { + "epoch": 0.8762706766917293, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 19568 + }, + { + "epoch": 0.8763308270676692, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 19569 + }, + { + "epoch": 0.876390977443609, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 19570 + }, + { + "epoch": 0.8764511278195489, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 19571 + }, + { + "epoch": 0.8765112781954887, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 19572 + }, + { + "epoch": 0.8765714285714286, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 19573 + }, + { + "epoch": 0.8766315789473684, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 19574 + }, + { + "epoch": 0.8766917293233083, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 19575 + }, + { + "epoch": 0.8767518796992482, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 19576 + }, + { + "epoch": 0.8768120300751879, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 19577 + }, + { + "epoch": 0.8768721804511278, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2257, + "step": 19578 + }, + { + "epoch": 0.8769323308270677, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 19579 + }, + { + "epoch": 0.8769924812030075, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 19580 + }, + { + "epoch": 0.8770526315789474, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 19581 + }, + { + "epoch": 0.8771127819548872, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 19582 + }, + { + "epoch": 0.877172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 19583 + }, + { + "epoch": 0.8772330827067669, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 19584 + }, + { + "epoch": 0.8772932330827068, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 19585 + }, + { + "epoch": 0.8773533834586467, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 19586 + }, + { + "epoch": 0.8774135338345864, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 19587 + }, + { + "epoch": 0.8774736842105263, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 19588 + }, + { + "epoch": 0.8775338345864662, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 19589 + }, + { + "epoch": 0.877593984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.19, + "step": 19590 + }, + { + "epoch": 0.8776541353383459, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 19591 + }, + { + "epoch": 0.8777142857142857, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 19592 + }, + { + "epoch": 0.8777744360902255, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 19593 + }, + { + "epoch": 0.8778345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 19594 + }, + { + "epoch": 0.8778947368421053, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 19595 + }, + { + "epoch": 0.8779548872180452, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 19596 + }, + { + "epoch": 0.8780150375939849, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 19597 + }, + { + "epoch": 0.8780751879699248, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1543, + "step": 19598 + }, + { + "epoch": 0.8781353383458647, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 19599 + }, + { + "epoch": 0.8781954887218045, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 19600 + }, + { + "epoch": 0.8782556390977444, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 19601 + }, + { + "epoch": 0.8783157894736842, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 19602 + }, + { + "epoch": 0.878375939849624, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 19603 + }, + { + "epoch": 0.8784360902255639, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 19604 + }, + { + "epoch": 0.8784962406015038, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 19605 + }, + { + "epoch": 0.8785563909774436, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 19606 + }, + { + "epoch": 0.8786165413533834, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 19607 + }, + { + "epoch": 0.8786766917293233, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 19608 + }, + { + "epoch": 0.8787368421052631, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 19609 + }, + { + "epoch": 0.878796992481203, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 19610 + }, + { + "epoch": 0.8788571428571429, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 19611 + }, + { + "epoch": 0.8789172932330827, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 19612 + }, + { + "epoch": 0.8789774436090225, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 19613 + }, + { + "epoch": 0.8790375939849624, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 19614 + }, + { + "epoch": 0.8790977443609023, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 19615 + }, + { + "epoch": 0.8791578947368421, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 19616 + }, + { + "epoch": 0.879218045112782, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 19617 + }, + { + "epoch": 0.8792781954887218, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 19618 + }, + { + "epoch": 0.8793383458646616, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 19619 + }, + { + "epoch": 0.8793984962406015, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.3425, + "step": 19620 + }, + { + "epoch": 0.8794586466165414, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 19621 + }, + { + "epoch": 0.8795187969924813, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 19622 + }, + { + "epoch": 0.879578947368421, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 19623 + }, + { + "epoch": 0.8796390977443609, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 19624 + }, + { + "epoch": 0.8796992481203008, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 19625 + }, + { + "epoch": 0.8797593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 19626 + }, + { + "epoch": 0.8798195488721805, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 19627 + }, + { + "epoch": 0.8798796992481203, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 19628 + }, + { + "epoch": 0.8799398496240601, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 19629 + }, + { + "epoch": 0.88, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 19630 + }, + { + "epoch": 0.8800601503759399, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 19631 + }, + { + "epoch": 0.8801203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 19632 + }, + { + "epoch": 0.8801804511278195, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 19633 + }, + { + "epoch": 0.8802406015037594, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 19634 + }, + { + "epoch": 0.8803007518796993, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 19635 + }, + { + "epoch": 0.8803609022556391, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 19636 + }, + { + "epoch": 0.880421052631579, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 19637 + }, + { + "epoch": 0.8804812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 19638 + }, + { + "epoch": 0.8805413533834586, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 19639 + }, + { + "epoch": 0.8806015037593985, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 19640 + }, + { + "epoch": 0.8806616541353384, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 19641 + }, + { + "epoch": 0.8807218045112782, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 19642 + }, + { + "epoch": 0.880781954887218, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 19643 + }, + { + "epoch": 0.8808421052631579, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 19644 + }, + { + "epoch": 0.8809022556390977, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 19645 + }, + { + "epoch": 0.8809624060150376, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 19646 + }, + { + "epoch": 0.8810225563909775, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 19647 + }, + { + "epoch": 0.8810827067669172, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 19648 + }, + { + "epoch": 0.8811428571428571, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 19649 + }, + { + "epoch": 0.881203007518797, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 19650 + }, + { + "epoch": 0.8812631578947369, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 19651 + }, + { + "epoch": 0.8813233082706767, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 19652 + }, + { + "epoch": 0.8813834586466165, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2387, + "step": 19653 + }, + { + "epoch": 0.8814436090225564, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 19654 + }, + { + "epoch": 0.8815037593984962, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 19655 + }, + { + "epoch": 0.8815639097744361, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 19656 + }, + { + "epoch": 0.881624060150376, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 19657 + }, + { + "epoch": 0.8816842105263157, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 19658 + }, + { + "epoch": 0.8817443609022556, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 19659 + }, + { + "epoch": 0.8818045112781955, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 19660 + }, + { + "epoch": 0.8818646616541354, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 19661 + }, + { + "epoch": 0.8819248120300752, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 19662 + }, + { + "epoch": 0.881984962406015, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 19663 + }, + { + "epoch": 0.8820451127819549, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 19664 + }, + { + "epoch": 0.8821052631578947, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 19665 + }, + { + "epoch": 0.8821654135338346, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 19666 + }, + { + "epoch": 0.8822255639097745, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 19667 + }, + { + "epoch": 0.8822857142857143, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 19668 + }, + { + "epoch": 0.8823458646616541, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 19669 + }, + { + "epoch": 0.882406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 19670 + }, + { + "epoch": 0.8824661654135338, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 19671 + }, + { + "epoch": 0.8825263157894737, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1705, + "step": 19672 + }, + { + "epoch": 0.8825864661654136, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 19673 + }, + { + "epoch": 0.8826466165413533, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2075, + "step": 19674 + }, + { + "epoch": 0.8827067669172932, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 19675 + }, + { + "epoch": 0.8827669172932331, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 19676 + }, + { + "epoch": 0.882827067669173, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 19677 + }, + { + "epoch": 0.8828872180451128, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 19678 + }, + { + "epoch": 0.8829473684210526, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 19679 + }, + { + "epoch": 0.8830075187969925, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 19680 + }, + { + "epoch": 0.8830676691729323, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 19681 + }, + { + "epoch": 0.8831278195488722, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 19682 + }, + { + "epoch": 0.8831879699248121, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 19683 + }, + { + "epoch": 0.8832481203007518, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 19684 + }, + { + "epoch": 0.8833082706766917, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 19685 + }, + { + "epoch": 0.8833684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 19686 + }, + { + "epoch": 0.8834285714285715, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 19687 + }, + { + "epoch": 0.8834887218045113, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 19688 + }, + { + "epoch": 0.8835488721804511, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 19689 + }, + { + "epoch": 0.883609022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 19690 + }, + { + "epoch": 0.8836691729323308, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 19691 + }, + { + "epoch": 0.8837293233082707, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 19692 + }, + { + "epoch": 0.8837894736842106, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 19693 + }, + { + "epoch": 0.8838496240601503, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 19694 + }, + { + "epoch": 0.8839097744360902, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 19695 + }, + { + "epoch": 0.8839699248120301, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 19696 + }, + { + "epoch": 0.8840300751879699, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 19697 + }, + { + "epoch": 0.8840902255639098, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 19698 + }, + { + "epoch": 0.8841503759398496, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 19699 + }, + { + "epoch": 0.8842105263157894, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 19700 + }, + { + "epoch": 0.8842706766917293, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 19701 + }, + { + "epoch": 0.8843308270676692, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2688, + "step": 19702 + }, + { + "epoch": 0.8843909774436091, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 19703 + }, + { + "epoch": 0.8844511278195488, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 19704 + }, + { + "epoch": 0.8845112781954887, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 19705 + }, + { + "epoch": 0.8845714285714286, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 19706 + }, + { + "epoch": 0.8846315789473684, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 19707 + }, + { + "epoch": 0.8846917293233083, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1641, + "step": 19708 + }, + { + "epoch": 0.8847518796992481, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 19709 + }, + { + "epoch": 0.8848120300751879, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 19710 + }, + { + "epoch": 0.8848721804511278, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 19711 + }, + { + "epoch": 0.8849323308270677, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 19712 + }, + { + "epoch": 0.8849924812030076, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 19713 + }, + { + "epoch": 0.8850526315789473, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 19714 + }, + { + "epoch": 0.8851127819548872, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 19715 + }, + { + "epoch": 0.8851729323308271, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 19716 + }, + { + "epoch": 0.8852330827067669, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 19717 + }, + { + "epoch": 0.8852932330827068, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 19718 + }, + { + "epoch": 0.8853533834586467, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 19719 + }, + { + "epoch": 0.8854135338345864, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 19720 + }, + { + "epoch": 0.8854736842105263, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 19721 + }, + { + "epoch": 0.8855338345864662, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 19722 + }, + { + "epoch": 0.885593984962406, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 19723 + }, + { + "epoch": 0.8856541353383459, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 19724 + }, + { + "epoch": 0.8857142857142857, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 19725 + }, + { + "epoch": 0.8857744360902255, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 19726 + }, + { + "epoch": 0.8858345864661654, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 19727 + }, + { + "epoch": 0.8858947368421053, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 19728 + }, + { + "epoch": 0.8859548872180452, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 19729 + }, + { + "epoch": 0.8860150375939849, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 19730 + }, + { + "epoch": 0.8860751879699248, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 19731 + }, + { + "epoch": 0.8861353383458647, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 19732 + }, + { + "epoch": 0.8861954887218045, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 19733 + }, + { + "epoch": 0.8862556390977444, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 19734 + }, + { + "epoch": 0.8863157894736842, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 19735 + }, + { + "epoch": 0.886375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 19736 + }, + { + "epoch": 0.8864360902255639, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 19737 + }, + { + "epoch": 0.8864962406015038, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 19738 + }, + { + "epoch": 0.8865563909774437, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 19739 + }, + { + "epoch": 0.8866165413533834, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 19740 + }, + { + "epoch": 0.8866766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.241, + "step": 19741 + }, + { + "epoch": 0.8867368421052632, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 19742 + }, + { + "epoch": 0.886796992481203, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9309, + "step": 19743 + }, + { + "epoch": 0.8868571428571429, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 19744 + }, + { + "epoch": 0.8869172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 19745 + }, + { + "epoch": 0.8869774436090225, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 19746 + }, + { + "epoch": 0.8870375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 19747 + }, + { + "epoch": 0.8870977443609023, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 19748 + }, + { + "epoch": 0.8871578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 19749 + }, + { + "epoch": 0.8872180451127819, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 19750 + }, + { + "epoch": 0.8872781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 19751 + }, + { + "epoch": 0.8873383458646616, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 19752 + }, + { + "epoch": 0.8873984962406015, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 19753 + }, + { + "epoch": 0.8874586466165414, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 19754 + }, + { + "epoch": 0.8875187969924812, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 19755 + }, + { + "epoch": 0.887578947368421, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 19756 + }, + { + "epoch": 0.8876390977443609, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 19757 + }, + { + "epoch": 0.8876992481203008, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 19758 + }, + { + "epoch": 0.8877593984962406, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.268, + "step": 19759 + }, + { + "epoch": 0.8878195488721804, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 19760 + }, + { + "epoch": 0.8878796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 19761 + }, + { + "epoch": 0.8879398496240601, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 19762 + }, + { + "epoch": 0.888, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.3151, + "step": 19763 + }, + { + "epoch": 0.8880601503759399, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1167, + "step": 19764 + }, + { + "epoch": 0.8881203007518798, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 19765 + }, + { + "epoch": 0.8881804511278195, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 19766 + }, + { + "epoch": 0.8882406015037594, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 19767 + }, + { + "epoch": 0.8883007518796993, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 19768 + }, + { + "epoch": 0.8883609022556391, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 19769 + }, + { + "epoch": 0.888421052631579, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 19770 + }, + { + "epoch": 0.8884812030075188, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 19771 + }, + { + "epoch": 0.8885413533834586, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 19772 + }, + { + "epoch": 0.8886015037593985, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 19773 + }, + { + "epoch": 0.8886616541353384, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 19774 + }, + { + "epoch": 0.8887218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 19775 + }, + { + "epoch": 0.888781954887218, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 19776 + }, + { + "epoch": 0.8888421052631579, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 19777 + }, + { + "epoch": 0.8889022556390977, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 19778 + }, + { + "epoch": 0.8889624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 19779 + }, + { + "epoch": 0.8890225563909775, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 19780 + }, + { + "epoch": 0.8890827067669173, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 19781 + }, + { + "epoch": 0.8891428571428571, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1767, + "step": 19782 + }, + { + "epoch": 0.889203007518797, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 19783 + }, + { + "epoch": 0.8892631578947369, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 19784 + }, + { + "epoch": 0.8893233082706767, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 19785 + }, + { + "epoch": 0.8893834586466165, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 19786 + }, + { + "epoch": 0.8894436090225564, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 19787 + }, + { + "epoch": 0.8895037593984962, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 19788 + }, + { + "epoch": 0.8895639097744361, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 19789 + }, + { + "epoch": 0.889624060150376, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 19790 + }, + { + "epoch": 0.8896842105263157, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 19791 + }, + { + "epoch": 0.8897443609022556, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 19792 + }, + { + "epoch": 0.8898045112781955, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1865, + "step": 19793 + }, + { + "epoch": 0.8898646616541354, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 19794 + }, + { + "epoch": 0.8899248120300752, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 19795 + }, + { + "epoch": 0.889984962406015, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 19796 + }, + { + "epoch": 0.8900451127819549, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 19797 + }, + { + "epoch": 0.8901052631578947, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1334, + "step": 19798 + }, + { + "epoch": 0.8901654135338346, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 19799 + }, + { + "epoch": 0.8902255639097745, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 19800 + }, + { + "epoch": 0.8902857142857142, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 19801 + }, + { + "epoch": 0.8903458646616541, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2525, + "step": 19802 + }, + { + "epoch": 0.890406015037594, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 19803 + }, + { + "epoch": 0.8904661654135338, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 19804 + }, + { + "epoch": 0.8905263157894737, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 19805 + }, + { + "epoch": 0.8905864661654135, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 19806 + }, + { + "epoch": 0.8906466165413534, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.216, + "step": 19807 + }, + { + "epoch": 0.8907067669172932, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 19808 + }, + { + "epoch": 0.8907669172932331, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 19809 + }, + { + "epoch": 0.890827067669173, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2116, + "step": 19810 + }, + { + "epoch": 0.8908872180451127, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 19811 + }, + { + "epoch": 0.8909473684210526, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 19812 + }, + { + "epoch": 0.8910075187969925, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 19813 + }, + { + "epoch": 0.8910676691729323, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 19814 + }, + { + "epoch": 0.8911278195488722, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 19815 + }, + { + "epoch": 0.8911879699248121, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 19816 + }, + { + "epoch": 0.8912481203007518, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 19817 + }, + { + "epoch": 0.8913082706766917, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1582, + "step": 19818 + }, + { + "epoch": 0.8913684210526316, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 19819 + }, + { + "epoch": 0.8914285714285715, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 19820 + }, + { + "epoch": 0.8914887218045113, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 19821 + }, + { + "epoch": 0.8915488721804511, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 19822 + }, + { + "epoch": 0.891609022556391, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 19823 + }, + { + "epoch": 0.8916691729323308, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 19824 + }, + { + "epoch": 0.8917293233082707, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 19825 + }, + { + "epoch": 0.8917894736842106, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 19826 + }, + { + "epoch": 0.8918496240601503, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 19827 + }, + { + "epoch": 0.8919097744360902, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 19828 + }, + { + "epoch": 0.8919699248120301, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 19829 + }, + { + "epoch": 0.89203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 19830 + }, + { + "epoch": 0.8920902255639098, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 19831 + }, + { + "epoch": 0.8921503759398496, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 19832 + }, + { + "epoch": 0.8922105263157895, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 19833 + }, + { + "epoch": 0.8922706766917293, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 19834 + }, + { + "epoch": 0.8923308270676692, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 19835 + }, + { + "epoch": 0.8923909774436091, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 19836 + }, + { + "epoch": 0.8924511278195488, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 19837 + }, + { + "epoch": 0.8925112781954887, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 19838 + }, + { + "epoch": 0.8925714285714286, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 19839 + }, + { + "epoch": 0.8926315789473684, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 19840 + }, + { + "epoch": 0.8926917293233083, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 19841 + }, + { + "epoch": 0.8927518796992481, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 19842 + }, + { + "epoch": 0.892812030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 19843 + }, + { + "epoch": 0.8928721804511278, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 19844 + }, + { + "epoch": 0.8929323308270677, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 19845 + }, + { + "epoch": 0.8929924812030076, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 19846 + }, + { + "epoch": 0.8930526315789473, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 19847 + }, + { + "epoch": 0.8931127819548872, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 19848 + }, + { + "epoch": 0.8931729323308271, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 19849 + }, + { + "epoch": 0.8932330827067669, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 19850 + }, + { + "epoch": 0.8932932330827068, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 19851 + }, + { + "epoch": 0.8933533834586466, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 19852 + }, + { + "epoch": 0.8934135338345864, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 19853 + }, + { + "epoch": 0.8934736842105263, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 19854 + }, + { + "epoch": 0.8935338345864662, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 19855 + }, + { + "epoch": 0.893593984962406, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 19856 + }, + { + "epoch": 0.8936541353383458, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 19857 + }, + { + "epoch": 0.8937142857142857, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 19858 + }, + { + "epoch": 0.8937744360902256, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 19859 + }, + { + "epoch": 0.8938345864661654, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 19860 + }, + { + "epoch": 0.8938947368421053, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 19861 + }, + { + "epoch": 0.8939548872180451, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 19862 + }, + { + "epoch": 0.8940150375939849, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 19863 + }, + { + "epoch": 0.8940751879699248, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.1526, + "step": 19864 + }, + { + "epoch": 0.8941353383458647, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 19865 + }, + { + "epoch": 0.8941954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 19866 + }, + { + "epoch": 0.8942556390977444, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1714, + "step": 19867 + }, + { + "epoch": 0.8943157894736842, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 19868 + }, + { + "epoch": 0.894375939849624, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1339, + "step": 19869 + }, + { + "epoch": 0.8944360902255639, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 19870 + }, + { + "epoch": 0.8944962406015038, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 19871 + }, + { + "epoch": 0.8945563909774437, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 19872 + }, + { + "epoch": 0.8946165413533834, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 19873 + }, + { + "epoch": 0.8946766917293233, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 19874 + }, + { + "epoch": 0.8947368421052632, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 19875 + }, + { + "epoch": 0.894796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 19876 + }, + { + "epoch": 0.8948571428571429, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 19877 + }, + { + "epoch": 0.8949172932330827, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 19878 + }, + { + "epoch": 0.8949774436090225, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 19879 + }, + { + "epoch": 0.8950375939849624, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 19880 + }, + { + "epoch": 0.8950977443609023, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 19881 + }, + { + "epoch": 0.8951578947368422, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2597, + "step": 19882 + }, + { + "epoch": 0.8952180451127819, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 19883 + }, + { + "epoch": 0.8952781954887218, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 19884 + }, + { + "epoch": 0.8953383458646617, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 19885 + }, + { + "epoch": 0.8953984962406015, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 19886 + }, + { + "epoch": 0.8954586466165414, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 19887 + }, + { + "epoch": 0.8955187969924812, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 19888 + }, + { + "epoch": 0.895578947368421, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 19889 + }, + { + "epoch": 0.8956390977443609, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 19890 + }, + { + "epoch": 0.8956992481203008, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 19891 + }, + { + "epoch": 0.8957593984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 19892 + }, + { + "epoch": 0.8958195488721804, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 19893 + }, + { + "epoch": 0.8958796992481203, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 19894 + }, + { + "epoch": 0.8959398496240601, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 19895 + }, + { + "epoch": 0.896, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 19896 + }, + { + "epoch": 0.8960601503759399, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 19897 + }, + { + "epoch": 0.8961203007518797, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 19898 + }, + { + "epoch": 0.8961804511278195, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 19899 + }, + { + "epoch": 0.8962406015037594, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1467, + "step": 19900 + }, + { + "epoch": 0.8963007518796993, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 19901 + }, + { + "epoch": 0.8963609022556391, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 19902 + }, + { + "epoch": 0.8964210526315789, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 19903 + }, + { + "epoch": 0.8964812030075188, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 19904 + }, + { + "epoch": 0.8965413533834586, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 19905 + }, + { + "epoch": 0.8966015037593985, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 19906 + }, + { + "epoch": 0.8966616541353384, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 19907 + }, + { + "epoch": 0.8967218045112781, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 19908 + }, + { + "epoch": 0.896781954887218, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1326, + "step": 19909 + }, + { + "epoch": 0.8968421052631579, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 19910 + }, + { + "epoch": 0.8969022556390978, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 19911 + }, + { + "epoch": 0.8969624060150376, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 19912 + }, + { + "epoch": 0.8970225563909775, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 19913 + }, + { + "epoch": 0.8970827067669173, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 19914 + }, + { + "epoch": 0.8971428571428571, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 19915 + }, + { + "epoch": 0.897203007518797, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 19916 + }, + { + "epoch": 0.8972631578947369, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 19917 + }, + { + "epoch": 0.8973233082706767, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 19918 + }, + { + "epoch": 0.8973834586466165, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1333, + "step": 19919 + }, + { + "epoch": 0.8974436090225564, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 19920 + }, + { + "epoch": 0.8975037593984962, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 19921 + }, + { + "epoch": 0.8975639097744361, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 19922 + }, + { + "epoch": 0.897624060150376, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 19923 + }, + { + "epoch": 0.8976842105263158, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 19924 + }, + { + "epoch": 0.8977443609022556, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 19925 + }, + { + "epoch": 0.8978045112781955, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 19926 + }, + { + "epoch": 0.8978646616541354, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 19927 + }, + { + "epoch": 0.8979248120300752, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 19928 + }, + { + "epoch": 0.897984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 19929 + }, + { + "epoch": 0.8980451127819549, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 19930 + }, + { + "epoch": 0.8981052631578947, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 19931 + }, + { + "epoch": 0.8981654135338346, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 19932 + }, + { + "epoch": 0.8982255639097745, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 19933 + }, + { + "epoch": 0.8982857142857142, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 19934 + }, + { + "epoch": 0.8983458646616541, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 19935 + }, + { + "epoch": 0.898406015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 19936 + }, + { + "epoch": 0.8984661654135339, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 19937 + }, + { + "epoch": 0.8985263157894737, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1552, + "step": 19938 + }, + { + "epoch": 0.8985864661654135, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 19939 + }, + { + "epoch": 0.8986466165413534, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 19940 + }, + { + "epoch": 0.8987067669172932, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 19941 + }, + { + "epoch": 0.8987669172932331, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 19942 + }, + { + "epoch": 0.898827067669173, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 19943 + }, + { + "epoch": 0.8988872180451127, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 19944 + }, + { + "epoch": 0.8989473684210526, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2403, + "step": 19945 + }, + { + "epoch": 0.8990075187969925, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1623, + "step": 19946 + }, + { + "epoch": 0.8990676691729323, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 19947 + }, + { + "epoch": 0.8991278195488722, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 19948 + }, + { + "epoch": 0.899187969924812, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 19949 + }, + { + "epoch": 0.8992481203007519, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 19950 + }, + { + "epoch": 0.8993082706766917, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 19951 + }, + { + "epoch": 0.8993684210526316, + "grad_norm": 1.5546875, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 19952 + }, + { + "epoch": 0.8994285714285715, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 19953 + }, + { + "epoch": 0.8994887218045112, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 19954 + }, + { + "epoch": 0.8995488721804511, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 19955 + }, + { + "epoch": 0.899609022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 19956 + }, + { + "epoch": 0.8996691729323308, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 19957 + }, + { + "epoch": 0.8997293233082707, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 19958 + }, + { + "epoch": 0.8997894736842105, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 19959 + }, + { + "epoch": 0.8998496240601503, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 19960 + }, + { + "epoch": 0.8999097744360902, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 19961 + }, + { + "epoch": 0.8999699248120301, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 19962 + }, + { + "epoch": 0.90003007518797, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 19963 + }, + { + "epoch": 0.9000902255639098, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 19964 + }, + { + "epoch": 0.9001503759398496, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 19965 + }, + { + "epoch": 0.9002105263157895, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 19966 + }, + { + "epoch": 0.9002706766917293, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 19967 + }, + { + "epoch": 0.9003308270676692, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 19968 + }, + { + "epoch": 0.9003909774436091, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 19969 + }, + { + "epoch": 0.9004511278195488, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 19970 + }, + { + "epoch": 0.9005112781954887, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 19971 + }, + { + "epoch": 0.9005714285714286, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 19972 + }, + { + "epoch": 0.9006315789473684, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 19973 + }, + { + "epoch": 0.9006917293233083, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1612, + "step": 19974 + }, + { + "epoch": 0.9007518796992481, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1167, + "step": 19975 + }, + { + "epoch": 0.900812030075188, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 19976 + }, + { + "epoch": 0.9008721804511278, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 19977 + }, + { + "epoch": 0.9009323308270677, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 19978 + }, + { + "epoch": 0.9009924812030076, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 19979 + }, + { + "epoch": 0.9010526315789473, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 19980 + }, + { + "epoch": 0.9011127819548872, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 19981 + }, + { + "epoch": 0.9011729323308271, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 19982 + }, + { + "epoch": 0.9012330827067669, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 19983 + }, + { + "epoch": 0.9012932330827068, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2222, + "step": 19984 + }, + { + "epoch": 0.9013533834586466, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 19985 + }, + { + "epoch": 0.9014135338345864, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 19986 + }, + { + "epoch": 0.9014736842105263, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 19987 + }, + { + "epoch": 0.9015338345864662, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 19988 + }, + { + "epoch": 0.9015939849624061, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 19989 + }, + { + "epoch": 0.9016541353383458, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 19990 + }, + { + "epoch": 0.9017142857142857, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 19991 + }, + { + "epoch": 0.9017744360902256, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 19992 + }, + { + "epoch": 0.9018345864661654, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 19993 + }, + { + "epoch": 0.9018947368421053, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 19994 + }, + { + "epoch": 0.9019548872180451, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 19995 + }, + { + "epoch": 0.9020150375939849, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 19996 + }, + { + "epoch": 0.9020751879699248, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 19997 + }, + { + "epoch": 0.9021353383458647, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 19998 + }, + { + "epoch": 0.9021954887218046, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2169, + "step": 19999 + }, + { + "epoch": 0.9022556390977443, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 20000 + }, + { + "epoch": 0.9023157894736842, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2087, + "step": 20001 + }, + { + "epoch": 0.902375939849624, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 20002 + }, + { + "epoch": 0.9024360902255639, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 20003 + }, + { + "epoch": 0.9024962406015038, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 20004 + }, + { + "epoch": 0.9025563909774436, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 20005 + }, + { + "epoch": 0.9026165413533834, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 0.9215, + "step": 20006 + }, + { + "epoch": 0.9026766917293233, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 20007 + }, + { + "epoch": 0.9027368421052632, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 20008 + }, + { + "epoch": 0.902796992481203, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 20009 + }, + { + "epoch": 0.9028571428571428, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 20010 + }, + { + "epoch": 0.9029172932330827, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 20011 + }, + { + "epoch": 0.9029774436090225, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 20012 + }, + { + "epoch": 0.9030375939849624, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 20013 + }, + { + "epoch": 0.9030977443609023, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.2576, + "step": 20014 + }, + { + "epoch": 0.9031578947368422, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 20015 + }, + { + "epoch": 0.9032180451127819, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 20016 + }, + { + "epoch": 0.9032781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 20017 + }, + { + "epoch": 0.9033383458646617, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 20018 + }, + { + "epoch": 0.9033984962406015, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 20019 + }, + { + "epoch": 0.9034586466165414, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.932, + "step": 20020 + }, + { + "epoch": 0.9035187969924812, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 20021 + }, + { + "epoch": 0.903578947368421, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 20022 + }, + { + "epoch": 0.9036390977443609, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 20023 + }, + { + "epoch": 0.9036992481203008, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 20024 + }, + { + "epoch": 0.9037593984962407, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 20025 + }, + { + "epoch": 0.9038195488721804, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 20026 + }, + { + "epoch": 0.9038796992481203, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 20027 + }, + { + "epoch": 0.9039398496240602, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 20028 + }, + { + "epoch": 0.904, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 20029 + }, + { + "epoch": 0.9040601503759399, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 20030 + }, + { + "epoch": 0.9041203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 20031 + }, + { + "epoch": 0.9041804511278195, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 20032 + }, + { + "epoch": 0.9042406015037594, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 20033 + }, + { + "epoch": 0.9043007518796993, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 20034 + }, + { + "epoch": 0.9043609022556391, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 20035 + }, + { + "epoch": 0.9044210526315789, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 20036 + }, + { + "epoch": 0.9044812030075188, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 20037 + }, + { + "epoch": 0.9045413533834586, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 20038 + }, + { + "epoch": 0.9046015037593985, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2437, + "step": 20039 + }, + { + "epoch": 0.9046616541353384, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2086, + "step": 20040 + }, + { + "epoch": 0.9047218045112781, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 20041 + }, + { + "epoch": 0.904781954887218, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 20042 + }, + { + "epoch": 0.9048421052631579, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 20043 + }, + { + "epoch": 0.9049022556390978, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 20044 + }, + { + "epoch": 0.9049624060150376, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 20045 + }, + { + "epoch": 0.9050225563909774, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 20046 + }, + { + "epoch": 0.9050827067669173, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 20047 + }, + { + "epoch": 0.9051428571428571, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 20048 + }, + { + "epoch": 0.905203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2159, + "step": 20049 + }, + { + "epoch": 0.9052631578947369, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 20050 + }, + { + "epoch": 0.9053233082706766, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3256, + "step": 20051 + }, + { + "epoch": 0.9053834586466165, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 20052 + }, + { + "epoch": 0.9054436090225564, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 20053 + }, + { + "epoch": 0.9055037593984963, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 20054 + }, + { + "epoch": 0.9055639097744361, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 20055 + }, + { + "epoch": 0.9056240601503759, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 20056 + }, + { + "epoch": 0.9056842105263158, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 20057 + }, + { + "epoch": 0.9057443609022556, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 20058 + }, + { + "epoch": 0.9058045112781955, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 20059 + }, + { + "epoch": 0.9058646616541354, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 20060 + }, + { + "epoch": 0.9059248120300751, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 20061 + }, + { + "epoch": 0.905984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 20062 + }, + { + "epoch": 0.9060451127819549, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 20063 + }, + { + "epoch": 0.9061052631578947, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 20064 + }, + { + "epoch": 0.9061654135338346, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 20065 + }, + { + "epoch": 0.9062255639097745, + "grad_norm": 1.9140625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 20066 + }, + { + "epoch": 0.9062857142857143, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 20067 + }, + { + "epoch": 0.9063458646616541, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 20068 + }, + { + "epoch": 0.906406015037594, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 20069 + }, + { + "epoch": 0.9064661654135339, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 20070 + }, + { + "epoch": 0.9065263157894737, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1254, + "step": 20071 + }, + { + "epoch": 0.9065864661654135, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2632, + "step": 20072 + }, + { + "epoch": 0.9066466165413534, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 20073 + }, + { + "epoch": 0.9067067669172932, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 20074 + }, + { + "epoch": 0.9067669172932331, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 20075 + }, + { + "epoch": 0.906827067669173, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 20076 + }, + { + "epoch": 0.9068872180451127, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 20077 + }, + { + "epoch": 0.9069473684210526, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 20078 + }, + { + "epoch": 0.9070075187969925, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 20079 + }, + { + "epoch": 0.9070676691729324, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 20080 + }, + { + "epoch": 0.9071278195488722, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 20081 + }, + { + "epoch": 0.907187969924812, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 20082 + }, + { + "epoch": 0.9072481203007519, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 20083 + }, + { + "epoch": 0.9073082706766917, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 20084 + }, + { + "epoch": 0.9073684210526316, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 20085 + }, + { + "epoch": 0.9074285714285715, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 20086 + }, + { + "epoch": 0.9074887218045112, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 20087 + }, + { + "epoch": 0.9075488721804511, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 20088 + }, + { + "epoch": 0.907609022556391, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 20089 + }, + { + "epoch": 0.9076691729323308, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 20090 + }, + { + "epoch": 0.9077293233082707, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 20091 + }, + { + "epoch": 0.9077894736842105, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 20092 + }, + { + "epoch": 0.9078496240601504, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 20093 + }, + { + "epoch": 0.9079097744360902, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 20094 + }, + { + "epoch": 0.9079699248120301, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 20095 + }, + { + "epoch": 0.90803007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 20096 + }, + { + "epoch": 0.9080902255639097, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 20097 + }, + { + "epoch": 0.9081503759398496, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 20098 + }, + { + "epoch": 0.9082105263157895, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 20099 + }, + { + "epoch": 0.9082706766917293, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 20100 + }, + { + "epoch": 0.9083308270676692, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 20101 + }, + { + "epoch": 0.908390977443609, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 20102 + }, + { + "epoch": 0.9084511278195488, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2521, + "step": 20103 + }, + { + "epoch": 0.9085112781954887, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 20104 + }, + { + "epoch": 0.9085714285714286, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 20105 + }, + { + "epoch": 0.9086315789473685, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 20106 + }, + { + "epoch": 0.9086917293233082, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 20107 + }, + { + "epoch": 0.9087518796992481, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 20108 + }, + { + "epoch": 0.908812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 20109 + }, + { + "epoch": 0.9088721804511278, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 20110 + }, + { + "epoch": 0.9089323308270677, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 20111 + }, + { + "epoch": 0.9089924812030076, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 20112 + }, + { + "epoch": 0.9090526315789473, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 20113 + }, + { + "epoch": 0.9091127819548872, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 20114 + }, + { + "epoch": 0.9091729323308271, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 20115 + }, + { + "epoch": 0.909233082706767, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 20116 + }, + { + "epoch": 0.9092932330827068, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 20117 + }, + { + "epoch": 0.9093533834586466, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1832, + "step": 20118 + }, + { + "epoch": 0.9094135338345865, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 20119 + }, + { + "epoch": 0.9094736842105263, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 20120 + }, + { + "epoch": 0.9095338345864662, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 20121 + }, + { + "epoch": 0.9095939849624061, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.208, + "step": 20122 + }, + { + "epoch": 0.9096541353383458, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 20123 + }, + { + "epoch": 0.9097142857142857, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 20124 + }, + { + "epoch": 0.9097744360902256, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 20125 + }, + { + "epoch": 0.9098345864661654, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 20126 + }, + { + "epoch": 0.9098947368421053, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 20127 + }, + { + "epoch": 0.9099548872180451, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 20128 + }, + { + "epoch": 0.9100150375939849, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 20129 + }, + { + "epoch": 0.9100751879699248, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 20130 + }, + { + "epoch": 0.9101353383458647, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 20131 + }, + { + "epoch": 0.9101954887218046, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 20132 + }, + { + "epoch": 0.9102556390977443, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 20133 + }, + { + "epoch": 0.9103157894736842, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 20134 + }, + { + "epoch": 0.9103759398496241, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 20135 + }, + { + "epoch": 0.9104360902255639, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 20136 + }, + { + "epoch": 0.9104962406015038, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 20137 + }, + { + "epoch": 0.9105563909774436, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 20138 + }, + { + "epoch": 0.9106165413533834, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 20139 + }, + { + "epoch": 0.9106766917293233, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 20140 + }, + { + "epoch": 0.9107368421052632, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1932, + "step": 20141 + }, + { + "epoch": 0.910796992481203, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 20142 + }, + { + "epoch": 0.9108571428571428, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 20143 + }, + { + "epoch": 0.9109172932330827, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 20144 + }, + { + "epoch": 0.9109774436090226, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 20145 + }, + { + "epoch": 0.9110375939849624, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 20146 + }, + { + "epoch": 0.9110977443609023, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1847, + "step": 20147 + }, + { + "epoch": 0.9111578947368421, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 20148 + }, + { + "epoch": 0.9112180451127819, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 20149 + }, + { + "epoch": 0.9112781954887218, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 20150 + }, + { + "epoch": 0.9113383458646617, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 20151 + }, + { + "epoch": 0.9113984962406015, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 20152 + }, + { + "epoch": 0.9114586466165413, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1856, + "step": 20153 + }, + { + "epoch": 0.9115187969924812, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 20154 + }, + { + "epoch": 0.911578947368421, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 20155 + }, + { + "epoch": 0.9116390977443609, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 20156 + }, + { + "epoch": 0.9116992481203008, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 20157 + }, + { + "epoch": 0.9117593984962405, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 20158 + }, + { + "epoch": 0.9118195488721804, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 20159 + }, + { + "epoch": 0.9118796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 20160 + }, + { + "epoch": 0.9119398496240602, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 20161 + }, + { + "epoch": 0.912, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 20162 + }, + { + "epoch": 0.9120601503759399, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 20163 + }, + { + "epoch": 0.9121203007518797, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.214, + "step": 20164 + }, + { + "epoch": 0.9121804511278195, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 20165 + }, + { + "epoch": 0.9122406015037594, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 20166 + }, + { + "epoch": 0.9123007518796993, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 20167 + }, + { + "epoch": 0.9123609022556392, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 20168 + }, + { + "epoch": 0.9124210526315789, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 20169 + }, + { + "epoch": 0.9124812030075188, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 20170 + }, + { + "epoch": 0.9125413533834587, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 20171 + }, + { + "epoch": 0.9126015037593985, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 20172 + }, + { + "epoch": 0.9126616541353384, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 20173 + }, + { + "epoch": 0.9127218045112782, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 20174 + }, + { + "epoch": 0.912781954887218, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 20175 + }, + { + "epoch": 0.9128421052631579, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 20176 + }, + { + "epoch": 0.9129022556390978, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 20177 + }, + { + "epoch": 0.9129624060150376, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 20178 + }, + { + "epoch": 0.9130225563909774, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 20179 + }, + { + "epoch": 0.9130827067669173, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 20180 + }, + { + "epoch": 0.9131428571428571, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 20181 + }, + { + "epoch": 0.913203007518797, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 20182 + }, + { + "epoch": 0.9132631578947369, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 20183 + }, + { + "epoch": 0.9133233082706766, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 1.2547, + "step": 20184 + }, + { + "epoch": 0.9133834586466165, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 20185 + }, + { + "epoch": 0.9134436090225564, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 20186 + }, + { + "epoch": 0.9135037593984963, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 20187 + }, + { + "epoch": 0.9135639097744361, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 20188 + }, + { + "epoch": 0.9136240601503759, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 20189 + }, + { + "epoch": 0.9136842105263158, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 20190 + }, + { + "epoch": 0.9137443609022556, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 20191 + }, + { + "epoch": 0.9138045112781955, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.1834, + "step": 20192 + }, + { + "epoch": 0.9138646616541354, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 20193 + }, + { + "epoch": 0.9139248120300751, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 20194 + }, + { + "epoch": 0.913984962406015, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 20195 + }, + { + "epoch": 0.9140451127819549, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 20196 + }, + { + "epoch": 0.9141052631578948, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 20197 + }, + { + "epoch": 0.9141654135338346, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 20198 + }, + { + "epoch": 0.9142255639097744, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 20199 + }, + { + "epoch": 0.9142857142857143, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 20200 + }, + { + "epoch": 0.9143458646616541, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1991, + "step": 20201 + }, + { + "epoch": 0.914406015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 20202 + }, + { + "epoch": 0.9144661654135339, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1778, + "step": 20203 + }, + { + "epoch": 0.9145263157894736, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1623, + "step": 20204 + }, + { + "epoch": 0.9145864661654135, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 20205 + }, + { + "epoch": 0.9146466165413534, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 20206 + }, + { + "epoch": 0.9147067669172932, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 20207 + }, + { + "epoch": 0.9147669172932331, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 20208 + }, + { + "epoch": 0.9148270676691729, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 20209 + }, + { + "epoch": 0.9148872180451127, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 20210 + }, + { + "epoch": 0.9149473684210526, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 20211 + }, + { + "epoch": 0.9150075187969925, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1864, + "step": 20212 + }, + { + "epoch": 0.9150676691729324, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 20213 + }, + { + "epoch": 0.9151278195488722, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 20214 + }, + { + "epoch": 0.915187969924812, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 20215 + }, + { + "epoch": 0.9152481203007519, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 20216 + }, + { + "epoch": 0.9153082706766917, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 20217 + }, + { + "epoch": 0.9153684210526316, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 20218 + }, + { + "epoch": 0.9154285714285715, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 20219 + }, + { + "epoch": 0.9154887218045112, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 20220 + }, + { + "epoch": 0.9155488721804511, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 20221 + }, + { + "epoch": 0.915609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1526, + "step": 20222 + }, + { + "epoch": 0.9156691729323309, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 20223 + }, + { + "epoch": 0.9157293233082707, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 20224 + }, + { + "epoch": 0.9157894736842105, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 20225 + }, + { + "epoch": 0.9158496240601504, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 20226 + }, + { + "epoch": 0.9159097744360902, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 20227 + }, + { + "epoch": 0.9159699248120301, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 20228 + }, + { + "epoch": 0.91603007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 20229 + }, + { + "epoch": 0.9160902255639097, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 20230 + }, + { + "epoch": 0.9161503759398496, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 20231 + }, + { + "epoch": 0.9162105263157895, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 20232 + }, + { + "epoch": 0.9162706766917293, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 20233 + }, + { + "epoch": 0.9163308270676692, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 20234 + }, + { + "epoch": 0.916390977443609, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 20235 + }, + { + "epoch": 0.9164511278195488, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 20236 + }, + { + "epoch": 0.9165112781954887, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 20237 + }, + { + "epoch": 0.9165714285714286, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 20238 + }, + { + "epoch": 0.9166315789473685, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 20239 + }, + { + "epoch": 0.9166917293233082, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1921, + "step": 20240 + }, + { + "epoch": 0.9167518796992481, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 20241 + }, + { + "epoch": 0.916812030075188, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 20242 + }, + { + "epoch": 0.9168721804511278, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 20243 + }, + { + "epoch": 0.9169323308270677, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 20244 + }, + { + "epoch": 0.9169924812030075, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 20245 + }, + { + "epoch": 0.9170526315789473, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 20246 + }, + { + "epoch": 0.9171127819548872, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 20247 + }, + { + "epoch": 0.9171729323308271, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 20248 + }, + { + "epoch": 0.917233082706767, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 20249 + }, + { + "epoch": 0.9172932330827067, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 20250 + }, + { + "epoch": 0.9173533834586466, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 20251 + }, + { + "epoch": 0.9174135338345865, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 20252 + }, + { + "epoch": 0.9174736842105263, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 20253 + }, + { + "epoch": 0.9175338345864662, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 20254 + }, + { + "epoch": 0.917593984962406, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 20255 + }, + { + "epoch": 0.9176541353383458, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 20256 + }, + { + "epoch": 0.9177142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 20257 + }, + { + "epoch": 0.9177744360902256, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 20258 + }, + { + "epoch": 0.9178345864661654, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 20259 + }, + { + "epoch": 0.9178947368421052, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 20260 + }, + { + "epoch": 0.9179548872180451, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 20261 + }, + { + "epoch": 0.918015037593985, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 20262 + }, + { + "epoch": 0.9180751879699248, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3047, + "step": 20263 + }, + { + "epoch": 0.9181353383458647, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 20264 + }, + { + "epoch": 0.9181954887218046, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 20265 + }, + { + "epoch": 0.9182556390977443, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 20266 + }, + { + "epoch": 0.9183157894736842, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 20267 + }, + { + "epoch": 0.9183759398496241, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 20268 + }, + { + "epoch": 0.9184360902255639, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 20269 + }, + { + "epoch": 0.9184962406015038, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 20270 + }, + { + "epoch": 0.9185563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 20271 + }, + { + "epoch": 0.9186165413533834, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 20272 + }, + { + "epoch": 0.9186766917293233, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 20273 + }, + { + "epoch": 0.9187368421052632, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 20274 + }, + { + "epoch": 0.9187969924812031, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 20275 + }, + { + "epoch": 0.9188571428571428, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 20276 + }, + { + "epoch": 0.9189172932330827, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2483, + "step": 20277 + }, + { + "epoch": 0.9189774436090226, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 20278 + }, + { + "epoch": 0.9190375939849624, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 20279 + }, + { + "epoch": 0.9190977443609023, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 20280 + }, + { + "epoch": 0.9191578947368421, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 20281 + }, + { + "epoch": 0.9192180451127819, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 20282 + }, + { + "epoch": 0.9192781954887218, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 20283 + }, + { + "epoch": 0.9193383458646617, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 20284 + }, + { + "epoch": 0.9193984962406015, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 20285 + }, + { + "epoch": 0.9194586466165413, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 20286 + }, + { + "epoch": 0.9195187969924812, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.3099, + "step": 20287 + }, + { + "epoch": 0.919578947368421, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 20288 + }, + { + "epoch": 0.9196390977443609, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.249, + "step": 20289 + }, + { + "epoch": 0.9196992481203008, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 20290 + }, + { + "epoch": 0.9197593984962406, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 20291 + }, + { + "epoch": 0.9198195488721804, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 20292 + }, + { + "epoch": 0.9198796992481203, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 20293 + }, + { + "epoch": 0.9199398496240602, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1421, + "step": 20294 + }, + { + "epoch": 0.92, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 20295 + }, + { + "epoch": 0.9200601503759398, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 20296 + }, + { + "epoch": 0.9201203007518797, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 20297 + }, + { + "epoch": 0.9201804511278195, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 20298 + }, + { + "epoch": 0.9202406015037594, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 20299 + }, + { + "epoch": 0.9203007518796993, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2122, + "step": 20300 + }, + { + "epoch": 0.920360902255639, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 20301 + }, + { + "epoch": 0.9204210526315789, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 20302 + }, + { + "epoch": 0.9204812030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 20303 + }, + { + "epoch": 0.9205413533834587, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 20304 + }, + { + "epoch": 0.9206015037593985, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 20305 + }, + { + "epoch": 0.9206616541353383, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 20306 + }, + { + "epoch": 0.9207218045112782, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 20307 + }, + { + "epoch": 0.920781954887218, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 20308 + }, + { + "epoch": 0.9208421052631579, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 20309 + }, + { + "epoch": 0.9209022556390978, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 20310 + }, + { + "epoch": 0.9209624060150376, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2125, + "step": 20311 + }, + { + "epoch": 0.9210225563909774, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 20312 + }, + { + "epoch": 0.9210827067669173, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 20313 + }, + { + "epoch": 0.9211428571428572, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 20314 + }, + { + "epoch": 0.921203007518797, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 20315 + }, + { + "epoch": 0.9212631578947369, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 20316 + }, + { + "epoch": 0.9213233082706767, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 20317 + }, + { + "epoch": 0.9213834586466165, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 20318 + }, + { + "epoch": 0.9214436090225564, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 20319 + }, + { + "epoch": 0.9215037593984963, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 20320 + }, + { + "epoch": 0.9215639097744361, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 20321 + }, + { + "epoch": 0.9216240601503759, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 20322 + }, + { + "epoch": 0.9216842105263158, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 20323 + }, + { + "epoch": 0.9217443609022556, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 20324 + }, + { + "epoch": 0.9218045112781955, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 20325 + }, + { + "epoch": 0.9218646616541354, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 20326 + }, + { + "epoch": 0.9219248120300751, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 20327 + }, + { + "epoch": 0.921984962406015, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 20328 + }, + { + "epoch": 0.9220451127819549, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 20329 + }, + { + "epoch": 0.9221052631578948, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 20330 + }, + { + "epoch": 0.9221654135338346, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 20331 + }, + { + "epoch": 0.9222255639097744, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 20332 + }, + { + "epoch": 0.9222857142857143, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 20333 + }, + { + "epoch": 0.9223458646616541, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2543, + "step": 20334 + }, + { + "epoch": 0.922406015037594, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 20335 + }, + { + "epoch": 0.9224661654135339, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 20336 + }, + { + "epoch": 0.9225263157894736, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1718, + "step": 20337 + }, + { + "epoch": 0.9225864661654135, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 20338 + }, + { + "epoch": 0.9226466165413534, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 20339 + }, + { + "epoch": 0.9227067669172933, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 20340 + }, + { + "epoch": 0.9227669172932331, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 20341 + }, + { + "epoch": 0.9228270676691729, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 20342 + }, + { + "epoch": 0.9228872180451128, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 20343 + }, + { + "epoch": 0.9229473684210526, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 20344 + }, + { + "epoch": 0.9230075187969925, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 20345 + }, + { + "epoch": 0.9230676691729324, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 20346 + }, + { + "epoch": 0.9231278195488721, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 20347 + }, + { + "epoch": 0.923187969924812, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 20348 + }, + { + "epoch": 0.9232481203007519, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 20349 + }, + { + "epoch": 0.9233082706766917, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 20350 + }, + { + "epoch": 0.9233684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 20351 + }, + { + "epoch": 0.9234285714285714, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 20352 + }, + { + "epoch": 0.9234887218045112, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 20353 + }, + { + "epoch": 0.9235488721804511, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1738, + "step": 20354 + }, + { + "epoch": 0.923609022556391, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 20355 + }, + { + "epoch": 0.9236691729323309, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 20356 + }, + { + "epoch": 0.9237293233082706, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1696, + "step": 20357 + }, + { + "epoch": 0.9237894736842105, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 20358 + }, + { + "epoch": 0.9238496240601504, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2885, + "step": 20359 + }, + { + "epoch": 0.9239097744360902, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 20360 + }, + { + "epoch": 0.9239699248120301, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 20361 + }, + { + "epoch": 0.92403007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 20362 + }, + { + "epoch": 0.9240902255639097, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 20363 + }, + { + "epoch": 0.9241503759398496, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 20364 + }, + { + "epoch": 0.9242105263157895, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 20365 + }, + { + "epoch": 0.9242706766917294, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 20366 + }, + { + "epoch": 0.9243308270676692, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 20367 + }, + { + "epoch": 0.924390977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 20368 + }, + { + "epoch": 0.9244511278195489, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 20369 + }, + { + "epoch": 0.9245112781954887, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 20370 + }, + { + "epoch": 0.9245714285714286, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 20371 + }, + { + "epoch": 0.9246315789473685, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 20372 + }, + { + "epoch": 0.9246917293233082, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 20373 + }, + { + "epoch": 0.9247518796992481, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 20374 + }, + { + "epoch": 0.924812030075188, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 20375 + }, + { + "epoch": 0.9248721804511278, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 20376 + }, + { + "epoch": 0.9249323308270677, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 20377 + }, + { + "epoch": 0.9249924812030075, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 20378 + }, + { + "epoch": 0.9250526315789473, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 20379 + }, + { + "epoch": 0.9251127819548872, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 20380 + }, + { + "epoch": 0.9251729323308271, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 20381 + }, + { + "epoch": 0.925233082706767, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 20382 + }, + { + "epoch": 0.9252932330827067, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 20383 + }, + { + "epoch": 0.9253533834586466, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 20384 + }, + { + "epoch": 0.9254135338345865, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 20385 + }, + { + "epoch": 0.9254736842105263, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 20386 + }, + { + "epoch": 0.9255338345864662, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 20387 + }, + { + "epoch": 0.925593984962406, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 20388 + }, + { + "epoch": 0.9256541353383458, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 20389 + }, + { + "epoch": 0.9257142857142857, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2309, + "step": 20390 + }, + { + "epoch": 0.9257744360902256, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 20391 + }, + { + "epoch": 0.9258345864661655, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 20392 + }, + { + "epoch": 0.9258947368421052, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 20393 + }, + { + "epoch": 0.9259548872180451, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 20394 + }, + { + "epoch": 0.926015037593985, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 20395 + }, + { + "epoch": 0.9260751879699248, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 20396 + }, + { + "epoch": 0.9261353383458647, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 20397 + }, + { + "epoch": 0.9261954887218045, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 20398 + }, + { + "epoch": 0.9262556390977443, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 20399 + }, + { + "epoch": 0.9263157894736842, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 20400 + }, + { + "epoch": 0.9263759398496241, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 20401 + }, + { + "epoch": 0.926436090225564, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 20402 + }, + { + "epoch": 0.9264962406015037, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 20403 + }, + { + "epoch": 0.9265563909774436, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 20404 + }, + { + "epoch": 0.9266165413533834, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1772, + "step": 20405 + }, + { + "epoch": 0.9266766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 20406 + }, + { + "epoch": 0.9267368421052632, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 20407 + }, + { + "epoch": 0.926796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 20408 + }, + { + "epoch": 0.9268571428571428, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 20409 + }, + { + "epoch": 0.9269172932330827, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 20410 + }, + { + "epoch": 0.9269774436090226, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 20411 + }, + { + "epoch": 0.9270375939849624, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1314, + "step": 20412 + }, + { + "epoch": 0.9270977443609023, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 20413 + }, + { + "epoch": 0.9271578947368421, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 20414 + }, + { + "epoch": 0.9272180451127819, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 20415 + }, + { + "epoch": 0.9272781954887218, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 20416 + }, + { + "epoch": 0.9273383458646617, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 20417 + }, + { + "epoch": 0.9273984962406016, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 20418 + }, + { + "epoch": 0.9274586466165413, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 20419 + }, + { + "epoch": 0.9275187969924812, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 20420 + }, + { + "epoch": 0.9275789473684211, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 20421 + }, + { + "epoch": 0.9276390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 20422 + }, + { + "epoch": 0.9276992481203008, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 20423 + }, + { + "epoch": 0.9277593984962406, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 20424 + }, + { + "epoch": 0.9278195488721804, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.3679, + "step": 20425 + }, + { + "epoch": 0.9278796992481203, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1636, + "step": 20426 + }, + { + "epoch": 0.9279398496240602, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 20427 + }, + { + "epoch": 0.928, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 20428 + }, + { + "epoch": 0.9280601503759398, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 20429 + }, + { + "epoch": 0.9281203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 20430 + }, + { + "epoch": 0.9281804511278196, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 20431 + }, + { + "epoch": 0.9282406015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 20432 + }, + { + "epoch": 0.9283007518796993, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 20433 + }, + { + "epoch": 0.928360902255639, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.325, + "step": 20434 + }, + { + "epoch": 0.9284210526315789, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1831, + "step": 20435 + }, + { + "epoch": 0.9284812030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 20436 + }, + { + "epoch": 0.9285413533834587, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 20437 + }, + { + "epoch": 0.9286015037593985, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 20438 + }, + { + "epoch": 0.9286616541353383, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 20439 + }, + { + "epoch": 0.9287218045112782, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 20440 + }, + { + "epoch": 0.928781954887218, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 20441 + }, + { + "epoch": 0.9288421052631579, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 20442 + }, + { + "epoch": 0.9289022556390978, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 20443 + }, + { + "epoch": 0.9289624060150375, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 20444 + }, + { + "epoch": 0.9290225563909774, + "grad_norm": 1.53125, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 20445 + }, + { + "epoch": 0.9290827067669173, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 20446 + }, + { + "epoch": 0.9291428571428572, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 20447 + }, + { + "epoch": 0.929203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 20448 + }, + { + "epoch": 0.9292631578947368, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 20449 + }, + { + "epoch": 0.9293233082706767, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 20450 + }, + { + "epoch": 0.9293834586466165, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 20451 + }, + { + "epoch": 0.9294436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 20452 + }, + { + "epoch": 0.9295037593984963, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1271, + "step": 20453 + }, + { + "epoch": 0.929563909774436, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 20454 + }, + { + "epoch": 0.9296240601503759, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.271, + "step": 20455 + }, + { + "epoch": 0.9296842105263158, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 20456 + }, + { + "epoch": 0.9297443609022557, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 20457 + }, + { + "epoch": 0.9298045112781955, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 20458 + }, + { + "epoch": 0.9298646616541354, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 20459 + }, + { + "epoch": 0.9299248120300752, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 20460 + }, + { + "epoch": 0.929984962406015, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 20461 + }, + { + "epoch": 0.9300451127819549, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 20462 + }, + { + "epoch": 0.9301052631578948, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 20463 + }, + { + "epoch": 0.9301654135338346, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 20464 + }, + { + "epoch": 0.9302255639097744, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 20465 + }, + { + "epoch": 0.9302857142857143, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 20466 + }, + { + "epoch": 0.9303458646616541, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 20467 + }, + { + "epoch": 0.930406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 20468 + }, + { + "epoch": 0.9304661654135339, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 20469 + }, + { + "epoch": 0.9305263157894736, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 20470 + }, + { + "epoch": 0.9305864661654135, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 20471 + }, + { + "epoch": 0.9306466165413534, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 20472 + }, + { + "epoch": 0.9307067669172933, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 20473 + }, + { + "epoch": 0.9307669172932331, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 20474 + }, + { + "epoch": 0.9308270676691729, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 20475 + }, + { + "epoch": 0.9308872180451128, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 20476 + }, + { + "epoch": 0.9309473684210526, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 20477 + }, + { + "epoch": 0.9310075187969925, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 20478 + }, + { + "epoch": 0.9310676691729324, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1564, + "step": 20479 + }, + { + "epoch": 0.9311278195488721, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.214, + "step": 20480 + }, + { + "epoch": 0.931187969924812, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 20481 + }, + { + "epoch": 0.9312481203007519, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 20482 + }, + { + "epoch": 0.9313082706766918, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 20483 + }, + { + "epoch": 0.9313684210526316, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2017, + "step": 20484 + }, + { + "epoch": 0.9314285714285714, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 20485 + }, + { + "epoch": 0.9314887218045113, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2062, + "step": 20486 + }, + { + "epoch": 0.9315488721804511, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 20487 + }, + { + "epoch": 0.931609022556391, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 20488 + }, + { + "epoch": 0.9316691729323309, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 20489 + }, + { + "epoch": 0.9317293233082706, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 20490 + }, + { + "epoch": 0.9317894736842105, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 20491 + }, + { + "epoch": 0.9318496240601504, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 20492 + }, + { + "epoch": 0.9319097744360902, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 20493 + }, + { + "epoch": 0.9319699248120301, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 20494 + }, + { + "epoch": 0.9320300751879699, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 20495 + }, + { + "epoch": 0.9320902255639097, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 20496 + }, + { + "epoch": 0.9321503759398496, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 20497 + }, + { + "epoch": 0.9322105263157895, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 20498 + }, + { + "epoch": 0.9322706766917294, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 20499 + }, + { + "epoch": 0.9323308270676691, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 20500 + }, + { + "epoch": 0.932390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 20501 + }, + { + "epoch": 0.9324511278195489, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 20502 + }, + { + "epoch": 0.9325112781954887, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 20503 + }, + { + "epoch": 0.9325714285714286, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 20504 + }, + { + "epoch": 0.9326315789473684, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 20505 + }, + { + "epoch": 0.9326917293233082, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 20506 + }, + { + "epoch": 0.9327518796992481, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 20507 + }, + { + "epoch": 0.932812030075188, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1768, + "step": 20508 + }, + { + "epoch": 0.9328721804511279, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 20509 + }, + { + "epoch": 0.9329323308270677, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 20510 + }, + { + "epoch": 0.9329924812030075, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 20511 + }, + { + "epoch": 0.9330526315789474, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 20512 + }, + { + "epoch": 0.9331127819548872, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 20513 + }, + { + "epoch": 0.9331729323308271, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 20514 + }, + { + "epoch": 0.933233082706767, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 20515 + }, + { + "epoch": 0.9332932330827067, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 20516 + }, + { + "epoch": 0.9333533834586466, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 20517 + }, + { + "epoch": 0.9334135338345865, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 20518 + }, + { + "epoch": 0.9334736842105263, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 20519 + }, + { + "epoch": 0.9335338345864662, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 20520 + }, + { + "epoch": 0.933593984962406, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 20521 + }, + { + "epoch": 0.9336541353383458, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 20522 + }, + { + "epoch": 0.9337142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 20523 + }, + { + "epoch": 0.9337744360902256, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 20524 + }, + { + "epoch": 0.9338345864661655, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 20525 + }, + { + "epoch": 0.9338947368421052, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 20526 + }, + { + "epoch": 0.9339548872180451, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 20527 + }, + { + "epoch": 0.934015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 20528 + }, + { + "epoch": 0.9340751879699248, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 20529 + }, + { + "epoch": 0.9341353383458647, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 20530 + }, + { + "epoch": 0.9341954887218045, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 20531 + }, + { + "epoch": 0.9342556390977443, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 20532 + }, + { + "epoch": 0.9343157894736842, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 20533 + }, + { + "epoch": 0.9343759398496241, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 20534 + }, + { + "epoch": 0.934436090225564, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 20535 + }, + { + "epoch": 0.9344962406015037, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 20536 + }, + { + "epoch": 0.9345563909774436, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 20537 + }, + { + "epoch": 0.9346165413533835, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 20538 + }, + { + "epoch": 0.9346766917293233, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 20539 + }, + { + "epoch": 0.9347368421052632, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 20540 + }, + { + "epoch": 0.934796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 20541 + }, + { + "epoch": 0.9348571428571428, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 20542 + }, + { + "epoch": 0.9349172932330827, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 20543 + }, + { + "epoch": 0.9349774436090226, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 20544 + }, + { + "epoch": 0.9350375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2053, + "step": 20545 + }, + { + "epoch": 0.9350977443609022, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 20546 + }, + { + "epoch": 0.9351578947368421, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 20547 + }, + { + "epoch": 0.935218045112782, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 20548 + }, + { + "epoch": 0.9352781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 20549 + }, + { + "epoch": 0.9353383458646617, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.156, + "step": 20550 + }, + { + "epoch": 0.9353984962406015, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 20551 + }, + { + "epoch": 0.9354586466165413, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1482, + "step": 20552 + }, + { + "epoch": 0.9355187969924812, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 20553 + }, + { + "epoch": 0.9355789473684211, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 20554 + }, + { + "epoch": 0.9356390977443609, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 20555 + }, + { + "epoch": 0.9356992481203007, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 20556 + }, + { + "epoch": 0.9357593984962406, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 20557 + }, + { + "epoch": 0.9358195488721804, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 20558 + }, + { + "epoch": 0.9358796992481203, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.2703, + "step": 20559 + }, + { + "epoch": 0.9359398496240602, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1962, + "step": 20560 + }, + { + "epoch": 0.936, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 20561 + }, + { + "epoch": 0.9360601503759398, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 20562 + }, + { + "epoch": 0.9361203007518797, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 20563 + }, + { + "epoch": 0.9361804511278196, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 20564 + }, + { + "epoch": 0.9362406015037594, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 20565 + }, + { + "epoch": 0.9363007518796993, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 20566 + }, + { + "epoch": 0.9363609022556391, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 20567 + }, + { + "epoch": 0.9364210526315789, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 20568 + }, + { + "epoch": 0.9364812030075188, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 20569 + }, + { + "epoch": 0.9365413533834587, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 20570 + }, + { + "epoch": 0.9366015037593985, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1585, + "step": 20571 + }, + { + "epoch": 0.9366616541353383, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 20572 + }, + { + "epoch": 0.9367218045112782, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 20573 + }, + { + "epoch": 0.936781954887218, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 20574 + }, + { + "epoch": 0.9368421052631579, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 20575 + }, + { + "epoch": 0.9369022556390978, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 20576 + }, + { + "epoch": 0.9369624060150376, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 20577 + }, + { + "epoch": 0.9370225563909774, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 20578 + }, + { + "epoch": 0.9370827067669173, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 20579 + }, + { + "epoch": 0.9371428571428572, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 20580 + }, + { + "epoch": 0.937203007518797, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 20581 + }, + { + "epoch": 0.9372631578947368, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 20582 + }, + { + "epoch": 0.9373233082706767, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 20583 + }, + { + "epoch": 0.9373834586466165, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 20584 + }, + { + "epoch": 0.9374436090225564, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 20585 + }, + { + "epoch": 0.9375037593984963, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 20586 + }, + { + "epoch": 0.937563909774436, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 20587 + }, + { + "epoch": 0.9376240601503759, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 20588 + }, + { + "epoch": 0.9376842105263158, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 20589 + }, + { + "epoch": 0.9377443609022557, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 20590 + }, + { + "epoch": 0.9378045112781955, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2008, + "step": 20591 + }, + { + "epoch": 0.9378646616541353, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 20592 + }, + { + "epoch": 0.9379248120300752, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 20593 + }, + { + "epoch": 0.937984962406015, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 20594 + }, + { + "epoch": 0.9380451127819549, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 20595 + }, + { + "epoch": 0.9381052631578948, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 20596 + }, + { + "epoch": 0.9381654135338345, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 20597 + }, + { + "epoch": 0.9382255639097744, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 20598 + }, + { + "epoch": 0.9382857142857143, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 20599 + }, + { + "epoch": 0.9383458646616541, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 20600 + }, + { + "epoch": 0.938406015037594, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 20601 + }, + { + "epoch": 0.9384661654135338, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1465, + "step": 20602 + }, + { + "epoch": 0.9385263157894737, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 20603 + }, + { + "epoch": 0.9385864661654135, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 20604 + }, + { + "epoch": 0.9386466165413534, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 20605 + }, + { + "epoch": 0.9387067669172933, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 20606 + }, + { + "epoch": 0.938766917293233, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 20607 + }, + { + "epoch": 0.9388270676691729, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 20608 + }, + { + "epoch": 0.9388872180451128, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 20609 + }, + { + "epoch": 0.9389473684210526, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 20610 + }, + { + "epoch": 0.9390075187969925, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 20611 + }, + { + "epoch": 0.9390676691729324, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 20612 + }, + { + "epoch": 0.9391278195488721, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 20613 + }, + { + "epoch": 0.939187969924812, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 20614 + }, + { + "epoch": 0.9392481203007519, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 20615 + }, + { + "epoch": 0.9393082706766918, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 20616 + }, + { + "epoch": 0.9393684210526316, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 20617 + }, + { + "epoch": 0.9394285714285714, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 20618 + }, + { + "epoch": 0.9394887218045113, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 20619 + }, + { + "epoch": 0.9395488721804511, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 20620 + }, + { + "epoch": 0.939609022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 20621 + }, + { + "epoch": 0.9396691729323309, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 20622 + }, + { + "epoch": 0.9397293233082706, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 20623 + }, + { + "epoch": 0.9397894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 20624 + }, + { + "epoch": 0.9398496240601504, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 20625 + }, + { + "epoch": 0.9399097744360903, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1567, + "step": 20626 + }, + { + "epoch": 0.9399699248120301, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 20627 + }, + { + "epoch": 0.9400300751879699, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 20628 + }, + { + "epoch": 0.9400902255639098, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 20629 + }, + { + "epoch": 0.9401503759398496, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 20630 + }, + { + "epoch": 0.9402105263157895, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 20631 + }, + { + "epoch": 0.9402706766917294, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 20632 + }, + { + "epoch": 0.9403308270676691, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 20633 + }, + { + "epoch": 0.940390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 20634 + }, + { + "epoch": 0.9404511278195489, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 20635 + }, + { + "epoch": 0.9405112781954887, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 20636 + }, + { + "epoch": 0.9405714285714286, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 20637 + }, + { + "epoch": 0.9406315789473684, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 20638 + }, + { + "epoch": 0.9406917293233082, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 20639 + }, + { + "epoch": 0.9407518796992481, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 20640 + }, + { + "epoch": 0.940812030075188, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1925, + "step": 20641 + }, + { + "epoch": 0.9408721804511279, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 20642 + }, + { + "epoch": 0.9409323308270676, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 20643 + }, + { + "epoch": 0.9409924812030075, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 20644 + }, + { + "epoch": 0.9410526315789474, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 20645 + }, + { + "epoch": 0.9411127819548872, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 20646 + }, + { + "epoch": 0.9411729323308271, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 20647 + }, + { + "epoch": 0.9412330827067669, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 20648 + }, + { + "epoch": 0.9412932330827067, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 20649 + }, + { + "epoch": 0.9413533834586466, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 20650 + }, + { + "epoch": 0.9414135338345865, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 20651 + }, + { + "epoch": 0.9414736842105264, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 20652 + }, + { + "epoch": 0.9415338345864661, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 20653 + }, + { + "epoch": 0.941593984962406, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 20654 + }, + { + "epoch": 0.9416541353383459, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 20655 + }, + { + "epoch": 0.9417142857142857, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 20656 + }, + { + "epoch": 0.9417744360902256, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 20657 + }, + { + "epoch": 0.9418345864661655, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 20658 + }, + { + "epoch": 0.9418947368421052, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.3157, + "step": 20659 + }, + { + "epoch": 0.9419548872180451, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 20660 + }, + { + "epoch": 0.942015037593985, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 20661 + }, + { + "epoch": 0.9420751879699248, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 20662 + }, + { + "epoch": 0.9421353383458647, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 20663 + }, + { + "epoch": 0.9421954887218045, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 20664 + }, + { + "epoch": 0.9422556390977443, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 20665 + }, + { + "epoch": 0.9423157894736842, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 20666 + }, + { + "epoch": 0.9423759398496241, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 20667 + }, + { + "epoch": 0.942436090225564, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 20668 + }, + { + "epoch": 0.9424962406015037, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 20669 + }, + { + "epoch": 0.9425563909774436, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 20670 + }, + { + "epoch": 0.9426165413533835, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 20671 + }, + { + "epoch": 0.9426766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 20672 + }, + { + "epoch": 0.9427368421052632, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 20673 + }, + { + "epoch": 0.942796992481203, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 20674 + }, + { + "epoch": 0.9428571428571428, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 20675 + }, + { + "epoch": 0.9429172932330827, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 20676 + }, + { + "epoch": 0.9429774436090226, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 20677 + }, + { + "epoch": 0.9430375939849625, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 20678 + }, + { + "epoch": 0.9430977443609022, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 20679 + }, + { + "epoch": 0.9431578947368421, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 20680 + }, + { + "epoch": 0.943218045112782, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 20681 + }, + { + "epoch": 0.9432781954887218, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 20682 + }, + { + "epoch": 0.9433383458646617, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 20683 + }, + { + "epoch": 0.9433984962406015, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 20684 + }, + { + "epoch": 0.9434586466165413, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 20685 + }, + { + "epoch": 0.9435187969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 20686 + }, + { + "epoch": 0.9435789473684211, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 20687 + }, + { + "epoch": 0.9436390977443609, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 20688 + }, + { + "epoch": 0.9436992481203007, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 20689 + }, + { + "epoch": 0.9437593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 20690 + }, + { + "epoch": 0.9438195488721804, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 20691 + }, + { + "epoch": 0.9438796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 20692 + }, + { + "epoch": 0.9439398496240602, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 20693 + }, + { + "epoch": 0.944, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1834, + "step": 20694 + }, + { + "epoch": 0.9440601503759398, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 20695 + }, + { + "epoch": 0.9441203007518797, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 20696 + }, + { + "epoch": 0.9441804511278196, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 20697 + }, + { + "epoch": 0.9442406015037594, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 20698 + }, + { + "epoch": 0.9443007518796992, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 20699 + }, + { + "epoch": 0.9443609022556391, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 20700 + }, + { + "epoch": 0.9444210526315789, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 20701 + }, + { + "epoch": 0.9444812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 20702 + }, + { + "epoch": 0.9445413533834587, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 20703 + }, + { + "epoch": 0.9446015037593984, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 20704 + }, + { + "epoch": 0.9446616541353383, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 20705 + }, + { + "epoch": 0.9447218045112782, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2085, + "step": 20706 + }, + { + "epoch": 0.9447819548872181, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 20707 + }, + { + "epoch": 0.9448421052631579, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 20708 + }, + { + "epoch": 0.9449022556390978, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 20709 + }, + { + "epoch": 0.9449624060150376, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 20710 + }, + { + "epoch": 0.9450225563909774, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 20711 + }, + { + "epoch": 0.9450827067669173, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 20712 + }, + { + "epoch": 0.9451428571428572, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 20713 + }, + { + "epoch": 0.945203007518797, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 20714 + }, + { + "epoch": 0.9452631578947368, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 20715 + }, + { + "epoch": 0.9453233082706767, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 20716 + }, + { + "epoch": 0.9453834586466165, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 20717 + }, + { + "epoch": 0.9454436090225564, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.147, + "step": 20718 + }, + { + "epoch": 0.9455037593984963, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 20719 + }, + { + "epoch": 0.945563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 20720 + }, + { + "epoch": 0.9456240601503759, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 20721 + }, + { + "epoch": 0.9456842105263158, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2154, + "step": 20722 + }, + { + "epoch": 0.9457443609022557, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 20723 + }, + { + "epoch": 0.9458045112781955, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 20724 + }, + { + "epoch": 0.9458646616541353, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 20725 + }, + { + "epoch": 0.9459248120300752, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 20726 + }, + { + "epoch": 0.945984962406015, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 20727 + }, + { + "epoch": 0.9460451127819549, + "grad_norm": 1.390625, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 20728 + }, + { + "epoch": 0.9461052631578948, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 20729 + }, + { + "epoch": 0.9461654135338345, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 20730 + }, + { + "epoch": 0.9462255639097744, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 20731 + }, + { + "epoch": 0.9462857142857143, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 20732 + }, + { + "epoch": 0.9463458646616542, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 20733 + }, + { + "epoch": 0.946406015037594, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2464, + "step": 20734 + }, + { + "epoch": 0.9464661654135338, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 20735 + }, + { + "epoch": 0.9465263157894737, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 20736 + }, + { + "epoch": 0.9465864661654135, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 20737 + }, + { + "epoch": 0.9466466165413534, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 20738 + }, + { + "epoch": 0.9467067669172933, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 20739 + }, + { + "epoch": 0.946766917293233, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 20740 + }, + { + "epoch": 0.9468270676691729, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 20741 + }, + { + "epoch": 0.9468872180451128, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 20742 + }, + { + "epoch": 0.9469473684210526, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1314, + "step": 20743 + }, + { + "epoch": 0.9470075187969925, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 20744 + }, + { + "epoch": 0.9470676691729323, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 20745 + }, + { + "epoch": 0.9471278195488722, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 20746 + }, + { + "epoch": 0.947187969924812, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 20747 + }, + { + "epoch": 0.9472481203007519, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 20748 + }, + { + "epoch": 0.9473082706766918, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 20749 + }, + { + "epoch": 0.9473684210526315, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 20750 + }, + { + "epoch": 0.9474285714285714, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 20751 + }, + { + "epoch": 0.9474887218045113, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 20752 + }, + { + "epoch": 0.9475488721804511, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 20753 + }, + { + "epoch": 0.947609022556391, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1254, + "step": 20754 + }, + { + "epoch": 0.9476691729323308, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 20755 + }, + { + "epoch": 0.9477293233082706, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 20756 + }, + { + "epoch": 0.9477894736842105, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 20757 + }, + { + "epoch": 0.9478496240601504, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 20758 + }, + { + "epoch": 0.9479097744360903, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 20759 + }, + { + "epoch": 0.9479699248120301, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 20760 + }, + { + "epoch": 0.9480300751879699, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 20761 + }, + { + "epoch": 0.9480902255639098, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 20762 + }, + { + "epoch": 0.9481503759398496, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 20763 + }, + { + "epoch": 0.9482105263157895, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 20764 + }, + { + "epoch": 0.9482706766917294, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 20765 + }, + { + "epoch": 0.9483308270676691, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 20766 + }, + { + "epoch": 0.948390977443609, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 20767 + }, + { + "epoch": 0.9484511278195489, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 20768 + }, + { + "epoch": 0.9485112781954887, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 20769 + }, + { + "epoch": 0.9485714285714286, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 20770 + }, + { + "epoch": 0.9486315789473684, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 20771 + }, + { + "epoch": 0.9486917293233083, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 20772 + }, + { + "epoch": 0.9487518796992481, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 20773 + }, + { + "epoch": 0.948812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 20774 + }, + { + "epoch": 0.9488721804511279, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 20775 + }, + { + "epoch": 0.9489323308270676, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 20776 + }, + { + "epoch": 0.9489924812030075, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 20777 + }, + { + "epoch": 0.9490526315789474, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 20778 + }, + { + "epoch": 0.9491127819548872, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 20779 + }, + { + "epoch": 0.9491729323308271, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 20780 + }, + { + "epoch": 0.9492330827067669, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 20781 + }, + { + "epoch": 0.9492932330827067, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 20782 + }, + { + "epoch": 0.9493533834586466, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 20783 + }, + { + "epoch": 0.9494135338345865, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 20784 + }, + { + "epoch": 0.9494736842105264, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 20785 + }, + { + "epoch": 0.9495338345864661, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 20786 + }, + { + "epoch": 0.949593984962406, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 20787 + }, + { + "epoch": 0.9496541353383459, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 20788 + }, + { + "epoch": 0.9497142857142857, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 20789 + }, + { + "epoch": 0.9497744360902256, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 20790 + }, + { + "epoch": 0.9498345864661654, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 20791 + }, + { + "epoch": 0.9498947368421052, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 20792 + }, + { + "epoch": 0.9499548872180451, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 20793 + }, + { + "epoch": 0.950015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 20794 + }, + { + "epoch": 0.9500751879699249, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 20795 + }, + { + "epoch": 0.9501353383458646, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 20796 + }, + { + "epoch": 0.9501954887218045, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 20797 + }, + { + "epoch": 0.9502556390977444, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 20798 + }, + { + "epoch": 0.9503157894736842, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 20799 + }, + { + "epoch": 0.9503759398496241, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 20800 + }, + { + "epoch": 0.9504360902255639, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 20801 + }, + { + "epoch": 0.9504962406015037, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 20802 + }, + { + "epoch": 0.9505563909774436, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 20803 + }, + { + "epoch": 0.9506165413533835, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 20804 + }, + { + "epoch": 0.9506766917293233, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 20805 + }, + { + "epoch": 0.9507368421052631, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 20806 + }, + { + "epoch": 0.950796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 20807 + }, + { + "epoch": 0.9508571428571428, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 20808 + }, + { + "epoch": 0.9509172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 20809 + }, + { + "epoch": 0.9509774436090226, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 20810 + }, + { + "epoch": 0.9510375939849625, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 20811 + }, + { + "epoch": 0.9510977443609022, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 20812 + }, + { + "epoch": 0.9511578947368421, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 20813 + }, + { + "epoch": 0.951218045112782, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 20814 + }, + { + "epoch": 0.9512781954887218, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 20815 + }, + { + "epoch": 0.9513383458646617, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 20816 + }, + { + "epoch": 0.9513984962406015, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 20817 + }, + { + "epoch": 0.9514586466165413, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 20818 + }, + { + "epoch": 0.9515187969924812, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 20819 + }, + { + "epoch": 0.9515789473684211, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 20820 + }, + { + "epoch": 0.951639097744361, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 20821 + }, + { + "epoch": 0.9516992481203007, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 20822 + }, + { + "epoch": 0.9517593984962406, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 20823 + }, + { + "epoch": 0.9518195488721805, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 20824 + }, + { + "epoch": 0.9518796992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1662, + "step": 20825 + }, + { + "epoch": 0.9519398496240602, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 20826 + }, + { + "epoch": 0.952, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 20827 + }, + { + "epoch": 0.9520601503759398, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 20828 + }, + { + "epoch": 0.9521203007518797, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 20829 + }, + { + "epoch": 0.9521804511278196, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 20830 + }, + { + "epoch": 0.9522406015037594, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 20831 + }, + { + "epoch": 0.9523007518796992, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 20832 + }, + { + "epoch": 0.9523609022556391, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 20833 + }, + { + "epoch": 0.952421052631579, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 20834 + }, + { + "epoch": 0.9524812030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 20835 + }, + { + "epoch": 0.9525413533834587, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 20836 + }, + { + "epoch": 0.9526015037593984, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 20837 + }, + { + "epoch": 0.9526616541353383, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 20838 + }, + { + "epoch": 0.9527218045112782, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 20839 + }, + { + "epoch": 0.9527819548872181, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 20840 + }, + { + "epoch": 0.9528421052631579, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2948, + "step": 20841 + }, + { + "epoch": 0.9529022556390977, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 20842 + }, + { + "epoch": 0.9529624060150376, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 20843 + }, + { + "epoch": 0.9530225563909774, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 20844 + }, + { + "epoch": 0.9530827067669173, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 20845 + }, + { + "epoch": 0.9531428571428572, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 20846 + }, + { + "epoch": 0.9532030075187969, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 20847 + }, + { + "epoch": 0.9532631578947368, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 20848 + }, + { + "epoch": 0.9533233082706767, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1797, + "step": 20849 + }, + { + "epoch": 0.9533834586466166, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 20850 + }, + { + "epoch": 0.9534436090225564, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.3117, + "step": 20851 + }, + { + "epoch": 0.9535037593984962, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 20852 + }, + { + "epoch": 0.9535639097744361, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 20853 + }, + { + "epoch": 0.9536240601503759, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 20854 + }, + { + "epoch": 0.9536842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 20855 + }, + { + "epoch": 0.9537443609022557, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 20856 + }, + { + "epoch": 0.9538045112781955, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 20857 + }, + { + "epoch": 0.9538646616541353, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 20858 + }, + { + "epoch": 0.9539248120300752, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 20859 + }, + { + "epoch": 0.953984962406015, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 20860 + }, + { + "epoch": 0.9540451127819549, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 20861 + }, + { + "epoch": 0.9541052631578948, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 20862 + }, + { + "epoch": 0.9541654135338345, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 20863 + }, + { + "epoch": 0.9542255639097744, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 20864 + }, + { + "epoch": 0.9542857142857143, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 20865 + }, + { + "epoch": 0.9543458646616542, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 20866 + }, + { + "epoch": 0.954406015037594, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 20867 + }, + { + "epoch": 0.9544661654135338, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 20868 + }, + { + "epoch": 0.9545263157894737, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1919, + "step": 20869 + }, + { + "epoch": 0.9545864661654135, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 20870 + }, + { + "epoch": 0.9546466165413534, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 20871 + }, + { + "epoch": 0.9547067669172933, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 20872 + }, + { + "epoch": 0.954766917293233, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 20873 + }, + { + "epoch": 0.9548270676691729, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 20874 + }, + { + "epoch": 0.9548872180451128, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 20875 + }, + { + "epoch": 0.9549473684210527, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 20876 + }, + { + "epoch": 0.9550075187969925, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 20877 + }, + { + "epoch": 0.9550676691729323, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 20878 + }, + { + "epoch": 0.9551278195488722, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 20879 + }, + { + "epoch": 0.955187969924812, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 20880 + }, + { + "epoch": 0.9552481203007519, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 20881 + }, + { + "epoch": 0.9553082706766918, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 20882 + }, + { + "epoch": 0.9553684210526315, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 20883 + }, + { + "epoch": 0.9554285714285714, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 20884 + }, + { + "epoch": 0.9554887218045113, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 20885 + }, + { + "epoch": 0.9555488721804511, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.215, + "step": 20886 + }, + { + "epoch": 0.955609022556391, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.2864, + "step": 20887 + }, + { + "epoch": 0.9556691729323308, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 20888 + }, + { + "epoch": 0.9557293233082707, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 20889 + }, + { + "epoch": 0.9557894736842105, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 20890 + }, + { + "epoch": 0.9558496240601504, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 20891 + }, + { + "epoch": 0.9559097744360903, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 20892 + }, + { + "epoch": 0.95596992481203, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 20893 + }, + { + "epoch": 0.9560300751879699, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 20894 + }, + { + "epoch": 0.9560902255639098, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 20895 + }, + { + "epoch": 0.9561503759398496, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 20896 + }, + { + "epoch": 0.9562105263157895, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 20897 + }, + { + "epoch": 0.9562706766917293, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 20898 + }, + { + "epoch": 0.9563308270676691, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2129, + "step": 20899 + }, + { + "epoch": 0.956390977443609, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 20900 + }, + { + "epoch": 0.9564511278195489, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 20901 + }, + { + "epoch": 0.9565112781954888, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 20902 + }, + { + "epoch": 0.9565714285714285, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 20903 + }, + { + "epoch": 0.9566315789473684, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 20904 + }, + { + "epoch": 0.9566917293233083, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 20905 + }, + { + "epoch": 0.9567518796992481, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 20906 + }, + { + "epoch": 0.956812030075188, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 20907 + }, + { + "epoch": 0.9568721804511279, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 20908 + }, + { + "epoch": 0.9569323308270676, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 20909 + }, + { + "epoch": 0.9569924812030075, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 20910 + }, + { + "epoch": 0.9570526315789474, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 20911 + }, + { + "epoch": 0.9571127819548872, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 20912 + }, + { + "epoch": 0.9571729323308271, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 20913 + }, + { + "epoch": 0.9572330827067669, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 20914 + }, + { + "epoch": 0.9572932330827068, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 20915 + }, + { + "epoch": 0.9573533834586466, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 20916 + }, + { + "epoch": 0.9574135338345865, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 20917 + }, + { + "epoch": 0.9574736842105264, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 20918 + }, + { + "epoch": 0.9575338345864661, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 20919 + }, + { + "epoch": 0.957593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 20920 + }, + { + "epoch": 0.9576541353383459, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 20921 + }, + { + "epoch": 0.9577142857142857, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 20922 + }, + { + "epoch": 0.9577744360902256, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 20923 + }, + { + "epoch": 0.9578345864661654, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 20924 + }, + { + "epoch": 0.9578947368421052, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 20925 + }, + { + "epoch": 0.9579548872180451, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 20926 + }, + { + "epoch": 0.958015037593985, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 20927 + }, + { + "epoch": 0.9580751879699249, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 20928 + }, + { + "epoch": 0.9581353383458646, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 20929 + }, + { + "epoch": 0.9581954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 20930 + }, + { + "epoch": 0.9582556390977444, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 20931 + }, + { + "epoch": 0.9583157894736842, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 20932 + }, + { + "epoch": 0.9583759398496241, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 20933 + }, + { + "epoch": 0.9584360902255639, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 20934 + }, + { + "epoch": 0.9584962406015037, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 20935 + }, + { + "epoch": 0.9585563909774436, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 20936 + }, + { + "epoch": 0.9586165413533835, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 20937 + }, + { + "epoch": 0.9586766917293233, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 20938 + }, + { + "epoch": 0.9587368421052631, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 20939 + }, + { + "epoch": 0.958796992481203, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 20940 + }, + { + "epoch": 0.9588571428571429, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 20941 + }, + { + "epoch": 0.9589172932330827, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 20942 + }, + { + "epoch": 0.9589774436090226, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 20943 + }, + { + "epoch": 0.9590375939849624, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 20944 + }, + { + "epoch": 0.9590977443609022, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 20945 + }, + { + "epoch": 0.9591578947368421, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 20946 + }, + { + "epoch": 0.959218045112782, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 20947 + }, + { + "epoch": 0.9592781954887218, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 20948 + }, + { + "epoch": 0.9593383458646616, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 20949 + }, + { + "epoch": 0.9593984962406015, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 20950 + }, + { + "epoch": 0.9594586466165413, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 20951 + }, + { + "epoch": 0.9595187969924812, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 20952 + }, + { + "epoch": 0.9595789473684211, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 20953 + }, + { + "epoch": 0.9596390977443608, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 20954 + }, + { + "epoch": 0.9596992481203007, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 20955 + }, + { + "epoch": 0.9597593984962406, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 20956 + }, + { + "epoch": 0.9598195488721805, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 20957 + }, + { + "epoch": 0.9598796992481203, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 20958 + }, + { + "epoch": 0.9599398496240602, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 20959 + }, + { + "epoch": 0.96, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 20960 + }, + { + "epoch": 0.9600601503759398, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 20961 + }, + { + "epoch": 0.9601203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 20962 + }, + { + "epoch": 0.9601804511278196, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 20963 + }, + { + "epoch": 0.9602406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 20964 + }, + { + "epoch": 0.9603007518796992, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 20965 + }, + { + "epoch": 0.9603609022556391, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 20966 + }, + { + "epoch": 0.960421052631579, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 20967 + }, + { + "epoch": 0.9604812030075188, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 20968 + }, + { + "epoch": 0.9605413533834587, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 20969 + }, + { + "epoch": 0.9606015037593985, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 20970 + }, + { + "epoch": 0.9606616541353383, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 20971 + }, + { + "epoch": 0.9607218045112782, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 20972 + }, + { + "epoch": 0.9607819548872181, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 20973 + }, + { + "epoch": 0.9608421052631579, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 20974 + }, + { + "epoch": 0.9609022556390977, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 20975 + }, + { + "epoch": 0.9609624060150376, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 20976 + }, + { + "epoch": 0.9610225563909774, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 20977 + }, + { + "epoch": 0.9610827067669173, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 20978 + }, + { + "epoch": 0.9611428571428572, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 20979 + }, + { + "epoch": 0.961203007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 20980 + }, + { + "epoch": 0.9612631578947368, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 20981 + }, + { + "epoch": 0.9613233082706767, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 20982 + }, + { + "epoch": 0.9613834586466166, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 20983 + }, + { + "epoch": 0.9614436090225564, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 20984 + }, + { + "epoch": 0.9615037593984962, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 20985 + }, + { + "epoch": 0.9615639097744361, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 20986 + }, + { + "epoch": 0.9616240601503759, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 20987 + }, + { + "epoch": 0.9616842105263158, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 20988 + }, + { + "epoch": 0.9617443609022557, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 20989 + }, + { + "epoch": 0.9618045112781954, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 20990 + }, + { + "epoch": 0.9618646616541353, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 20991 + }, + { + "epoch": 0.9619248120300752, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 20992 + }, + { + "epoch": 0.961984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 20993 + }, + { + "epoch": 0.9620451127819549, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 20994 + }, + { + "epoch": 0.9621052631578947, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 20995 + }, + { + "epoch": 0.9621654135338346, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 20996 + }, + { + "epoch": 0.9622255639097744, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 20997 + }, + { + "epoch": 0.9622857142857143, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 20998 + }, + { + "epoch": 0.9623458646616542, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 20999 + }, + { + "epoch": 0.9624060150375939, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 21000 + }, + { + "epoch": 0.9624661654135338, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2032, + "step": 21001 + }, + { + "epoch": 0.9625263157894737, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 21002 + }, + { + "epoch": 0.9625864661654135, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 21003 + }, + { + "epoch": 0.9626466165413534, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 21004 + }, + { + "epoch": 0.9627067669172933, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 21005 + }, + { + "epoch": 0.962766917293233, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 21006 + }, + { + "epoch": 0.9628270676691729, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.2148, + "step": 21007 + }, + { + "epoch": 0.9628872180451128, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 21008 + }, + { + "epoch": 0.9629473684210527, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1421, + "step": 21009 + }, + { + "epoch": 0.9630075187969925, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 21010 + }, + { + "epoch": 0.9630676691729323, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 21011 + }, + { + "epoch": 0.9631278195488722, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 21012 + }, + { + "epoch": 0.963187969924812, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 21013 + }, + { + "epoch": 0.9632481203007519, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 21014 + }, + { + "epoch": 0.9633082706766918, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 21015 + }, + { + "epoch": 0.9633684210526315, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 21016 + }, + { + "epoch": 0.9634285714285714, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 21017 + }, + { + "epoch": 0.9634887218045113, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2337, + "step": 21018 + }, + { + "epoch": 0.9635488721804512, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 21019 + }, + { + "epoch": 0.963609022556391, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 21020 + }, + { + "epoch": 0.9636691729323308, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 21021 + }, + { + "epoch": 0.9637293233082707, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 21022 + }, + { + "epoch": 0.9637894736842105, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 21023 + }, + { + "epoch": 0.9638496240601504, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 21024 + }, + { + "epoch": 0.9639097744360903, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 21025 + }, + { + "epoch": 0.96396992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 21026 + }, + { + "epoch": 0.9640300751879699, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 21027 + }, + { + "epoch": 0.9640902255639098, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 21028 + }, + { + "epoch": 0.9641503759398496, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 21029 + }, + { + "epoch": 0.9642105263157895, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 21030 + }, + { + "epoch": 0.9642706766917293, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2263, + "step": 21031 + }, + { + "epoch": 0.9643308270676691, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 21032 + }, + { + "epoch": 0.964390977443609, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 21033 + }, + { + "epoch": 0.9644511278195489, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 21034 + }, + { + "epoch": 0.9645112781954888, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 21035 + }, + { + "epoch": 0.9645714285714285, + "grad_norm": 1.609375, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 21036 + }, + { + "epoch": 0.9646315789473684, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 21037 + }, + { + "epoch": 0.9646917293233083, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 21038 + }, + { + "epoch": 0.9647518796992481, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 21039 + }, + { + "epoch": 0.964812030075188, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 21040 + }, + { + "epoch": 0.9648721804511278, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 21041 + }, + { + "epoch": 0.9649323308270676, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 21042 + }, + { + "epoch": 0.9649924812030075, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 21043 + }, + { + "epoch": 0.9650526315789474, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 21044 + }, + { + "epoch": 0.9651127819548873, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 21045 + }, + { + "epoch": 0.965172932330827, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 21046 + }, + { + "epoch": 0.9652330827067669, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1865, + "step": 21047 + }, + { + "epoch": 0.9652932330827068, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 21048 + }, + { + "epoch": 0.9653533834586466, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 21049 + }, + { + "epoch": 0.9654135338345865, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 21050 + }, + { + "epoch": 0.9654736842105263, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 21051 + }, + { + "epoch": 0.9655338345864661, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 21052 + }, + { + "epoch": 0.965593984962406, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 21053 + }, + { + "epoch": 0.9656541353383459, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 21054 + }, + { + "epoch": 0.9657142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 21055 + }, + { + "epoch": 0.9657744360902256, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 21056 + }, + { + "epoch": 0.9658345864661654, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 21057 + }, + { + "epoch": 0.9658947368421053, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 21058 + }, + { + "epoch": 0.9659548872180451, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 21059 + }, + { + "epoch": 0.966015037593985, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 21060 + }, + { + "epoch": 0.9660751879699249, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 21061 + }, + { + "epoch": 0.9661353383458646, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 21062 + }, + { + "epoch": 0.9661954887218045, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 21063 + }, + { + "epoch": 0.9662556390977444, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 21064 + }, + { + "epoch": 0.9663157894736842, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.2313, + "step": 21065 + }, + { + "epoch": 0.9663759398496241, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 21066 + }, + { + "epoch": 0.9664360902255639, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 21067 + }, + { + "epoch": 0.9664962406015037, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 21068 + }, + { + "epoch": 0.9665563909774436, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 21069 + }, + { + "epoch": 0.9666165413533835, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 21070 + }, + { + "epoch": 0.9666766917293234, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 21071 + }, + { + "epoch": 0.9667368421052631, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 21072 + }, + { + "epoch": 0.966796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 21073 + }, + { + "epoch": 0.9668571428571429, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2161, + "step": 21074 + }, + { + "epoch": 0.9669172932330827, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 21075 + }, + { + "epoch": 0.9669774436090226, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 21076 + }, + { + "epoch": 0.9670375939849624, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 21077 + }, + { + "epoch": 0.9670977443609022, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 21078 + }, + { + "epoch": 0.9671578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 21079 + }, + { + "epoch": 0.967218045112782, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 21080 + }, + { + "epoch": 0.9672781954887218, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 21081 + }, + { + "epoch": 0.9673383458646616, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 21082 + }, + { + "epoch": 0.9673984962406015, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 21083 + }, + { + "epoch": 0.9674586466165414, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 21084 + }, + { + "epoch": 0.9675187969924812, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 21085 + }, + { + "epoch": 0.9675789473684211, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 21086 + }, + { + "epoch": 0.9676390977443609, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 21087 + }, + { + "epoch": 0.9676992481203007, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 21088 + }, + { + "epoch": 0.9677593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 21089 + }, + { + "epoch": 0.9678195488721805, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 21090 + }, + { + "epoch": 0.9678796992481203, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 21091 + }, + { + "epoch": 0.9679398496240601, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 21092 + }, + { + "epoch": 0.968, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 21093 + }, + { + "epoch": 0.9680601503759398, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 21094 + }, + { + "epoch": 0.9681203007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 21095 + }, + { + "epoch": 0.9681804511278196, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 21096 + }, + { + "epoch": 0.9682406015037593, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 21097 + }, + { + "epoch": 0.9683007518796992, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 21098 + }, + { + "epoch": 0.9683609022556391, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.32, + "step": 21099 + }, + { + "epoch": 0.968421052631579, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 21100 + }, + { + "epoch": 0.9684812030075188, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 21101 + }, + { + "epoch": 0.9685413533834586, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 21102 + }, + { + "epoch": 0.9686015037593985, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 21103 + }, + { + "epoch": 0.9686616541353383, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 21104 + }, + { + "epoch": 0.9687218045112782, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 21105 + }, + { + "epoch": 0.9687819548872181, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 21106 + }, + { + "epoch": 0.968842105263158, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 21107 + }, + { + "epoch": 0.9689022556390977, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 21108 + }, + { + "epoch": 0.9689624060150376, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2524, + "step": 21109 + }, + { + "epoch": 0.9690225563909775, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 21110 + }, + { + "epoch": 0.9690827067669173, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2306, + "step": 21111 + }, + { + "epoch": 0.9691428571428572, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 21112 + }, + { + "epoch": 0.969203007518797, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 21113 + }, + { + "epoch": 0.9692631578947368, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 21114 + }, + { + "epoch": 0.9693233082706767, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2109, + "step": 21115 + }, + { + "epoch": 0.9693834586466166, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 21116 + }, + { + "epoch": 0.9694436090225564, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 21117 + }, + { + "epoch": 0.9695037593984962, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 21118 + }, + { + "epoch": 0.9695639097744361, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2343, + "step": 21119 + }, + { + "epoch": 0.9696240601503759, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 21120 + }, + { + "epoch": 0.9696842105263158, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 21121 + }, + { + "epoch": 0.9697443609022557, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 21122 + }, + { + "epoch": 0.9698045112781954, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1996, + "step": 21123 + }, + { + "epoch": 0.9698646616541353, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 21124 + }, + { + "epoch": 0.9699248120300752, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2133, + "step": 21125 + }, + { + "epoch": 0.9699849624060151, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 21126 + }, + { + "epoch": 0.9700451127819549, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 21127 + }, + { + "epoch": 0.9701052631578947, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 21128 + }, + { + "epoch": 0.9701654135338346, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 21129 + }, + { + "epoch": 0.9702255639097744, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 21130 + }, + { + "epoch": 0.9702857142857143, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 21131 + }, + { + "epoch": 0.9703458646616542, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 21132 + }, + { + "epoch": 0.9704060150375939, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 21133 + }, + { + "epoch": 0.9704661654135338, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 21134 + }, + { + "epoch": 0.9705263157894737, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.4561, + "step": 21135 + }, + { + "epoch": 0.9705864661654136, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 21136 + }, + { + "epoch": 0.9706466165413534, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 21137 + }, + { + "epoch": 0.9707067669172932, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 21138 + }, + { + "epoch": 0.9707669172932331, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 21139 + }, + { + "epoch": 0.9708270676691729, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 21140 + }, + { + "epoch": 0.9708872180451128, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 21141 + }, + { + "epoch": 0.9709473684210527, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 21142 + }, + { + "epoch": 0.9710075187969924, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 21143 + }, + { + "epoch": 0.9710676691729323, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 21144 + }, + { + "epoch": 0.9711278195488722, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 21145 + }, + { + "epoch": 0.971187969924812, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2731, + "step": 21146 + }, + { + "epoch": 0.9712481203007519, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 21147 + }, + { + "epoch": 0.9713082706766917, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 21148 + }, + { + "epoch": 0.9713684210526315, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 21149 + }, + { + "epoch": 0.9714285714285714, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 21150 + }, + { + "epoch": 0.9714887218045113, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 21151 + }, + { + "epoch": 0.9715488721804512, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 21152 + }, + { + "epoch": 0.9716090225563909, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 21153 + }, + { + "epoch": 0.9716691729323308, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 21154 + }, + { + "epoch": 0.9717293233082707, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 21155 + }, + { + "epoch": 0.9717894736842105, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 21156 + }, + { + "epoch": 0.9718496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 21157 + }, + { + "epoch": 0.9719097744360903, + "grad_norm": 1.875, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 21158 + }, + { + "epoch": 0.97196992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 21159 + }, + { + "epoch": 0.9720300751879699, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 21160 + }, + { + "epoch": 0.9720902255639098, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 21161 + }, + { + "epoch": 0.9721503759398497, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 21162 + }, + { + "epoch": 0.9722105263157895, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 21163 + }, + { + "epoch": 0.9722706766917293, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 21164 + }, + { + "epoch": 0.9723308270676692, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 21165 + }, + { + "epoch": 0.972390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 21166 + }, + { + "epoch": 0.9724511278195489, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 21167 + }, + { + "epoch": 0.9725112781954888, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 21168 + }, + { + "epoch": 0.9725714285714285, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 21169 + }, + { + "epoch": 0.9726315789473684, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 21170 + }, + { + "epoch": 0.9726917293233083, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 21171 + }, + { + "epoch": 0.9727518796992481, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 21172 + }, + { + "epoch": 0.972812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 21173 + }, + { + "epoch": 0.9728721804511278, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 21174 + }, + { + "epoch": 0.9729323308270676, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 21175 + }, + { + "epoch": 0.9729924812030075, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 21176 + }, + { + "epoch": 0.9730526315789474, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 21177 + }, + { + "epoch": 0.9731127819548873, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 21178 + }, + { + "epoch": 0.973172932330827, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 21179 + }, + { + "epoch": 0.9732330827067669, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 21180 + }, + { + "epoch": 0.9732932330827068, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 21181 + }, + { + "epoch": 0.9733533834586466, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 21182 + }, + { + "epoch": 0.9734135338345865, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 21183 + }, + { + "epoch": 0.9734736842105263, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 21184 + }, + { + "epoch": 0.9735338345864661, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 21185 + }, + { + "epoch": 0.973593984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 21186 + }, + { + "epoch": 0.9736541353383459, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 21187 + }, + { + "epoch": 0.9737142857142858, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 21188 + }, + { + "epoch": 0.9737744360902255, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 21189 + }, + { + "epoch": 0.9738345864661654, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 21190 + }, + { + "epoch": 0.9738947368421053, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 21191 + }, + { + "epoch": 0.9739548872180451, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 21192 + }, + { + "epoch": 0.974015037593985, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 21193 + }, + { + "epoch": 0.9740751879699248, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 21194 + }, + { + "epoch": 0.9741353383458646, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 21195 + }, + { + "epoch": 0.9741954887218045, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 21196 + }, + { + "epoch": 0.9742556390977444, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 21197 + }, + { + "epoch": 0.9743157894736842, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 21198 + }, + { + "epoch": 0.974375939849624, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 21199 + }, + { + "epoch": 0.9744360902255639, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 21200 + }, + { + "epoch": 0.9744962406015037, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 21201 + }, + { + "epoch": 0.9745563909774436, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 21202 + }, + { + "epoch": 0.9746165413533835, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 21203 + }, + { + "epoch": 0.9746766917293234, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2957, + "step": 21204 + }, + { + "epoch": 0.9747368421052631, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 21205 + }, + { + "epoch": 0.974796992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 21206 + }, + { + "epoch": 0.9748571428571429, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 21207 + }, + { + "epoch": 0.9749172932330827, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1954, + "step": 21208 + }, + { + "epoch": 0.9749774436090226, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 21209 + }, + { + "epoch": 0.9750375939849624, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 21210 + }, + { + "epoch": 0.9750977443609022, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 21211 + }, + { + "epoch": 0.9751578947368421, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 21212 + }, + { + "epoch": 0.975218045112782, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 21213 + }, + { + "epoch": 0.9752781954887219, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 21214 + }, + { + "epoch": 0.9753383458646616, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 21215 + }, + { + "epoch": 0.9753984962406015, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 21216 + }, + { + "epoch": 0.9754586466165414, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 21217 + }, + { + "epoch": 0.9755187969924812, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2078, + "step": 21218 + }, + { + "epoch": 0.9755789473684211, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1937, + "step": 21219 + }, + { + "epoch": 0.9756390977443609, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 21220 + }, + { + "epoch": 0.9756992481203007, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 21221 + }, + { + "epoch": 0.9757593984962406, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 21222 + }, + { + "epoch": 0.9758195488721805, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 21223 + }, + { + "epoch": 0.9758796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 21224 + }, + { + "epoch": 0.9759398496240601, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 21225 + }, + { + "epoch": 0.976, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 21226 + }, + { + "epoch": 0.9760601503759399, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 21227 + }, + { + "epoch": 0.9761203007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 21228 + }, + { + "epoch": 0.9761804511278196, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 21229 + }, + { + "epoch": 0.9762406015037594, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 21230 + }, + { + "epoch": 0.9763007518796992, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 21231 + }, + { + "epoch": 0.9763609022556391, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 21232 + }, + { + "epoch": 0.976421052631579, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 21233 + }, + { + "epoch": 0.9764812030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 21234 + }, + { + "epoch": 0.9765413533834586, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 21235 + }, + { + "epoch": 0.9766015037593985, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 21236 + }, + { + "epoch": 0.9766616541353383, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 21237 + }, + { + "epoch": 0.9767218045112782, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 21238 + }, + { + "epoch": 0.9767819548872181, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 21239 + }, + { + "epoch": 0.9768421052631578, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 21240 + }, + { + "epoch": 0.9769022556390977, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1271, + "step": 21241 + }, + { + "epoch": 0.9769624060150376, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 21242 + }, + { + "epoch": 0.9770225563909775, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 21243 + }, + { + "epoch": 0.9770827067669173, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 21244 + }, + { + "epoch": 0.9771428571428571, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 21245 + }, + { + "epoch": 0.977203007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 21246 + }, + { + "epoch": 0.9772631578947368, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 21247 + }, + { + "epoch": 0.9773233082706767, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 21248 + }, + { + "epoch": 0.9773834586466166, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 21249 + }, + { + "epoch": 0.9774436090225563, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 21250 + }, + { + "epoch": 0.9775037593984962, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 21251 + }, + { + "epoch": 0.9775639097744361, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 21252 + }, + { + "epoch": 0.977624060150376, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.3295, + "step": 21253 + }, + { + "epoch": 0.9776842105263158, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 21254 + }, + { + "epoch": 0.9777443609022557, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 21255 + }, + { + "epoch": 0.9778045112781955, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 21256 + }, + { + "epoch": 0.9778646616541353, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 21257 + }, + { + "epoch": 0.9779248120300752, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 21258 + }, + { + "epoch": 0.9779849624060151, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 21259 + }, + { + "epoch": 0.9780451127819549, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 21260 + }, + { + "epoch": 0.9781052631578947, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 21261 + }, + { + "epoch": 0.9781654135338346, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 21262 + }, + { + "epoch": 0.9782255639097744, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 21263 + }, + { + "epoch": 0.9782857142857143, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 21264 + }, + { + "epoch": 0.9783458646616542, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 21265 + }, + { + "epoch": 0.978406015037594, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 21266 + }, + { + "epoch": 0.9784661654135338, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 21267 + }, + { + "epoch": 0.9785263157894737, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 21268 + }, + { + "epoch": 0.9785864661654136, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 21269 + }, + { + "epoch": 0.9786466165413534, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 21270 + }, + { + "epoch": 0.9787067669172932, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9375, + "step": 21271 + }, + { + "epoch": 0.9787669172932331, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 21272 + }, + { + "epoch": 0.9788270676691729, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 21273 + }, + { + "epoch": 0.9788872180451128, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 21274 + }, + { + "epoch": 0.9789473684210527, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 21275 + }, + { + "epoch": 0.9790075187969924, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 21276 + }, + { + "epoch": 0.9790676691729323, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 21277 + }, + { + "epoch": 0.9791278195488722, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 21278 + }, + { + "epoch": 0.979187969924812, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 21279 + }, + { + "epoch": 0.9792481203007519, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 21280 + }, + { + "epoch": 0.9793082706766917, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 21281 + }, + { + "epoch": 0.9793684210526316, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 21282 + }, + { + "epoch": 0.9794285714285714, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 21283 + }, + { + "epoch": 0.9794887218045113, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 21284 + }, + { + "epoch": 0.9795488721804512, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 21285 + }, + { + "epoch": 0.9796090225563909, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 21286 + }, + { + "epoch": 0.9796691729323308, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 21287 + }, + { + "epoch": 0.9797293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 21288 + }, + { + "epoch": 0.9797894736842105, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 21289 + }, + { + "epoch": 0.9798496240601504, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2307, + "step": 21290 + }, + { + "epoch": 0.9799097744360902, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 21291 + }, + { + "epoch": 0.97996992481203, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 21292 + }, + { + "epoch": 0.9800300751879699, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 21293 + }, + { + "epoch": 0.9800902255639098, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 21294 + }, + { + "epoch": 0.9801503759398497, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 21295 + }, + { + "epoch": 0.9802105263157894, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 21296 + }, + { + "epoch": 0.9802706766917293, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 21297 + }, + { + "epoch": 0.9803308270676692, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 21298 + }, + { + "epoch": 0.980390977443609, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 21299 + }, + { + "epoch": 0.9804511278195489, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 21300 + }, + { + "epoch": 0.9805112781954887, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1843, + "step": 21301 + }, + { + "epoch": 0.9805714285714285, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 21302 + }, + { + "epoch": 0.9806315789473684, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 21303 + }, + { + "epoch": 0.9806917293233083, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 21304 + }, + { + "epoch": 0.9807518796992482, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 21305 + }, + { + "epoch": 0.980812030075188, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 21306 + }, + { + "epoch": 0.9808721804511278, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 21307 + }, + { + "epoch": 0.9809323308270677, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 21308 + }, + { + "epoch": 0.9809924812030075, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 21309 + }, + { + "epoch": 0.9810526315789474, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 21310 + }, + { + "epoch": 0.9811127819548873, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 21311 + }, + { + "epoch": 0.981172932330827, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 21312 + }, + { + "epoch": 0.9812330827067669, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 21313 + }, + { + "epoch": 0.9812932330827068, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 21314 + }, + { + "epoch": 0.9813533834586466, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 21315 + }, + { + "epoch": 0.9814135338345865, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 21316 + }, + { + "epoch": 0.9814736842105263, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 21317 + }, + { + "epoch": 0.9815338345864661, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 21318 + }, + { + "epoch": 0.981593984962406, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1847, + "step": 21319 + }, + { + "epoch": 0.9816541353383459, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 21320 + }, + { + "epoch": 0.9817142857142858, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 21321 + }, + { + "epoch": 0.9817744360902255, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 21322 + }, + { + "epoch": 0.9818345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 21323 + }, + { + "epoch": 0.9818947368421053, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 21324 + }, + { + "epoch": 0.9819548872180451, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 21325 + }, + { + "epoch": 0.982015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 21326 + }, + { + "epoch": 0.9820751879699248, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 21327 + }, + { + "epoch": 0.9821353383458646, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 21328 + }, + { + "epoch": 0.9821954887218045, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.2002, + "step": 21329 + }, + { + "epoch": 0.9822556390977444, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 21330 + }, + { + "epoch": 0.9823157894736843, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 21331 + }, + { + "epoch": 0.982375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 21332 + }, + { + "epoch": 0.9824360902255639, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 21333 + }, + { + "epoch": 0.9824962406015038, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2147, + "step": 21334 + }, + { + "epoch": 0.9825563909774436, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 21335 + }, + { + "epoch": 0.9826165413533835, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 21336 + }, + { + "epoch": 0.9826766917293233, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 21337 + }, + { + "epoch": 0.9827368421052631, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 21338 + }, + { + "epoch": 0.982796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 21339 + }, + { + "epoch": 0.9828571428571429, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 21340 + }, + { + "epoch": 0.9829172932330827, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 21341 + }, + { + "epoch": 0.9829774436090225, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 21342 + }, + { + "epoch": 0.9830375939849624, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 21343 + }, + { + "epoch": 0.9830977443609022, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 21344 + }, + { + "epoch": 0.9831578947368421, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 21345 + }, + { + "epoch": 0.983218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 21346 + }, + { + "epoch": 0.9832781954887218, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 21347 + }, + { + "epoch": 0.9833383458646616, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 21348 + }, + { + "epoch": 0.9833984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 21349 + }, + { + "epoch": 0.9834586466165414, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 21350 + }, + { + "epoch": 0.9835187969924812, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 21351 + }, + { + "epoch": 0.983578947368421, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 21352 + }, + { + "epoch": 0.9836390977443609, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 21353 + }, + { + "epoch": 0.9836992481203007, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 21354 + }, + { + "epoch": 0.9837593984962406, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 21355 + }, + { + "epoch": 0.9838195488721805, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 21356 + }, + { + "epoch": 0.9838796992481204, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 21357 + }, + { + "epoch": 0.9839398496240601, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 21358 + }, + { + "epoch": 0.984, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 21359 + }, + { + "epoch": 0.9840601503759399, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2148, + "step": 21360 + }, + { + "epoch": 0.9841203007518797, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 21361 + }, + { + "epoch": 0.9841804511278196, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 21362 + }, + { + "epoch": 0.9842406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 21363 + }, + { + "epoch": 0.9843007518796992, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 21364 + }, + { + "epoch": 0.9843609022556391, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 21365 + }, + { + "epoch": 0.984421052631579, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 21366 + }, + { + "epoch": 0.9844812030075188, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 21367 + }, + { + "epoch": 0.9845413533834586, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 21368 + }, + { + "epoch": 0.9846015037593985, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1714, + "step": 21369 + }, + { + "epoch": 0.9846616541353383, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1918, + "step": 21370 + }, + { + "epoch": 0.9847218045112782, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 21371 + }, + { + "epoch": 0.9847819548872181, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 21372 + }, + { + "epoch": 0.9848421052631579, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 21373 + }, + { + "epoch": 0.9849022556390977, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 21374 + }, + { + "epoch": 0.9849624060150376, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 21375 + }, + { + "epoch": 0.9850225563909775, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 21376 + }, + { + "epoch": 0.9850827067669173, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 21377 + }, + { + "epoch": 0.9851428571428571, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 21378 + }, + { + "epoch": 0.985203007518797, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 21379 + }, + { + "epoch": 0.9852631578947368, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 21380 + }, + { + "epoch": 0.9853233082706767, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 21381 + }, + { + "epoch": 0.9853834586466166, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 21382 + }, + { + "epoch": 0.9854436090225563, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 21383 + }, + { + "epoch": 0.9855037593984962, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 21384 + }, + { + "epoch": 0.9855639097744361, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 21385 + }, + { + "epoch": 0.985624060150376, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 21386 + }, + { + "epoch": 0.9856842105263158, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 21387 + }, + { + "epoch": 0.9857443609022556, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 21388 + }, + { + "epoch": 0.9858045112781955, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 21389 + }, + { + "epoch": 0.9858646616541353, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 21390 + }, + { + "epoch": 0.9859248120300752, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 21391 + }, + { + "epoch": 0.9859849624060151, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 21392 + }, + { + "epoch": 0.9860451127819548, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 21393 + }, + { + "epoch": 0.9861052631578947, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 21394 + }, + { + "epoch": 0.9861654135338346, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 21395 + }, + { + "epoch": 0.9862255639097744, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1465, + "step": 21396 + }, + { + "epoch": 0.9862857142857143, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 21397 + }, + { + "epoch": 0.9863458646616541, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 21398 + }, + { + "epoch": 0.986406015037594, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 21399 + }, + { + "epoch": 0.9864661654135338, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 21400 + }, + { + "epoch": 0.9865263157894737, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 21401 + }, + { + "epoch": 0.9865864661654136, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 21402 + }, + { + "epoch": 0.9866466165413534, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 21403 + }, + { + "epoch": 0.9867067669172932, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 21404 + }, + { + "epoch": 0.9867669172932331, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 21405 + }, + { + "epoch": 0.9868270676691729, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 21406 + }, + { + "epoch": 0.9868872180451128, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 21407 + }, + { + "epoch": 0.9869473684210527, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 21408 + }, + { + "epoch": 0.9870075187969924, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 21409 + }, + { + "epoch": 0.9870676691729323, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 21410 + }, + { + "epoch": 0.9871278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 21411 + }, + { + "epoch": 0.9871879699248121, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 21412 + }, + { + "epoch": 0.9872481203007519, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 21413 + }, + { + "epoch": 0.9873082706766917, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 21414 + }, + { + "epoch": 0.9873684210526316, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 21415 + }, + { + "epoch": 0.9874285714285714, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 21416 + }, + { + "epoch": 0.9874887218045113, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 21417 + }, + { + "epoch": 0.9875488721804512, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 21418 + }, + { + "epoch": 0.9876090225563909, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 21419 + }, + { + "epoch": 0.9876691729323308, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 21420 + }, + { + "epoch": 0.9877293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 21421 + }, + { + "epoch": 0.9877894736842106, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 21422 + }, + { + "epoch": 0.9878496240601504, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 21423 + }, + { + "epoch": 0.9879097744360902, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 21424 + }, + { + "epoch": 0.98796992481203, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 21425 + }, + { + "epoch": 0.9880300751879699, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1816, + "step": 21426 + }, + { + "epoch": 0.9880902255639098, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 21427 + }, + { + "epoch": 0.9881503759398497, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 21428 + }, + { + "epoch": 0.9882105263157894, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 21429 + }, + { + "epoch": 0.9882706766917293, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 21430 + }, + { + "epoch": 0.9883308270676692, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 21431 + }, + { + "epoch": 0.988390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 21432 + }, + { + "epoch": 0.9884511278195489, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 21433 + }, + { + "epoch": 0.9885112781954887, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 21434 + }, + { + "epoch": 0.9885714285714285, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 21435 + }, + { + "epoch": 0.9886315789473684, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 21436 + }, + { + "epoch": 0.9886917293233083, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 21437 + }, + { + "epoch": 0.9887518796992482, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 21438 + }, + { + "epoch": 0.9888120300751879, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 21439 + }, + { + "epoch": 0.9888721804511278, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 21440 + }, + { + "epoch": 0.9889323308270677, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2029, + "step": 21441 + }, + { + "epoch": 0.9889924812030075, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 21442 + }, + { + "epoch": 0.9890526315789474, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 21443 + }, + { + "epoch": 0.9891127819548872, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 21444 + }, + { + "epoch": 0.989172932330827, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 21445 + }, + { + "epoch": 0.9892330827067669, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 21446 + }, + { + "epoch": 0.9892932330827068, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 21447 + }, + { + "epoch": 0.9893533834586467, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 21448 + }, + { + "epoch": 0.9894135338345864, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1928, + "step": 21449 + }, + { + "epoch": 0.9894736842105263, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 21450 + }, + { + "epoch": 0.9895338345864662, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 21451 + }, + { + "epoch": 0.989593984962406, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 21452 + }, + { + "epoch": 0.9896541353383459, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 21453 + }, + { + "epoch": 0.9897142857142858, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 21454 + }, + { + "epoch": 0.9897744360902255, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 21455 + }, + { + "epoch": 0.9898345864661654, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 21456 + }, + { + "epoch": 0.9898947368421053, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1912, + "step": 21457 + }, + { + "epoch": 0.9899548872180451, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 21458 + }, + { + "epoch": 0.990015037593985, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 21459 + }, + { + "epoch": 0.9900751879699248, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 21460 + }, + { + "epoch": 0.9901353383458646, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 21461 + }, + { + "epoch": 0.9901954887218045, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 21462 + }, + { + "epoch": 0.9902556390977444, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 21463 + }, + { + "epoch": 0.9903157894736843, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 21464 + }, + { + "epoch": 0.990375939849624, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 21465 + }, + { + "epoch": 0.9904360902255639, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 21466 + }, + { + "epoch": 0.9904962406015038, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 21467 + }, + { + "epoch": 0.9905563909774436, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.16, + "step": 21468 + }, + { + "epoch": 0.9906165413533835, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 21469 + }, + { + "epoch": 0.9906766917293233, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 21470 + }, + { + "epoch": 0.9907368421052631, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 21471 + }, + { + "epoch": 0.990796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 21472 + }, + { + "epoch": 0.9908571428571429, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 21473 + }, + { + "epoch": 0.9909172932330828, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 21474 + }, + { + "epoch": 0.9909774436090225, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 21475 + }, + { + "epoch": 0.9910375939849624, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2734, + "step": 21476 + }, + { + "epoch": 0.9910977443609023, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 21477 + }, + { + "epoch": 0.9911578947368421, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2897, + "step": 21478 + }, + { + "epoch": 0.991218045112782, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 21479 + }, + { + "epoch": 0.9912781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 21480 + }, + { + "epoch": 0.9913383458646616, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 21481 + }, + { + "epoch": 0.9913984962406015, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 21482 + }, + { + "epoch": 0.9914586466165414, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1339, + "step": 21483 + }, + { + "epoch": 0.9915187969924812, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 21484 + }, + { + "epoch": 0.991578947368421, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 21485 + }, + { + "epoch": 0.9916390977443609, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 21486 + }, + { + "epoch": 0.9916992481203007, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 21487 + }, + { + "epoch": 0.9917593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 21488 + }, + { + "epoch": 0.9918195488721805, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1567, + "step": 21489 + }, + { + "epoch": 0.9918796992481203, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 21490 + }, + { + "epoch": 0.9919398496240601, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 21491 + }, + { + "epoch": 0.992, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 21492 + }, + { + "epoch": 0.9920601503759399, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 21493 + }, + { + "epoch": 0.9921203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 21494 + }, + { + "epoch": 0.9921804511278195, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 21495 + }, + { + "epoch": 0.9922406015037594, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 21496 + }, + { + "epoch": 0.9923007518796992, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 21497 + }, + { + "epoch": 0.9923609022556391, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 21498 + }, + { + "epoch": 0.992421052631579, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 21499 + }, + { + "epoch": 0.9924812030075187, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 21500 + }, + { + "epoch": 0.9925413533834586, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 21501 + }, + { + "epoch": 0.9926015037593985, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 21502 + }, + { + "epoch": 0.9926616541353384, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2038, + "step": 21503 + }, + { + "epoch": 0.9927218045112782, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 21504 + }, + { + "epoch": 0.9927819548872181, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 21505 + }, + { + "epoch": 0.9928421052631579, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 21506 + }, + { + "epoch": 0.9929022556390977, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1907, + "step": 21507 + }, + { + "epoch": 0.9929624060150376, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 21508 + }, + { + "epoch": 0.9930225563909775, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 21509 + }, + { + "epoch": 0.9930827067669173, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 21510 + }, + { + "epoch": 0.9931428571428571, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 21511 + }, + { + "epoch": 0.993203007518797, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 21512 + }, + { + "epoch": 0.9932631578947368, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2893, + "step": 21513 + }, + { + "epoch": 0.9933233082706767, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 21514 + }, + { + "epoch": 0.9933834586466166, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 21515 + }, + { + "epoch": 0.9934436090225564, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 21516 + }, + { + "epoch": 0.9935037593984962, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 21517 + }, + { + "epoch": 0.9935639097744361, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 21518 + }, + { + "epoch": 0.993624060150376, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 21519 + }, + { + "epoch": 0.9936842105263158, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 21520 + }, + { + "epoch": 0.9937443609022556, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 21521 + }, + { + "epoch": 0.9938045112781955, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 21522 + }, + { + "epoch": 0.9938646616541353, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 21523 + }, + { + "epoch": 0.9939248120300752, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 21524 + }, + { + "epoch": 0.9939849624060151, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 21525 + }, + { + "epoch": 0.9940451127819548, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 21526 + }, + { + "epoch": 0.9941052631578947, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 21527 + }, + { + "epoch": 0.9941654135338346, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 21528 + }, + { + "epoch": 0.9942255639097745, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 21529 + }, + { + "epoch": 0.9942857142857143, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 21530 + }, + { + "epoch": 0.9943458646616541, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 21531 + }, + { + "epoch": 0.994406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 21532 + }, + { + "epoch": 0.9944661654135338, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 21533 + }, + { + "epoch": 0.9945263157894737, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 21534 + }, + { + "epoch": 0.9945864661654136, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 21535 + }, + { + "epoch": 0.9946466165413533, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 21536 + }, + { + "epoch": 0.9947067669172932, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 21537 + }, + { + "epoch": 0.9947669172932331, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 21538 + }, + { + "epoch": 0.994827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 21539 + }, + { + "epoch": 0.9948872180451128, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 21540 + }, + { + "epoch": 0.9949473684210526, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 21541 + }, + { + "epoch": 0.9950075187969925, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 21542 + }, + { + "epoch": 0.9950676691729323, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 21543 + }, + { + "epoch": 0.9951278195488722, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 21544 + }, + { + "epoch": 0.9951879699248121, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1635, + "step": 21545 + }, + { + "epoch": 0.9952481203007518, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 21546 + }, + { + "epoch": 0.9953082706766917, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 21547 + }, + { + "epoch": 0.9953684210526316, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 21548 + }, + { + "epoch": 0.9954285714285714, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 21549 + }, + { + "epoch": 0.9954887218045113, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 21550 + }, + { + "epoch": 0.9955488721804512, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 21551 + }, + { + "epoch": 0.9956090225563909, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 21552 + }, + { + "epoch": 0.9956691729323308, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 21553 + }, + { + "epoch": 0.9957293233082707, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 21554 + }, + { + "epoch": 0.9957894736842106, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 21555 + }, + { + "epoch": 0.9958496240601504, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 21556 + }, + { + "epoch": 0.9959097744360902, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 21557 + }, + { + "epoch": 0.9959699248120301, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 21558 + }, + { + "epoch": 0.9960300751879699, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 21559 + }, + { + "epoch": 0.9960902255639098, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1916, + "step": 21560 + }, + { + "epoch": 0.9961503759398497, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 21561 + }, + { + "epoch": 0.9962105263157894, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 21562 + }, + { + "epoch": 0.9962706766917293, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 21563 + }, + { + "epoch": 0.9963308270676692, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 21564 + }, + { + "epoch": 0.996390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 21565 + }, + { + "epoch": 0.9964511278195489, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1761, + "step": 21566 + }, + { + "epoch": 0.9965112781954887, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 21567 + }, + { + "epoch": 0.9965714285714286, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 21568 + }, + { + "epoch": 0.9966315789473684, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 21569 + }, + { + "epoch": 0.9966917293233083, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 21570 + }, + { + "epoch": 0.9967518796992482, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 21571 + }, + { + "epoch": 0.9968120300751879, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 21572 + }, + { + "epoch": 0.9968721804511278, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 21573 + }, + { + "epoch": 0.9969323308270677, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 21574 + }, + { + "epoch": 0.9969924812030075, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 21575 + }, + { + "epoch": 0.9970526315789474, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 21576 + }, + { + "epoch": 0.9971127819548872, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 21577 + }, + { + "epoch": 0.997172932330827, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 21578 + }, + { + "epoch": 0.9972330827067669, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 21579 + }, + { + "epoch": 0.9972932330827068, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 21580 + }, + { + "epoch": 0.9973533834586467, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 21581 + }, + { + "epoch": 0.9974135338345864, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 21582 + }, + { + "epoch": 0.9974736842105263, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 21583 + }, + { + "epoch": 0.9975338345864662, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1929, + "step": 21584 + }, + { + "epoch": 0.997593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 21585 + }, + { + "epoch": 0.9976541353383459, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 21586 + }, + { + "epoch": 0.9977142857142857, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 21587 + }, + { + "epoch": 0.9977744360902255, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 21588 + }, + { + "epoch": 0.9978345864661654, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 21589 + }, + { + "epoch": 0.9978947368421053, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 21590 + }, + { + "epoch": 0.9979548872180452, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 21591 + }, + { + "epoch": 0.9980150375939849, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 21592 + }, + { + "epoch": 0.9980751879699248, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 21593 + }, + { + "epoch": 0.9981353383458647, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 21594 + }, + { + "epoch": 0.9981954887218045, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 21595 + }, + { + "epoch": 0.9982556390977444, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1643, + "step": 21596 + }, + { + "epoch": 0.9983157894736842, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 21597 + }, + { + "epoch": 0.998375939849624, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 21598 + }, + { + "epoch": 0.9984360902255639, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 21599 + }, + { + "epoch": 0.9984962406015038, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 21600 + }, + { + "epoch": 0.9985563909774436, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 21601 + }, + { + "epoch": 0.9986165413533835, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 21602 + }, + { + "epoch": 0.9986766917293233, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 21603 + }, + { + "epoch": 0.9987368421052631, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 21604 + }, + { + "epoch": 0.998796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 21605 + }, + { + "epoch": 0.9988571428571429, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 21606 + }, + { + "epoch": 0.9989172932330828, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1691, + "step": 21607 + }, + { + "epoch": 0.9989774436090225, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 21608 + }, + { + "epoch": 0.9990375939849624, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 21609 + }, + { + "epoch": 0.9990977443609023, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 21610 + }, + { + "epoch": 0.9991578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 21611 + }, + { + "epoch": 0.999218045112782, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 21612 + }, + { + "epoch": 0.9992781954887218, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 21613 + }, + { + "epoch": 0.9993383458646616, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 21614 + }, + { + "epoch": 0.9993984962406015, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 21615 + }, + { + "epoch": 0.9994586466165414, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1612, + "step": 21616 + }, + { + "epoch": 0.9995187969924813, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 21617 + }, + { + "epoch": 0.999578947368421, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 21618 + }, + { + "epoch": 0.9996390977443609, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 21619 + }, + { + "epoch": 0.9996992481203008, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 21620 + }, + { + "epoch": 0.9997593984962406, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 21621 + }, + { + "epoch": 0.9998195488721805, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 21622 + }, + { + "epoch": 0.9998796992481203, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 21623 + }, + { + "epoch": 0.9999398496240601, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 21624 + }, + { + "epoch": 1.0, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2122, + "step": 21625 + }, + { + "epoch": 1.0000601503759399, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 21626 + }, + { + "epoch": 1.0001203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 21627 + }, + { + "epoch": 1.0001804511278196, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 21628 + }, + { + "epoch": 1.0002406015037595, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 21629 + }, + { + "epoch": 1.0003007518796994, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 21630 + }, + { + "epoch": 1.000360902255639, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 21631 + }, + { + "epoch": 1.0004210526315789, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 21632 + }, + { + "epoch": 1.0004812030075187, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 21633 + }, + { + "epoch": 1.0005413533834586, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 21634 + }, + { + "epoch": 1.0006015037593985, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 21635 + }, + { + "epoch": 1.0006616541353384, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.939, + "step": 21636 + }, + { + "epoch": 1.0007218045112782, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 21637 + }, + { + "epoch": 1.000781954887218, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 21638 + }, + { + "epoch": 1.000842105263158, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 21639 + }, + { + "epoch": 1.0009022556390978, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 21640 + }, + { + "epoch": 1.0009624060150375, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2484, + "step": 21641 + }, + { + "epoch": 1.0010225563909774, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 21642 + }, + { + "epoch": 1.0010827067669172, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 21643 + }, + { + "epoch": 1.0011428571428571, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 21644 + }, + { + "epoch": 1.001203007518797, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 21645 + }, + { + "epoch": 1.0012631578947369, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 21646 + }, + { + "epoch": 1.0013233082706767, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 21647 + }, + { + "epoch": 1.0013834586466166, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 21648 + }, + { + "epoch": 1.0014436090225565, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 21649 + }, + { + "epoch": 1.0015037593984963, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 21650 + }, + { + "epoch": 1.001563909774436, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 21651 + }, + { + "epoch": 1.0016240601503759, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 21652 + }, + { + "epoch": 1.0016842105263157, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1652, + "step": 21653 + }, + { + "epoch": 1.0017443609022556, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 21654 + }, + { + "epoch": 1.0018045112781955, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 21655 + }, + { + "epoch": 1.0018646616541353, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 21656 + }, + { + "epoch": 1.0019248120300752, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 21657 + }, + { + "epoch": 1.001984962406015, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 21658 + }, + { + "epoch": 1.002045112781955, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 21659 + }, + { + "epoch": 1.0021052631578948, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 21660 + }, + { + "epoch": 1.0021654135338345, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 21661 + }, + { + "epoch": 1.0022255639097744, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 21662 + }, + { + "epoch": 1.0022857142857142, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 21663 + }, + { + "epoch": 1.002345864661654, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 21664 + }, + { + "epoch": 1.002406015037594, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 21665 + }, + { + "epoch": 1.0024661654135338, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 21666 + }, + { + "epoch": 1.0025263157894737, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 21667 + }, + { + "epoch": 1.0025864661654136, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 21668 + }, + { + "epoch": 1.0026466165413535, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 21669 + }, + { + "epoch": 1.0027067669172933, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 21670 + }, + { + "epoch": 1.002766917293233, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 21671 + }, + { + "epoch": 1.0028270676691728, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1911, + "step": 21672 + }, + { + "epoch": 1.0028872180451127, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 21673 + }, + { + "epoch": 1.0029473684210526, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 21674 + }, + { + "epoch": 1.0030075187969925, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 21675 + }, + { + "epoch": 1.0030676691729323, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 21676 + }, + { + "epoch": 1.0031278195488722, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 21677 + }, + { + "epoch": 1.003187969924812, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 21678 + }, + { + "epoch": 1.003248120300752, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 21679 + }, + { + "epoch": 1.0033082706766918, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 21680 + }, + { + "epoch": 1.0033684210526317, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 21681 + }, + { + "epoch": 1.0034285714285713, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 21682 + }, + { + "epoch": 1.0034887218045112, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 21683 + }, + { + "epoch": 1.003548872180451, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 21684 + }, + { + "epoch": 1.003609022556391, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 21685 + }, + { + "epoch": 1.0036691729323308, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 21686 + }, + { + "epoch": 1.0037293233082707, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 21687 + }, + { + "epoch": 1.0037894736842106, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 21688 + }, + { + "epoch": 1.0038496240601504, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 21689 + }, + { + "epoch": 1.0039097744360903, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 21690 + }, + { + "epoch": 1.0039699248120302, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2143, + "step": 21691 + }, + { + "epoch": 1.0040300751879698, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 21692 + }, + { + "epoch": 1.0040902255639097, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 21693 + }, + { + "epoch": 1.0041503759398496, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 21694 + }, + { + "epoch": 1.0042105263157894, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 21695 + }, + { + "epoch": 1.0042706766917293, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 21696 + }, + { + "epoch": 1.0043308270676692, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 21697 + }, + { + "epoch": 1.004390977443609, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.2194, + "step": 21698 + }, + { + "epoch": 1.004451127819549, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 21699 + }, + { + "epoch": 1.0045112781954888, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 21700 + }, + { + "epoch": 1.0045714285714287, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 21701 + }, + { + "epoch": 1.0046315789473683, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2082, + "step": 21702 + }, + { + "epoch": 1.0046917293233082, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 21703 + }, + { + "epoch": 1.004751879699248, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 21704 + }, + { + "epoch": 1.004812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 21705 + }, + { + "epoch": 1.0048721804511278, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 21706 + }, + { + "epoch": 1.0049323308270677, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 21707 + }, + { + "epoch": 1.0049924812030075, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 21708 + }, + { + "epoch": 1.0050526315789474, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 21709 + }, + { + "epoch": 1.0051127819548873, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 21710 + }, + { + "epoch": 1.0051729323308272, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 21711 + }, + { + "epoch": 1.0052330827067668, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 21712 + }, + { + "epoch": 1.0052932330827067, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1414, + "step": 21713 + }, + { + "epoch": 1.0053533834586466, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 21714 + }, + { + "epoch": 1.0054135338345864, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 21715 + }, + { + "epoch": 1.0054736842105263, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 21716 + }, + { + "epoch": 1.0055338345864662, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 21717 + }, + { + "epoch": 1.005593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 21718 + }, + { + "epoch": 1.005654135338346, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1897, + "step": 21719 + }, + { + "epoch": 1.0057142857142858, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 21720 + }, + { + "epoch": 1.0057744360902257, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 21721 + }, + { + "epoch": 1.0058345864661653, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 21722 + }, + { + "epoch": 1.0058947368421052, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 21723 + }, + { + "epoch": 1.005954887218045, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 21724 + }, + { + "epoch": 1.006015037593985, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 21725 + }, + { + "epoch": 1.0060751879699248, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 21726 + }, + { + "epoch": 1.0061353383458647, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 21727 + }, + { + "epoch": 1.0061954887218045, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 21728 + }, + { + "epoch": 1.0062556390977444, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 21729 + }, + { + "epoch": 1.0063157894736843, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 21730 + }, + { + "epoch": 1.0063759398496241, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 21731 + }, + { + "epoch": 1.006436090225564, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 21732 + }, + { + "epoch": 1.0064962406015037, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 21733 + }, + { + "epoch": 1.0065563909774435, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 21734 + }, + { + "epoch": 1.0066165413533834, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 21735 + }, + { + "epoch": 1.0066766917293233, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1647, + "step": 21736 + }, + { + "epoch": 1.0067368421052632, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 21737 + }, + { + "epoch": 1.006796992481203, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 21738 + }, + { + "epoch": 1.006857142857143, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 21739 + }, + { + "epoch": 1.0069172932330828, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1512, + "step": 21740 + }, + { + "epoch": 1.0069774436090226, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 21741 + }, + { + "epoch": 1.0070375939849625, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2315, + "step": 21742 + }, + { + "epoch": 1.0070977443609022, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 21743 + }, + { + "epoch": 1.007157894736842, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 21744 + }, + { + "epoch": 1.007218045112782, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 21745 + }, + { + "epoch": 1.0072781954887218, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 21746 + }, + { + "epoch": 1.0073383458646616, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 21747 + }, + { + "epoch": 1.0073984962406015, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 21748 + }, + { + "epoch": 1.0074586466165414, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 21749 + }, + { + "epoch": 1.0075187969924813, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 21750 + }, + { + "epoch": 1.0075789473684211, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 21751 + }, + { + "epoch": 1.007639097744361, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 21752 + }, + { + "epoch": 1.0076992481203007, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 21753 + }, + { + "epoch": 1.0077593984962405, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 21754 + }, + { + "epoch": 1.0078195488721804, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 21755 + }, + { + "epoch": 1.0078796992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 21756 + }, + { + "epoch": 1.0079398496240601, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 21757 + }, + { + "epoch": 1.008, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 21758 + }, + { + "epoch": 1.0080601503759399, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 21759 + }, + { + "epoch": 1.0081203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 21760 + }, + { + "epoch": 1.0081804511278196, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 21761 + }, + { + "epoch": 1.0082406015037595, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 21762 + }, + { + "epoch": 1.0083007518796991, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 21763 + }, + { + "epoch": 1.008360902255639, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 21764 + }, + { + "epoch": 1.0084210526315789, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 21765 + }, + { + "epoch": 1.0084812030075188, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 21766 + }, + { + "epoch": 1.0085413533834586, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 21767 + }, + { + "epoch": 1.0086015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 21768 + }, + { + "epoch": 1.0086616541353384, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 21769 + }, + { + "epoch": 1.0087218045112782, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 21770 + }, + { + "epoch": 1.0087819548872181, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 21771 + }, + { + "epoch": 1.008842105263158, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 21772 + }, + { + "epoch": 1.0089022556390979, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 21773 + }, + { + "epoch": 1.0089624060150375, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 21774 + }, + { + "epoch": 1.0090225563909774, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 21775 + }, + { + "epoch": 1.0090827067669172, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 21776 + }, + { + "epoch": 1.0091428571428571, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 21777 + }, + { + "epoch": 1.009203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 21778 + }, + { + "epoch": 1.0092631578947369, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 21779 + }, + { + "epoch": 1.0093233082706767, + "grad_norm": 1.609375, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 21780 + }, + { + "epoch": 1.0093834586466166, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 21781 + }, + { + "epoch": 1.0094436090225565, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 21782 + }, + { + "epoch": 1.0095037593984963, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 21783 + }, + { + "epoch": 1.009563909774436, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 21784 + }, + { + "epoch": 1.0096240601503759, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 21785 + }, + { + "epoch": 1.0096842105263157, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 21786 + }, + { + "epoch": 1.0097443609022556, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1884, + "step": 21787 + }, + { + "epoch": 1.0098045112781955, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 21788 + }, + { + "epoch": 1.0098646616541354, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 21789 + }, + { + "epoch": 1.0099248120300752, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 21790 + }, + { + "epoch": 1.009984962406015, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 21791 + }, + { + "epoch": 1.010045112781955, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 21792 + }, + { + "epoch": 1.0101052631578948, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 21793 + }, + { + "epoch": 1.0101654135338345, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 21794 + }, + { + "epoch": 1.0102255639097744, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 21795 + }, + { + "epoch": 1.0102857142857142, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 21796 + }, + { + "epoch": 1.010345864661654, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 21797 + }, + { + "epoch": 1.010406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 21798 + }, + { + "epoch": 1.0104661654135338, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 21799 + }, + { + "epoch": 1.0105263157894737, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 21800 + }, + { + "epoch": 1.0105864661654136, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 21801 + }, + { + "epoch": 1.0106466165413535, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 21802 + }, + { + "epoch": 1.0107067669172933, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 21803 + }, + { + "epoch": 1.010766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 21804 + }, + { + "epoch": 1.0108270676691729, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 21805 + }, + { + "epoch": 1.0108872180451127, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 21806 + }, + { + "epoch": 1.0109473684210526, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 21807 + }, + { + "epoch": 1.0110075187969925, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 21808 + }, + { + "epoch": 1.0110676691729323, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 21809 + }, + { + "epoch": 1.0111278195488722, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 21810 + }, + { + "epoch": 1.011187969924812, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 21811 + }, + { + "epoch": 1.011248120300752, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 21812 + }, + { + "epoch": 1.0113082706766918, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 21813 + }, + { + "epoch": 1.0113684210526315, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 21814 + }, + { + "epoch": 1.0114285714285713, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 21815 + }, + { + "epoch": 1.0114887218045112, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 21816 + }, + { + "epoch": 1.011548872180451, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 21817 + }, + { + "epoch": 1.011609022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 21818 + }, + { + "epoch": 1.0116691729323308, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 21819 + }, + { + "epoch": 1.0117293233082707, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 21820 + }, + { + "epoch": 1.0117894736842106, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 21821 + }, + { + "epoch": 1.0118496240601504, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 21822 + }, + { + "epoch": 1.0119097744360903, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.19, + "step": 21823 + }, + { + "epoch": 1.0119699248120302, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 21824 + }, + { + "epoch": 1.0120300751879698, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 21825 + }, + { + "epoch": 1.0120902255639097, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 21826 + }, + { + "epoch": 1.0121503759398496, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 21827 + }, + { + "epoch": 1.0122105263157894, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 21828 + }, + { + "epoch": 1.0122706766917293, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 21829 + }, + { + "epoch": 1.0123308270676692, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 21830 + }, + { + "epoch": 1.012390977443609, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 21831 + }, + { + "epoch": 1.012451127819549, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 21832 + }, + { + "epoch": 1.0125112781954888, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 21833 + }, + { + "epoch": 1.0125714285714287, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 21834 + }, + { + "epoch": 1.0126315789473683, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 21835 + }, + { + "epoch": 1.0126917293233082, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 21836 + }, + { + "epoch": 1.012751879699248, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.182, + "step": 21837 + }, + { + "epoch": 1.012812030075188, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 21838 + }, + { + "epoch": 1.0128721804511278, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1971, + "step": 21839 + }, + { + "epoch": 1.0129323308270677, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 21840 + }, + { + "epoch": 1.0129924812030076, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2214, + "step": 21841 + }, + { + "epoch": 1.0130526315789474, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 21842 + }, + { + "epoch": 1.0131127819548873, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 21843 + }, + { + "epoch": 1.0131729323308272, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 21844 + }, + { + "epoch": 1.0132330827067668, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 21845 + }, + { + "epoch": 1.0132932330827067, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 21846 + }, + { + "epoch": 1.0133533834586466, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 21847 + }, + { + "epoch": 1.0134135338345864, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 21848 + }, + { + "epoch": 1.0134736842105263, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 21849 + }, + { + "epoch": 1.0135338345864662, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 21850 + }, + { + "epoch": 1.013593984962406, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 21851 + }, + { + "epoch": 1.013654135338346, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 21852 + }, + { + "epoch": 1.0137142857142858, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 21853 + }, + { + "epoch": 1.0137744360902257, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 21854 + }, + { + "epoch": 1.0138345864661653, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 21855 + }, + { + "epoch": 1.0138947368421052, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 21856 + }, + { + "epoch": 1.013954887218045, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 21857 + }, + { + "epoch": 1.014015037593985, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 21858 + }, + { + "epoch": 1.0140751879699248, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 21859 + }, + { + "epoch": 1.0141353383458647, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 21860 + }, + { + "epoch": 1.0141954887218045, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 21861 + }, + { + "epoch": 1.0142556390977444, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 21862 + }, + { + "epoch": 1.0143157894736843, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1177, + "step": 21863 + }, + { + "epoch": 1.0143759398496242, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 21864 + }, + { + "epoch": 1.0144360902255638, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1825, + "step": 21865 + }, + { + "epoch": 1.0144962406015037, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 21866 + }, + { + "epoch": 1.0145563909774435, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 21867 + }, + { + "epoch": 1.0146165413533834, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 21868 + }, + { + "epoch": 1.0146766917293233, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 21869 + }, + { + "epoch": 1.0147368421052632, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 21870 + }, + { + "epoch": 1.014796992481203, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 21871 + }, + { + "epoch": 1.014857142857143, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 21872 + }, + { + "epoch": 1.0149172932330828, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 21873 + }, + { + "epoch": 1.0149774436090226, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 21874 + }, + { + "epoch": 1.0150375939849625, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 21875 + }, + { + "epoch": 1.0150977443609022, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 21876 + }, + { + "epoch": 1.015157894736842, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 21877 + }, + { + "epoch": 1.015218045112782, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 21878 + }, + { + "epoch": 1.0152781954887218, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 21879 + }, + { + "epoch": 1.0153383458646617, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 21880 + }, + { + "epoch": 1.0153984962406015, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 21881 + }, + { + "epoch": 1.0154586466165414, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 21882 + }, + { + "epoch": 1.0155187969924813, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 21883 + }, + { + "epoch": 1.0155789473684211, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 21884 + }, + { + "epoch": 1.015639097744361, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 21885 + }, + { + "epoch": 1.0156992481203007, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 21886 + }, + { + "epoch": 1.0157593984962405, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 21887 + }, + { + "epoch": 1.0158195488721804, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 21888 + }, + { + "epoch": 1.0158796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 21889 + }, + { + "epoch": 1.0159398496240601, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 21890 + }, + { + "epoch": 1.016, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 21891 + }, + { + "epoch": 1.0160601503759399, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.2014, + "step": 21892 + }, + { + "epoch": 1.0161203007518798, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 21893 + }, + { + "epoch": 1.0161804511278196, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 21894 + }, + { + "epoch": 1.0162406015037595, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 21895 + }, + { + "epoch": 1.0163007518796991, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 21896 + }, + { + "epoch": 1.016360902255639, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 21897 + }, + { + "epoch": 1.016421052631579, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 21898 + }, + { + "epoch": 1.0164812030075188, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 21899 + }, + { + "epoch": 1.0165413533834586, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 21900 + }, + { + "epoch": 1.0166015037593985, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 21901 + }, + { + "epoch": 1.0166616541353384, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 21902 + }, + { + "epoch": 1.0167218045112782, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 21903 + }, + { + "epoch": 1.0167819548872181, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 21904 + }, + { + "epoch": 1.016842105263158, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 21905 + }, + { + "epoch": 1.0169022556390976, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 21906 + }, + { + "epoch": 1.0169624060150375, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 21907 + }, + { + "epoch": 1.0170225563909774, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 21908 + }, + { + "epoch": 1.0170827067669173, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 21909 + }, + { + "epoch": 1.0171428571428571, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 21910 + }, + { + "epoch": 1.017203007518797, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 21911 + }, + { + "epoch": 1.0172631578947369, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 21912 + }, + { + "epoch": 1.0173233082706767, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 21913 + }, + { + "epoch": 1.0173834586466166, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 21914 + }, + { + "epoch": 1.0174436090225565, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 21915 + }, + { + "epoch": 1.0175037593984961, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 21916 + }, + { + "epoch": 1.017563909774436, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 21917 + }, + { + "epoch": 1.0176240601503759, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 21918 + }, + { + "epoch": 1.0176842105263157, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 21919 + }, + { + "epoch": 1.0177443609022556, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 21920 + }, + { + "epoch": 1.0178045112781955, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 21921 + }, + { + "epoch": 1.0178646616541354, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 21922 + }, + { + "epoch": 1.0179248120300752, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 21923 + }, + { + "epoch": 1.017984962406015, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 21924 + }, + { + "epoch": 1.018045112781955, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 21925 + }, + { + "epoch": 1.0181052631578948, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 21926 + }, + { + "epoch": 1.0181654135338345, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 21927 + }, + { + "epoch": 1.0182255639097744, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 21928 + }, + { + "epoch": 1.0182857142857142, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 21929 + }, + { + "epoch": 1.018345864661654, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 21930 + }, + { + "epoch": 1.018406015037594, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 21931 + }, + { + "epoch": 1.0184661654135339, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 21932 + }, + { + "epoch": 1.0185263157894737, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 21933 + }, + { + "epoch": 1.0185864661654136, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 21934 + }, + { + "epoch": 1.0186466165413535, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 21935 + }, + { + "epoch": 1.0187067669172933, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 21936 + }, + { + "epoch": 1.018766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 21937 + }, + { + "epoch": 1.0188270676691729, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 21938 + }, + { + "epoch": 1.0188872180451127, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 21939 + }, + { + "epoch": 1.0189473684210526, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 21940 + }, + { + "epoch": 1.0190075187969925, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 21941 + }, + { + "epoch": 1.0190676691729323, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 21942 + }, + { + "epoch": 1.0191278195488722, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 21943 + }, + { + "epoch": 1.019187969924812, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 21944 + }, + { + "epoch": 1.019248120300752, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 21945 + }, + { + "epoch": 1.0193082706766918, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 21946 + }, + { + "epoch": 1.0193684210526315, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 21947 + }, + { + "epoch": 1.0194285714285714, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 21948 + }, + { + "epoch": 1.0194887218045112, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 21949 + }, + { + "epoch": 1.019548872180451, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2067, + "step": 21950 + }, + { + "epoch": 1.019609022556391, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 21951 + }, + { + "epoch": 1.0196691729323308, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 21952 + }, + { + "epoch": 1.0197293233082707, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 21953 + }, + { + "epoch": 1.0197894736842106, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 21954 + }, + { + "epoch": 1.0198496240601505, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 21955 + }, + { + "epoch": 1.0199097744360903, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 21956 + }, + { + "epoch": 1.01996992481203, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 21957 + }, + { + "epoch": 1.0200300751879698, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 21958 + }, + { + "epoch": 1.0200902255639097, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 21959 + }, + { + "epoch": 1.0201503759398496, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 21960 + }, + { + "epoch": 1.0202105263157895, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 21961 + }, + { + "epoch": 1.0202706766917293, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 21962 + }, + { + "epoch": 1.0203308270676692, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 21963 + }, + { + "epoch": 1.020390977443609, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1482, + "step": 21964 + }, + { + "epoch": 1.020451127819549, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 21965 + }, + { + "epoch": 1.0205112781954888, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 21966 + }, + { + "epoch": 1.0205714285714285, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 21967 + }, + { + "epoch": 1.0206315789473683, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 21968 + }, + { + "epoch": 1.0206917293233082, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 21969 + }, + { + "epoch": 1.020751879699248, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 21970 + }, + { + "epoch": 1.020812030075188, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 21971 + }, + { + "epoch": 1.0208721804511278, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 21972 + }, + { + "epoch": 1.0209323308270677, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 21973 + }, + { + "epoch": 1.0209924812030076, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 21974 + }, + { + "epoch": 1.0210526315789474, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 21975 + }, + { + "epoch": 1.0211127819548873, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 21976 + }, + { + "epoch": 1.0211729323308272, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 21977 + }, + { + "epoch": 1.0212330827067668, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1695, + "step": 21978 + }, + { + "epoch": 1.0212932330827067, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 21979 + }, + { + "epoch": 1.0213533834586466, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 21980 + }, + { + "epoch": 1.0214135338345864, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 21981 + }, + { + "epoch": 1.0214736842105263, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 21982 + }, + { + "epoch": 1.0215338345864662, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 21983 + }, + { + "epoch": 1.021593984962406, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 21984 + }, + { + "epoch": 1.021654135338346, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 21985 + }, + { + "epoch": 1.0217142857142858, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 21986 + }, + { + "epoch": 1.0217744360902257, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 21987 + }, + { + "epoch": 1.0218345864661653, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 21988 + }, + { + "epoch": 1.0218947368421052, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 21989 + }, + { + "epoch": 1.021954887218045, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 21990 + }, + { + "epoch": 1.022015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 21991 + }, + { + "epoch": 1.0220751879699248, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 21992 + }, + { + "epoch": 1.0221353383458647, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 21993 + }, + { + "epoch": 1.0221954887218045, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 21994 + }, + { + "epoch": 1.0222556390977444, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 21995 + }, + { + "epoch": 1.0223157894736843, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 21996 + }, + { + "epoch": 1.0223759398496242, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 21997 + }, + { + "epoch": 1.0224360902255638, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 21998 + }, + { + "epoch": 1.0224962406015037, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 21999 + }, + { + "epoch": 1.0225563909774436, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 22000 + }, + { + "epoch": 1.0226165413533834, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 22001 + }, + { + "epoch": 1.0226766917293233, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 22002 + }, + { + "epoch": 1.0227368421052632, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 22003 + }, + { + "epoch": 1.022796992481203, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 22004 + }, + { + "epoch": 1.022857142857143, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 22005 + }, + { + "epoch": 1.0229172932330828, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 22006 + }, + { + "epoch": 1.0229774436090227, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 22007 + }, + { + "epoch": 1.0230375939849623, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 22008 + }, + { + "epoch": 1.0230977443609022, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 22009 + }, + { + "epoch": 1.023157894736842, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1767, + "step": 22010 + }, + { + "epoch": 1.023218045112782, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 22011 + }, + { + "epoch": 1.0232781954887218, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 22012 + }, + { + "epoch": 1.0233383458646617, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 22013 + }, + { + "epoch": 1.0233984962406015, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 22014 + }, + { + "epoch": 1.0234586466165414, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 22015 + }, + { + "epoch": 1.0235187969924813, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 22016 + }, + { + "epoch": 1.0235789473684211, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 22017 + }, + { + "epoch": 1.0236390977443608, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 22018 + }, + { + "epoch": 1.0236992481203007, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 22019 + }, + { + "epoch": 1.0237593984962405, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 22020 + }, + { + "epoch": 1.0238195488721804, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1884, + "step": 22021 + }, + { + "epoch": 1.0238796992481203, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 22022 + }, + { + "epoch": 1.0239398496240601, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 22023 + }, + { + "epoch": 1.024, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 22024 + }, + { + "epoch": 1.02406015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 22025 + }, + { + "epoch": 1.0241203007518798, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 22026 + }, + { + "epoch": 1.0241804511278196, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 22027 + }, + { + "epoch": 1.0242406015037595, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 22028 + }, + { + "epoch": 1.0243007518796992, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 22029 + }, + { + "epoch": 1.024360902255639, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 22030 + }, + { + "epoch": 1.024421052631579, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 22031 + }, + { + "epoch": 1.0244812030075188, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 22032 + }, + { + "epoch": 1.0245413533834586, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 22033 + }, + { + "epoch": 1.0246015037593985, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 22034 + }, + { + "epoch": 1.0246616541353384, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 22035 + }, + { + "epoch": 1.0247218045112783, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 22036 + }, + { + "epoch": 1.0247819548872181, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 22037 + }, + { + "epoch": 1.024842105263158, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 22038 + }, + { + "epoch": 1.0249022556390976, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 22039 + }, + { + "epoch": 1.0249624060150375, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2249, + "step": 22040 + }, + { + "epoch": 1.0250225563909774, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 22041 + }, + { + "epoch": 1.0250827067669173, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 22042 + }, + { + "epoch": 1.0251428571428571, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 22043 + }, + { + "epoch": 1.025203007518797, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 22044 + }, + { + "epoch": 1.0252631578947369, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 22045 + }, + { + "epoch": 1.0253233082706767, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 22046 + }, + { + "epoch": 1.0253834586466166, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 22047 + }, + { + "epoch": 1.0254436090225565, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 22048 + }, + { + "epoch": 1.0255037593984961, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 22049 + }, + { + "epoch": 1.025563909774436, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 22050 + }, + { + "epoch": 1.0256240601503759, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 22051 + }, + { + "epoch": 1.0256842105263158, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 22052 + }, + { + "epoch": 1.0257443609022556, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 22053 + }, + { + "epoch": 1.0258045112781955, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 22054 + }, + { + "epoch": 1.0258646616541354, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 22055 + }, + { + "epoch": 1.0259248120300752, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 22056 + }, + { + "epoch": 1.025984962406015, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 22057 + }, + { + "epoch": 1.026045112781955, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 22058 + }, + { + "epoch": 1.0261052631578946, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 22059 + }, + { + "epoch": 1.0261654135338345, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 22060 + }, + { + "epoch": 1.0262255639097744, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 22061 + }, + { + "epoch": 1.0262857142857142, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 22062 + }, + { + "epoch": 1.0263458646616541, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 22063 + }, + { + "epoch": 1.026406015037594, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 22064 + }, + { + "epoch": 1.0264661654135339, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 22065 + }, + { + "epoch": 1.0265263157894737, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 22066 + }, + { + "epoch": 1.0265864661654136, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 22067 + }, + { + "epoch": 1.0266466165413535, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 22068 + }, + { + "epoch": 1.0267067669172931, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1662, + "step": 22069 + }, + { + "epoch": 1.026766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 22070 + }, + { + "epoch": 1.0268270676691729, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 22071 + }, + { + "epoch": 1.0268872180451127, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 22072 + }, + { + "epoch": 1.0269473684210526, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 22073 + }, + { + "epoch": 1.0270075187969925, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 22074 + }, + { + "epoch": 1.0270676691729324, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1914, + "step": 22075 + }, + { + "epoch": 1.0271278195488722, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 22076 + }, + { + "epoch": 1.027187969924812, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 22077 + }, + { + "epoch": 1.027248120300752, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 22078 + }, + { + "epoch": 1.0273082706766918, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 22079 + }, + { + "epoch": 1.0273684210526315, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 22080 + }, + { + "epoch": 1.0274285714285714, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 22081 + }, + { + "epoch": 1.0274887218045112, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 22082 + }, + { + "epoch": 1.027548872180451, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 22083 + }, + { + "epoch": 1.027609022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 22084 + }, + { + "epoch": 1.0276691729323308, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 22085 + }, + { + "epoch": 1.0277293233082707, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 22086 + }, + { + "epoch": 1.0277894736842106, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 22087 + }, + { + "epoch": 1.0278496240601505, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 22088 + }, + { + "epoch": 1.0279097744360903, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 22089 + }, + { + "epoch": 1.02796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 22090 + }, + { + "epoch": 1.0280300751879698, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2016, + "step": 22091 + }, + { + "epoch": 1.0280902255639097, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 22092 + }, + { + "epoch": 1.0281503759398496, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 22093 + }, + { + "epoch": 1.0282105263157895, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 22094 + }, + { + "epoch": 1.0282706766917293, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 22095 + }, + { + "epoch": 1.0283308270676692, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 22096 + }, + { + "epoch": 1.028390977443609, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 22097 + }, + { + "epoch": 1.028451127819549, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 22098 + }, + { + "epoch": 1.0285112781954888, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 22099 + }, + { + "epoch": 1.0285714285714285, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 22100 + }, + { + "epoch": 1.0286315789473683, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 22101 + }, + { + "epoch": 1.0286917293233082, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 22102 + }, + { + "epoch": 1.028751879699248, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 22103 + }, + { + "epoch": 1.028812030075188, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 22104 + }, + { + "epoch": 1.0288721804511278, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 22105 + }, + { + "epoch": 1.0289323308270677, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 22106 + }, + { + "epoch": 1.0289924812030076, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 22107 + }, + { + "epoch": 1.0290526315789474, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 22108 + }, + { + "epoch": 1.0291127819548873, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 22109 + }, + { + "epoch": 1.029172932330827, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 22110 + }, + { + "epoch": 1.0292330827067668, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1819, + "step": 22111 + }, + { + "epoch": 1.0292932330827067, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 22112 + }, + { + "epoch": 1.0293533834586466, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 22113 + }, + { + "epoch": 1.0294135338345864, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1776, + "step": 22114 + }, + { + "epoch": 1.0294736842105263, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 22115 + }, + { + "epoch": 1.0295338345864662, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 22116 + }, + { + "epoch": 1.029593984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 22117 + }, + { + "epoch": 1.029654135338346, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 22118 + }, + { + "epoch": 1.0297142857142858, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 22119 + }, + { + "epoch": 1.0297744360902255, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 22120 + }, + { + "epoch": 1.0298345864661653, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 22121 + }, + { + "epoch": 1.0298947368421052, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 22122 + }, + { + "epoch": 1.029954887218045, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 22123 + }, + { + "epoch": 1.030015037593985, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 22124 + }, + { + "epoch": 1.0300751879699248, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 22125 + }, + { + "epoch": 1.0301353383458647, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 22126 + }, + { + "epoch": 1.0301954887218046, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 22127 + }, + { + "epoch": 1.0302556390977444, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 22128 + }, + { + "epoch": 1.0303157894736843, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 22129 + }, + { + "epoch": 1.0303759398496242, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 22130 + }, + { + "epoch": 1.0304360902255638, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 22131 + }, + { + "epoch": 1.0304962406015037, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 22132 + }, + { + "epoch": 1.0305563909774436, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 22133 + }, + { + "epoch": 1.0306165413533834, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 22134 + }, + { + "epoch": 1.0306766917293233, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 22135 + }, + { + "epoch": 1.0307368421052632, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 22136 + }, + { + "epoch": 1.030796992481203, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 22137 + }, + { + "epoch": 1.030857142857143, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 22138 + }, + { + "epoch": 1.0309172932330828, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 22139 + }, + { + "epoch": 1.0309774436090227, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 22140 + }, + { + "epoch": 1.0310375939849623, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 22141 + }, + { + "epoch": 1.0310977443609022, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 22142 + }, + { + "epoch": 1.031157894736842, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 22143 + }, + { + "epoch": 1.031218045112782, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 22144 + }, + { + "epoch": 1.0312781954887218, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2083, + "step": 22145 + }, + { + "epoch": 1.0313383458646617, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 22146 + }, + { + "epoch": 1.0313984962406015, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 22147 + }, + { + "epoch": 1.0314586466165414, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1814, + "step": 22148 + }, + { + "epoch": 1.0315187969924813, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 22149 + }, + { + "epoch": 1.0315789473684212, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 22150 + }, + { + "epoch": 1.0316390977443608, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 22151 + }, + { + "epoch": 1.0316992481203007, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 22152 + }, + { + "epoch": 1.0317593984962405, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 22153 + }, + { + "epoch": 1.0318195488721804, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 22154 + }, + { + "epoch": 1.0318796992481203, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 22155 + }, + { + "epoch": 1.0319398496240602, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 22156 + }, + { + "epoch": 1.032, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 22157 + }, + { + "epoch": 1.03206015037594, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 22158 + }, + { + "epoch": 1.0321203007518798, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 22159 + }, + { + "epoch": 1.0321804511278196, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 22160 + }, + { + "epoch": 1.0322406015037593, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 22161 + }, + { + "epoch": 1.0323007518796992, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 22162 + }, + { + "epoch": 1.032360902255639, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 22163 + }, + { + "epoch": 1.032421052631579, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 22164 + }, + { + "epoch": 1.0324812030075188, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 22165 + }, + { + "epoch": 1.0325413533834586, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2673, + "step": 22166 + }, + { + "epoch": 1.0326015037593985, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 22167 + }, + { + "epoch": 1.0326616541353384, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 22168 + }, + { + "epoch": 1.0327218045112783, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 22169 + }, + { + "epoch": 1.0327819548872181, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 22170 + }, + { + "epoch": 1.0328421052631578, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 22171 + }, + { + "epoch": 1.0329022556390977, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 22172 + }, + { + "epoch": 1.0329624060150375, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 22173 + }, + { + "epoch": 1.0330225563909774, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 22174 + }, + { + "epoch": 1.0330827067669173, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 22175 + }, + { + "epoch": 1.0331428571428571, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 22176 + }, + { + "epoch": 1.033203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 22177 + }, + { + "epoch": 1.0332631578947369, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 22178 + }, + { + "epoch": 1.0333233082706768, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 22179 + }, + { + "epoch": 1.0333834586466166, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 22180 + }, + { + "epoch": 1.0334436090225565, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 22181 + }, + { + "epoch": 1.0335037593984961, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 22182 + }, + { + "epoch": 1.033563909774436, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 22183 + }, + { + "epoch": 1.033624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 22184 + }, + { + "epoch": 1.0336842105263158, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 22185 + }, + { + "epoch": 1.0337443609022556, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 22186 + }, + { + "epoch": 1.0338045112781955, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 22187 + }, + { + "epoch": 1.0338646616541354, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 22188 + }, + { + "epoch": 1.0339248120300752, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 22189 + }, + { + "epoch": 1.0339849624060151, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 22190 + }, + { + "epoch": 1.034045112781955, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1723, + "step": 22191 + }, + { + "epoch": 1.0341052631578946, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 22192 + }, + { + "epoch": 1.0341654135338345, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 22193 + }, + { + "epoch": 1.0342255639097744, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1913, + "step": 22194 + }, + { + "epoch": 1.0342857142857143, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 22195 + }, + { + "epoch": 1.0343458646616541, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 22196 + }, + { + "epoch": 1.034406015037594, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 22197 + }, + { + "epoch": 1.0344661654135339, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 22198 + }, + { + "epoch": 1.0345263157894737, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 22199 + }, + { + "epoch": 1.0345864661654136, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1939, + "step": 22200 + }, + { + "epoch": 1.0346466165413535, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 22201 + }, + { + "epoch": 1.0347067669172931, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 22202 + }, + { + "epoch": 1.034766917293233, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 22203 + }, + { + "epoch": 1.0348270676691729, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 22204 + }, + { + "epoch": 1.0348872180451127, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 22205 + }, + { + "epoch": 1.0349473684210526, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 22206 + }, + { + "epoch": 1.0350075187969925, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 22207 + }, + { + "epoch": 1.0350676691729324, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 22208 + }, + { + "epoch": 1.0351278195488722, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 22209 + }, + { + "epoch": 1.035187969924812, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 22210 + }, + { + "epoch": 1.035248120300752, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 22211 + }, + { + "epoch": 1.0353082706766916, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 22212 + }, + { + "epoch": 1.0353684210526315, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 22213 + }, + { + "epoch": 1.0354285714285714, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 22214 + }, + { + "epoch": 1.0354887218045112, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 22215 + }, + { + "epoch": 1.035548872180451, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 22216 + }, + { + "epoch": 1.035609022556391, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 22217 + }, + { + "epoch": 1.0356691729323309, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 22218 + }, + { + "epoch": 1.0357293233082707, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 22219 + }, + { + "epoch": 1.0357894736842106, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 22220 + }, + { + "epoch": 1.0358496240601505, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 22221 + }, + { + "epoch": 1.0359097744360901, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 22222 + }, + { + "epoch": 1.03596992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 22223 + }, + { + "epoch": 1.0360300751879699, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 22224 + }, + { + "epoch": 1.0360902255639097, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 22225 + }, + { + "epoch": 1.0361503759398496, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1883, + "step": 22226 + }, + { + "epoch": 1.0362105263157895, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 22227 + }, + { + "epoch": 1.0362706766917293, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 22228 + }, + { + "epoch": 1.0363308270676692, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 22229 + }, + { + "epoch": 1.036390977443609, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 22230 + }, + { + "epoch": 1.036451127819549, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 22231 + }, + { + "epoch": 1.0365112781954888, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1768, + "step": 22232 + }, + { + "epoch": 1.0365714285714285, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 22233 + }, + { + "epoch": 1.0366315789473683, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 22234 + }, + { + "epoch": 1.0366917293233082, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 22235 + }, + { + "epoch": 1.036751879699248, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 22236 + }, + { + "epoch": 1.036812030075188, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 22237 + }, + { + "epoch": 1.0368721804511278, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 22238 + }, + { + "epoch": 1.0369323308270677, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 22239 + }, + { + "epoch": 1.0369924812030076, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 22240 + }, + { + "epoch": 1.0370526315789474, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1805, + "step": 22241 + }, + { + "epoch": 1.0371127819548873, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 22242 + }, + { + "epoch": 1.037172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 22243 + }, + { + "epoch": 1.0372330827067668, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 22244 + }, + { + "epoch": 1.0372932330827067, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 22245 + }, + { + "epoch": 1.0373533834586466, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1823, + "step": 22246 + }, + { + "epoch": 1.0374135338345865, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 22247 + }, + { + "epoch": 1.0374736842105263, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 22248 + }, + { + "epoch": 1.0375338345864662, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 22249 + }, + { + "epoch": 1.037593984962406, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2648, + "step": 22250 + }, + { + "epoch": 1.037654135338346, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 22251 + }, + { + "epoch": 1.0377142857142858, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 22252 + }, + { + "epoch": 1.0377744360902255, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 22253 + }, + { + "epoch": 1.0378345864661653, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 22254 + }, + { + "epoch": 1.0378947368421052, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 22255 + }, + { + "epoch": 1.037954887218045, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 22256 + }, + { + "epoch": 1.038015037593985, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 22257 + }, + { + "epoch": 1.0380751879699248, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 22258 + }, + { + "epoch": 1.0381353383458647, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 22259 + }, + { + "epoch": 1.0381954887218046, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 22260 + }, + { + "epoch": 1.0382556390977444, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 22261 + }, + { + "epoch": 1.0383157894736843, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 22262 + }, + { + "epoch": 1.0383759398496242, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 22263 + }, + { + "epoch": 1.0384360902255638, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 22264 + }, + { + "epoch": 1.0384962406015037, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 22265 + }, + { + "epoch": 1.0385563909774436, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 22266 + }, + { + "epoch": 1.0386165413533834, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 22267 + }, + { + "epoch": 1.0386766917293233, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 22268 + }, + { + "epoch": 1.0387368421052632, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 22269 + }, + { + "epoch": 1.038796992481203, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 22270 + }, + { + "epoch": 1.038857142857143, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 22271 + }, + { + "epoch": 1.0389172932330828, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 22272 + }, + { + "epoch": 1.0389774436090224, + "grad_norm": 0.4609375, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 22273 + }, + { + "epoch": 1.0390375939849623, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 22274 + }, + { + "epoch": 1.0390977443609022, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 22275 + }, + { + "epoch": 1.039157894736842, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 22276 + }, + { + "epoch": 1.039218045112782, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 22277 + }, + { + "epoch": 1.0392781954887218, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 22278 + }, + { + "epoch": 1.0393383458646617, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 22279 + }, + { + "epoch": 1.0393984962406015, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 22280 + }, + { + "epoch": 1.0394586466165414, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 22281 + }, + { + "epoch": 1.0395187969924813, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 22282 + }, + { + "epoch": 1.0395789473684212, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 22283 + }, + { + "epoch": 1.0396390977443608, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 22284 + }, + { + "epoch": 1.0396992481203007, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 22285 + }, + { + "epoch": 1.0397593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 22286 + }, + { + "epoch": 1.0398195488721804, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 22287 + }, + { + "epoch": 1.0398796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 22288 + }, + { + "epoch": 1.0399398496240602, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 22289 + }, + { + "epoch": 1.04, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 22290 + }, + { + "epoch": 1.04006015037594, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.214, + "step": 22291 + }, + { + "epoch": 1.0401203007518798, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 22292 + }, + { + "epoch": 1.0401804511278196, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 22293 + }, + { + "epoch": 1.0402406015037593, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 22294 + }, + { + "epoch": 1.0403007518796992, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 22295 + }, + { + "epoch": 1.040360902255639, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 22296 + }, + { + "epoch": 1.040421052631579, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 22297 + }, + { + "epoch": 1.0404812030075188, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 1.203, + "step": 22298 + }, + { + "epoch": 1.0405413533834587, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 22299 + }, + { + "epoch": 1.0406015037593985, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 22300 + }, + { + "epoch": 1.0406616541353384, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 22301 + }, + { + "epoch": 1.0407218045112783, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 22302 + }, + { + "epoch": 1.0407819548872181, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 22303 + }, + { + "epoch": 1.0408421052631578, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 22304 + }, + { + "epoch": 1.0409022556390977, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 22305 + }, + { + "epoch": 1.0409624060150375, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 22306 + }, + { + "epoch": 1.0410225563909774, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 22307 + }, + { + "epoch": 1.0410827067669173, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 22308 + }, + { + "epoch": 1.0411428571428571, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 22309 + }, + { + "epoch": 1.041203007518797, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 22310 + }, + { + "epoch": 1.041263157894737, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 22311 + }, + { + "epoch": 1.0413233082706768, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 22312 + }, + { + "epoch": 1.0413834586466166, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 22313 + }, + { + "epoch": 1.0414436090225565, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 22314 + }, + { + "epoch": 1.0415037593984962, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 22315 + }, + { + "epoch": 1.041563909774436, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 22316 + }, + { + "epoch": 1.041624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 22317 + }, + { + "epoch": 1.0416842105263158, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.146, + "step": 22318 + }, + { + "epoch": 1.0417443609022556, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 22319 + }, + { + "epoch": 1.0418045112781955, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2191, + "step": 22320 + }, + { + "epoch": 1.0418646616541354, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 22321 + }, + { + "epoch": 1.0419248120300753, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 22322 + }, + { + "epoch": 1.0419849624060151, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 22323 + }, + { + "epoch": 1.0420451127819548, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 22324 + }, + { + "epoch": 1.0421052631578946, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 22325 + }, + { + "epoch": 1.0421654135338345, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 22326 + }, + { + "epoch": 1.0422255639097744, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 22327 + }, + { + "epoch": 1.0422857142857143, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 22328 + }, + { + "epoch": 1.0423458646616541, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 22329 + }, + { + "epoch": 1.042406015037594, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 22330 + }, + { + "epoch": 1.0424661654135339, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 22331 + }, + { + "epoch": 1.0425263157894737, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 22332 + }, + { + "epoch": 1.0425864661654136, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 22333 + }, + { + "epoch": 1.0426466165413535, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 22334 + }, + { + "epoch": 1.0427067669172931, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 22335 + }, + { + "epoch": 1.042766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 22336 + }, + { + "epoch": 1.0428270676691729, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 22337 + }, + { + "epoch": 1.0428872180451128, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 22338 + }, + { + "epoch": 1.0429473684210526, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 22339 + }, + { + "epoch": 1.0430075187969925, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 22340 + }, + { + "epoch": 1.0430676691729324, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 22341 + }, + { + "epoch": 1.0431278195488722, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 22342 + }, + { + "epoch": 1.043187969924812, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 22343 + }, + { + "epoch": 1.043248120300752, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 22344 + }, + { + "epoch": 1.0433082706766916, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 22345 + }, + { + "epoch": 1.0433684210526315, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 22346 + }, + { + "epoch": 1.0434285714285714, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 22347 + }, + { + "epoch": 1.0434887218045112, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 22348 + }, + { + "epoch": 1.0435488721804511, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 22349 + }, + { + "epoch": 1.043609022556391, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 22350 + }, + { + "epoch": 1.0436691729323309, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 22351 + }, + { + "epoch": 1.0437293233082707, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 22352 + }, + { + "epoch": 1.0437894736842106, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 22353 + }, + { + "epoch": 1.0438496240601505, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 22354 + }, + { + "epoch": 1.0439097744360901, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 22355 + }, + { + "epoch": 1.04396992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 22356 + }, + { + "epoch": 1.0440300751879699, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 22357 + }, + { + "epoch": 1.0440902255639097, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 22358 + }, + { + "epoch": 1.0441503759398496, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 22359 + }, + { + "epoch": 1.0442105263157895, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 22360 + }, + { + "epoch": 1.0442706766917293, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 22361 + }, + { + "epoch": 1.0443308270676692, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 22362 + }, + { + "epoch": 1.044390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 22363 + }, + { + "epoch": 1.044451127819549, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1446, + "step": 22364 + }, + { + "epoch": 1.0445112781954888, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 22365 + }, + { + "epoch": 1.0445714285714285, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 22366 + }, + { + "epoch": 1.0446315789473684, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 22367 + }, + { + "epoch": 1.0446917293233082, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 22368 + }, + { + "epoch": 1.044751879699248, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 22369 + }, + { + "epoch": 1.044812030075188, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 22370 + }, + { + "epoch": 1.0448721804511278, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 22371 + }, + { + "epoch": 1.0449323308270677, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 22372 + }, + { + "epoch": 1.0449924812030076, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 22373 + }, + { + "epoch": 1.0450526315789475, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 22374 + }, + { + "epoch": 1.045112781954887, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 22375 + }, + { + "epoch": 1.045172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 22376 + }, + { + "epoch": 1.0452330827067668, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 22377 + }, + { + "epoch": 1.0452932330827067, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 22378 + }, + { + "epoch": 1.0453533834586466, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.3115, + "step": 22379 + }, + { + "epoch": 1.0454135338345865, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 22380 + }, + { + "epoch": 1.0454736842105263, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 22381 + }, + { + "epoch": 1.0455338345864662, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 22382 + }, + { + "epoch": 1.045593984962406, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 22383 + }, + { + "epoch": 1.045654135338346, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 22384 + }, + { + "epoch": 1.0457142857142858, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 22385 + }, + { + "epoch": 1.0457744360902255, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 22386 + }, + { + "epoch": 1.0458345864661653, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 22387 + }, + { + "epoch": 1.0458947368421052, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 22388 + }, + { + "epoch": 1.045954887218045, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 22389 + }, + { + "epoch": 1.046015037593985, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 22390 + }, + { + "epoch": 1.0460751879699248, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 22391 + }, + { + "epoch": 1.0461353383458647, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 22392 + }, + { + "epoch": 1.0461954887218046, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 22393 + }, + { + "epoch": 1.0462556390977444, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 22394 + }, + { + "epoch": 1.0463157894736843, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 22395 + }, + { + "epoch": 1.046375939849624, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 22396 + }, + { + "epoch": 1.0464360902255638, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 22397 + }, + { + "epoch": 1.0464962406015037, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 22398 + }, + { + "epoch": 1.0465563909774436, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 22399 + }, + { + "epoch": 1.0466165413533834, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 22400 + }, + { + "epoch": 1.0466766917293233, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 22401 + }, + { + "epoch": 1.0467368421052632, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 22402 + }, + { + "epoch": 1.046796992481203, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 22403 + }, + { + "epoch": 1.046857142857143, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 22404 + }, + { + "epoch": 1.0469172932330828, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 22405 + }, + { + "epoch": 1.0469774436090225, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 22406 + }, + { + "epoch": 1.0470375939849623, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 22407 + }, + { + "epoch": 1.0470977443609022, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 22408 + }, + { + "epoch": 1.047157894736842, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 22409 + }, + { + "epoch": 1.047218045112782, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 22410 + }, + { + "epoch": 1.0472781954887218, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2012, + "step": 22411 + }, + { + "epoch": 1.0473383458646617, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 22412 + }, + { + "epoch": 1.0473984962406016, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 22413 + }, + { + "epoch": 1.0474586466165414, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 22414 + }, + { + "epoch": 1.0475187969924813, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 22415 + }, + { + "epoch": 1.0475789473684212, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 22416 + }, + { + "epoch": 1.0476390977443608, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2215, + "step": 22417 + }, + { + "epoch": 1.0476992481203007, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 22418 + }, + { + "epoch": 1.0477593984962406, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 22419 + }, + { + "epoch": 1.0478195488721804, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 22420 + }, + { + "epoch": 1.0478796992481203, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 22421 + }, + { + "epoch": 1.0479398496240602, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 22422 + }, + { + "epoch": 1.048, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 22423 + }, + { + "epoch": 1.04806015037594, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 22424 + }, + { + "epoch": 1.0481203007518798, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 22425 + }, + { + "epoch": 1.0481804511278194, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 22426 + }, + { + "epoch": 1.0482406015037593, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 22427 + }, + { + "epoch": 1.0483007518796992, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 22428 + }, + { + "epoch": 1.048360902255639, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 22429 + }, + { + "epoch": 1.048421052631579, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 22430 + }, + { + "epoch": 1.0484812030075188, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 22431 + }, + { + "epoch": 1.0485413533834587, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1804, + "step": 22432 + }, + { + "epoch": 1.0486015037593985, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 22433 + }, + { + "epoch": 1.0486616541353384, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 22434 + }, + { + "epoch": 1.0487218045112783, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 22435 + }, + { + "epoch": 1.0487819548872181, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 22436 + }, + { + "epoch": 1.0488421052631578, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 22437 + }, + { + "epoch": 1.0489022556390977, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 22438 + }, + { + "epoch": 1.0489624060150375, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 22439 + }, + { + "epoch": 1.0490225563909774, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 22440 + }, + { + "epoch": 1.0490827067669173, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 22441 + }, + { + "epoch": 1.0491428571428572, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 22442 + }, + { + "epoch": 1.049203007518797, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 22443 + }, + { + "epoch": 1.049263157894737, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 22444 + }, + { + "epoch": 1.0493233082706768, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 22445 + }, + { + "epoch": 1.0493834586466166, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 22446 + }, + { + "epoch": 1.0494436090225563, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 22447 + }, + { + "epoch": 1.0495037593984962, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 22448 + }, + { + "epoch": 1.049563909774436, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 22449 + }, + { + "epoch": 1.049624060150376, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 22450 + }, + { + "epoch": 1.0496842105263158, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 22451 + }, + { + "epoch": 1.0497443609022556, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 22452 + }, + { + "epoch": 1.0498045112781955, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 22453 + }, + { + "epoch": 1.0498646616541354, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 22454 + }, + { + "epoch": 1.0499248120300753, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 22455 + }, + { + "epoch": 1.0499849624060151, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 22456 + }, + { + "epoch": 1.0500451127819548, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 22457 + }, + { + "epoch": 1.0501052631578947, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 22458 + }, + { + "epoch": 1.0501654135338345, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 22459 + }, + { + "epoch": 1.0502255639097744, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 22460 + }, + { + "epoch": 1.0502857142857143, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 22461 + }, + { + "epoch": 1.0503458646616541, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 22462 + }, + { + "epoch": 1.050406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 22463 + }, + { + "epoch": 1.0504661654135339, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 22464 + }, + { + "epoch": 1.0505263157894738, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1962, + "step": 22465 + }, + { + "epoch": 1.0505864661654136, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 22466 + }, + { + "epoch": 1.0506466165413535, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 22467 + }, + { + "epoch": 1.0507067669172931, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 22468 + }, + { + "epoch": 1.050766917293233, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1642, + "step": 22469 + }, + { + "epoch": 1.0508270676691729, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 22470 + }, + { + "epoch": 1.0508872180451128, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 22471 + }, + { + "epoch": 1.0509473684210526, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 22472 + }, + { + "epoch": 1.0510075187969925, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 22473 + }, + { + "epoch": 1.0510676691729324, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 22474 + }, + { + "epoch": 1.0511278195488722, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 22475 + }, + { + "epoch": 1.0511879699248121, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 22476 + }, + { + "epoch": 1.051248120300752, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 22477 + }, + { + "epoch": 1.0513082706766916, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 22478 + }, + { + "epoch": 1.0513684210526315, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 22479 + }, + { + "epoch": 1.0514285714285714, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 22480 + }, + { + "epoch": 1.0514887218045113, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 22481 + }, + { + "epoch": 1.0515488721804511, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 22482 + }, + { + "epoch": 1.051609022556391, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 22483 + }, + { + "epoch": 1.0516691729323309, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 22484 + }, + { + "epoch": 1.0517293233082707, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 22485 + }, + { + "epoch": 1.0517894736842106, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 22486 + }, + { + "epoch": 1.0518496240601505, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 22487 + }, + { + "epoch": 1.0519097744360901, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 22488 + }, + { + "epoch": 1.05196992481203, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 22489 + }, + { + "epoch": 1.0520300751879699, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 22490 + }, + { + "epoch": 1.0520902255639097, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 22491 + }, + { + "epoch": 1.0521503759398496, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 22492 + }, + { + "epoch": 1.0522105263157895, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 22493 + }, + { + "epoch": 1.0522706766917294, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 22494 + }, + { + "epoch": 1.0523308270676692, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 22495 + }, + { + "epoch": 1.052390977443609, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 22496 + }, + { + "epoch": 1.052451127819549, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 22497 + }, + { + "epoch": 1.0525112781954886, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 22498 + }, + { + "epoch": 1.0525714285714285, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 22499 + }, + { + "epoch": 1.0526315789473684, + "grad_norm": 1.4375, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 22500 + }, + { + "epoch": 1.0526917293233082, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2675, + "step": 22501 + }, + { + "epoch": 1.052751879699248, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 22502 + }, + { + "epoch": 1.052812030075188, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 22503 + }, + { + "epoch": 1.0528721804511278, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 22504 + }, + { + "epoch": 1.0529323308270677, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 22505 + }, + { + "epoch": 1.0529924812030076, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.2063, + "step": 22506 + }, + { + "epoch": 1.0530526315789475, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 22507 + }, + { + "epoch": 1.0531127819548871, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 22508 + }, + { + "epoch": 1.053172932330827, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 22509 + }, + { + "epoch": 1.0532330827067669, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 22510 + }, + { + "epoch": 1.0532932330827067, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 22511 + }, + { + "epoch": 1.0533533834586466, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 22512 + }, + { + "epoch": 1.0534135338345865, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2422, + "step": 22513 + }, + { + "epoch": 1.0534736842105263, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 22514 + }, + { + "epoch": 1.0535338345864662, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 22515 + }, + { + "epoch": 1.053593984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 22516 + }, + { + "epoch": 1.053654135338346, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 22517 + }, + { + "epoch": 1.0537142857142858, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 22518 + }, + { + "epoch": 1.0537744360902255, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 22519 + }, + { + "epoch": 1.0538345864661653, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 22520 + }, + { + "epoch": 1.0538947368421052, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 22521 + }, + { + "epoch": 1.053954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 22522 + }, + { + "epoch": 1.054015037593985, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 22523 + }, + { + "epoch": 1.0540751879699248, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 22524 + }, + { + "epoch": 1.0541353383458647, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 22525 + }, + { + "epoch": 1.0541954887218046, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 22526 + }, + { + "epoch": 1.0542556390977444, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 22527 + }, + { + "epoch": 1.0543157894736843, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 22528 + }, + { + "epoch": 1.054375939849624, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 22529 + }, + { + "epoch": 1.0544360902255638, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 22530 + }, + { + "epoch": 1.0544962406015037, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 22531 + }, + { + "epoch": 1.0545563909774436, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 22532 + }, + { + "epoch": 1.0546165413533835, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 22533 + }, + { + "epoch": 1.0546766917293233, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 22534 + }, + { + "epoch": 1.0547368421052632, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 22535 + }, + { + "epoch": 1.054796992481203, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 22536 + }, + { + "epoch": 1.054857142857143, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 22537 + }, + { + "epoch": 1.0549172932330828, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 22538 + }, + { + "epoch": 1.0549774436090225, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 22539 + }, + { + "epoch": 1.0550375939849623, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 22540 + }, + { + "epoch": 1.0550977443609022, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 22541 + }, + { + "epoch": 1.055157894736842, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 22542 + }, + { + "epoch": 1.055218045112782, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 22543 + }, + { + "epoch": 1.0552781954887218, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 22544 + }, + { + "epoch": 1.0553383458646617, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 22545 + }, + { + "epoch": 1.0553984962406016, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 22546 + }, + { + "epoch": 1.0554586466165414, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 22547 + }, + { + "epoch": 1.0555187969924813, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.147, + "step": 22548 + }, + { + "epoch": 1.055578947368421, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 22549 + }, + { + "epoch": 1.0556390977443608, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 22550 + }, + { + "epoch": 1.0556992481203007, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 22551 + }, + { + "epoch": 1.0557593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 22552 + }, + { + "epoch": 1.0558195488721804, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 22553 + }, + { + "epoch": 1.0558796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 22554 + }, + { + "epoch": 1.0559398496240602, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2117, + "step": 22555 + }, + { + "epoch": 1.056, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 22556 + }, + { + "epoch": 1.05606015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 22557 + }, + { + "epoch": 1.0561203007518798, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 22558 + }, + { + "epoch": 1.0561804511278194, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 22559 + }, + { + "epoch": 1.0562406015037593, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 22560 + }, + { + "epoch": 1.0563007518796992, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 22561 + }, + { + "epoch": 1.056360902255639, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 22562 + }, + { + "epoch": 1.056421052631579, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 22563 + }, + { + "epoch": 1.0564812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 22564 + }, + { + "epoch": 1.0565413533834587, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 22565 + }, + { + "epoch": 1.0566015037593985, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 22566 + }, + { + "epoch": 1.0566616541353384, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 22567 + }, + { + "epoch": 1.0567218045112783, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 22568 + }, + { + "epoch": 1.0567819548872182, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 22569 + }, + { + "epoch": 1.0568421052631578, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 22570 + }, + { + "epoch": 1.0569022556390977, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 22571 + }, + { + "epoch": 1.0569624060150375, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2269, + "step": 22572 + }, + { + "epoch": 1.0570225563909774, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 22573 + }, + { + "epoch": 1.0570827067669173, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 22574 + }, + { + "epoch": 1.0571428571428572, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 22575 + }, + { + "epoch": 1.057203007518797, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 22576 + }, + { + "epoch": 1.057263157894737, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 22577 + }, + { + "epoch": 1.0573233082706768, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 22578 + }, + { + "epoch": 1.0573834586466166, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 22579 + }, + { + "epoch": 1.0574436090225563, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 22580 + }, + { + "epoch": 1.0575037593984962, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 22581 + }, + { + "epoch": 1.057563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 22582 + }, + { + "epoch": 1.057624060150376, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 22583 + }, + { + "epoch": 1.0576842105263158, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 22584 + }, + { + "epoch": 1.0577443609022557, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 22585 + }, + { + "epoch": 1.0578045112781955, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 22586 + }, + { + "epoch": 1.0578646616541354, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 22587 + }, + { + "epoch": 1.0579248120300753, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 22588 + }, + { + "epoch": 1.0579849624060151, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 22589 + }, + { + "epoch": 1.0580451127819548, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 22590 + }, + { + "epoch": 1.0581052631578947, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 22591 + }, + { + "epoch": 1.0581654135338345, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 22592 + }, + { + "epoch": 1.0582255639097744, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 22593 + }, + { + "epoch": 1.0582857142857143, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 22594 + }, + { + "epoch": 1.0583458646616541, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 22595 + }, + { + "epoch": 1.058406015037594, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 22596 + }, + { + "epoch": 1.0584661654135339, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 22597 + }, + { + "epoch": 1.0585263157894738, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 22598 + }, + { + "epoch": 1.0585864661654136, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 22599 + }, + { + "epoch": 1.0586466165413533, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 22600 + }, + { + "epoch": 1.0587067669172932, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 22601 + }, + { + "epoch": 1.058766917293233, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 22602 + }, + { + "epoch": 1.058827067669173, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 22603 + }, + { + "epoch": 1.0588872180451128, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 22604 + }, + { + "epoch": 1.0589473684210526, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 22605 + }, + { + "epoch": 1.0590075187969925, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 22606 + }, + { + "epoch": 1.0590676691729324, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 22607 + }, + { + "epoch": 1.0591278195488723, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 22608 + }, + { + "epoch": 1.0591879699248121, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 22609 + }, + { + "epoch": 1.0592481203007518, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 22610 + }, + { + "epoch": 1.0593082706766916, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 22611 + }, + { + "epoch": 1.0593684210526315, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 22612 + }, + { + "epoch": 1.0594285714285714, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 22613 + }, + { + "epoch": 1.0594887218045113, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 22614 + }, + { + "epoch": 1.0595488721804511, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 22615 + }, + { + "epoch": 1.059609022556391, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 22616 + }, + { + "epoch": 1.0596691729323309, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 22617 + }, + { + "epoch": 1.0597293233082707, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 22618 + }, + { + "epoch": 1.0597894736842106, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 22619 + }, + { + "epoch": 1.0598496240601505, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 22620 + }, + { + "epoch": 1.0599097744360901, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 22621 + }, + { + "epoch": 1.05996992481203, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 22622 + }, + { + "epoch": 1.0600300751879699, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 22623 + }, + { + "epoch": 1.0600902255639097, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 22624 + }, + { + "epoch": 1.0601503759398496, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 22625 + }, + { + "epoch": 1.0602105263157895, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 22626 + }, + { + "epoch": 1.0602706766917294, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 22627 + }, + { + "epoch": 1.0603308270676692, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 22628 + }, + { + "epoch": 1.060390977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 22629 + }, + { + "epoch": 1.060451127819549, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 22630 + }, + { + "epoch": 1.0605112781954886, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 22631 + }, + { + "epoch": 1.0605714285714285, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 22632 + }, + { + "epoch": 1.0606315789473684, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 22633 + }, + { + "epoch": 1.0606917293233082, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 22634 + }, + { + "epoch": 1.0607518796992481, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 22635 + }, + { + "epoch": 1.060812030075188, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 22636 + }, + { + "epoch": 1.0608721804511279, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 22637 + }, + { + "epoch": 1.0609323308270677, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 22638 + }, + { + "epoch": 1.0609924812030076, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 22639 + }, + { + "epoch": 1.0610526315789475, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 22640 + }, + { + "epoch": 1.0611127819548871, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 22641 + }, + { + "epoch": 1.061172932330827, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 22642 + }, + { + "epoch": 1.0612330827067669, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 22643 + }, + { + "epoch": 1.0612932330827067, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 22644 + }, + { + "epoch": 1.0613533834586466, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 22645 + }, + { + "epoch": 1.0614135338345865, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 22646 + }, + { + "epoch": 1.0614736842105263, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 22647 + }, + { + "epoch": 1.0615338345864662, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 22648 + }, + { + "epoch": 1.061593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 22649 + }, + { + "epoch": 1.061654135338346, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 22650 + }, + { + "epoch": 1.0617142857142856, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1414, + "step": 22651 + }, + { + "epoch": 1.0617744360902255, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 22652 + }, + { + "epoch": 1.0618345864661654, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 22653 + }, + { + "epoch": 1.0618947368421052, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 22654 + }, + { + "epoch": 1.061954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 22655 + }, + { + "epoch": 1.062015037593985, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 22656 + }, + { + "epoch": 1.0620751879699248, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 22657 + }, + { + "epoch": 1.0621353383458647, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 22658 + }, + { + "epoch": 1.0621954887218046, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 22659 + }, + { + "epoch": 1.0622556390977445, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 22660 + }, + { + "epoch": 1.062315789473684, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 22661 + }, + { + "epoch": 1.062375939849624, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 22662 + }, + { + "epoch": 1.0624360902255638, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 22663 + }, + { + "epoch": 1.0624962406015037, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 22664 + }, + { + "epoch": 1.0625563909774436, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 22665 + }, + { + "epoch": 1.0626165413533835, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 22666 + }, + { + "epoch": 1.0626766917293233, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 22667 + }, + { + "epoch": 1.0627368421052632, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 22668 + }, + { + "epoch": 1.062796992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 22669 + }, + { + "epoch": 1.062857142857143, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 22670 + }, + { + "epoch": 1.0629172932330828, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 22671 + }, + { + "epoch": 1.0629774436090225, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 22672 + }, + { + "epoch": 1.0630375939849623, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 22673 + }, + { + "epoch": 1.0630977443609022, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 22674 + }, + { + "epoch": 1.063157894736842, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 22675 + }, + { + "epoch": 1.063218045112782, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 22676 + }, + { + "epoch": 1.0632781954887218, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 22677 + }, + { + "epoch": 1.0633383458646617, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 22678 + }, + { + "epoch": 1.0633984962406016, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 22679 + }, + { + "epoch": 1.0634586466165414, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 22680 + }, + { + "epoch": 1.063518796992481, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1271, + "step": 22681 + }, + { + "epoch": 1.063578947368421, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 22682 + }, + { + "epoch": 1.0636390977443608, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 22683 + }, + { + "epoch": 1.0636992481203007, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 22684 + }, + { + "epoch": 1.0637593984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 22685 + }, + { + "epoch": 1.0638195488721804, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 22686 + }, + { + "epoch": 1.0638796992481203, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 22687 + }, + { + "epoch": 1.0639398496240602, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 22688 + }, + { + "epoch": 1.064, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 22689 + }, + { + "epoch": 1.06406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 22690 + }, + { + "epoch": 1.0641203007518798, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 22691 + }, + { + "epoch": 1.0641804511278194, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 22692 + }, + { + "epoch": 1.0642406015037593, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 22693 + }, + { + "epoch": 1.0643007518796992, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 22694 + }, + { + "epoch": 1.064360902255639, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 22695 + }, + { + "epoch": 1.064421052631579, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 22696 + }, + { + "epoch": 1.0644812030075188, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 22697 + }, + { + "epoch": 1.0645413533834587, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 22698 + }, + { + "epoch": 1.0646015037593985, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 22699 + }, + { + "epoch": 1.0646616541353384, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.224, + "step": 22700 + }, + { + "epoch": 1.0647218045112783, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 22701 + }, + { + "epoch": 1.0647819548872182, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 22702 + }, + { + "epoch": 1.0648421052631578, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 22703 + }, + { + "epoch": 1.0649022556390977, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 22704 + }, + { + "epoch": 1.0649624060150376, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 22705 + }, + { + "epoch": 1.0650225563909774, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 22706 + }, + { + "epoch": 1.0650827067669173, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 22707 + }, + { + "epoch": 1.0651428571428572, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 22708 + }, + { + "epoch": 1.065203007518797, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 22709 + }, + { + "epoch": 1.065263157894737, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 22710 + }, + { + "epoch": 1.0653233082706768, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 22711 + }, + { + "epoch": 1.0653834586466164, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 22712 + }, + { + "epoch": 1.0654436090225563, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 22713 + }, + { + "epoch": 1.0655037593984962, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 22714 + }, + { + "epoch": 1.065563909774436, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 22715 + }, + { + "epoch": 1.065624060150376, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 22716 + }, + { + "epoch": 1.0656842105263158, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 22717 + }, + { + "epoch": 1.0657443609022557, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 22718 + }, + { + "epoch": 1.0658045112781955, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 22719 + }, + { + "epoch": 1.0658646616541354, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 22720 + }, + { + "epoch": 1.0659248120300753, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 22721 + }, + { + "epoch": 1.0659849624060151, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 22722 + }, + { + "epoch": 1.0660451127819548, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 22723 + }, + { + "epoch": 1.0661052631578947, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 22724 + }, + { + "epoch": 1.0661654135338345, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 22725 + }, + { + "epoch": 1.0662255639097744, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 22726 + }, + { + "epoch": 1.0662857142857143, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 22727 + }, + { + "epoch": 1.0663458646616542, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 22728 + }, + { + "epoch": 1.066406015037594, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 22729 + }, + { + "epoch": 1.066466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 22730 + }, + { + "epoch": 1.0665263157894738, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 22731 + }, + { + "epoch": 1.0665864661654134, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 22732 + }, + { + "epoch": 1.0666466165413533, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 22733 + }, + { + "epoch": 1.0667067669172932, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 22734 + }, + { + "epoch": 1.066766917293233, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 22735 + }, + { + "epoch": 1.066827067669173, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 22736 + }, + { + "epoch": 1.0668872180451128, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 22737 + }, + { + "epoch": 1.0669473684210526, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 22738 + }, + { + "epoch": 1.0670075187969925, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 22739 + }, + { + "epoch": 1.0670676691729324, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 22740 + }, + { + "epoch": 1.0671278195488723, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 22741 + }, + { + "epoch": 1.0671879699248121, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 22742 + }, + { + "epoch": 1.0672481203007518, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 22743 + }, + { + "epoch": 1.0673082706766917, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 22744 + }, + { + "epoch": 1.0673684210526315, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 22745 + }, + { + "epoch": 1.0674285714285714, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 22746 + }, + { + "epoch": 1.0674887218045113, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 22747 + }, + { + "epoch": 1.0675488721804511, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 22748 + }, + { + "epoch": 1.067609022556391, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 22749 + }, + { + "epoch": 1.0676691729323309, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 22750 + }, + { + "epoch": 1.0677293233082708, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 22751 + }, + { + "epoch": 1.0677894736842106, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 22752 + }, + { + "epoch": 1.0678496240601505, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 22753 + }, + { + "epoch": 1.0679097744360901, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 22754 + }, + { + "epoch": 1.06796992481203, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 22755 + }, + { + "epoch": 1.0680300751879699, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2626, + "step": 22756 + }, + { + "epoch": 1.0680902255639098, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 22757 + }, + { + "epoch": 1.0681503759398496, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 22758 + }, + { + "epoch": 1.0682105263157895, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 22759 + }, + { + "epoch": 1.0682706766917294, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1689, + "step": 22760 + }, + { + "epoch": 1.0683308270676692, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 22761 + }, + { + "epoch": 1.0683909774436091, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 22762 + }, + { + "epoch": 1.0684511278195488, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 22763 + }, + { + "epoch": 1.0685112781954886, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 22764 + }, + { + "epoch": 1.0685714285714285, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 22765 + }, + { + "epoch": 1.0686315789473684, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 22766 + }, + { + "epoch": 1.0686917293233082, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 22767 + }, + { + "epoch": 1.0687518796992481, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 22768 + }, + { + "epoch": 1.068812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 22769 + }, + { + "epoch": 1.0688721804511279, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 22770 + }, + { + "epoch": 1.0689323308270677, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 22771 + }, + { + "epoch": 1.0689924812030076, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 22772 + }, + { + "epoch": 1.0690526315789475, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 22773 + }, + { + "epoch": 1.0691127819548871, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 22774 + }, + { + "epoch": 1.069172932330827, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 22775 + }, + { + "epoch": 1.0692330827067669, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 22776 + }, + { + "epoch": 1.0692932330827067, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 22777 + }, + { + "epoch": 1.0693533834586466, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 22778 + }, + { + "epoch": 1.0694135338345865, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 22779 + }, + { + "epoch": 1.0694736842105264, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 22780 + }, + { + "epoch": 1.0695338345864662, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 22781 + }, + { + "epoch": 1.069593984962406, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 22782 + }, + { + "epoch": 1.069654135338346, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 22783 + }, + { + "epoch": 1.0697142857142856, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 22784 + }, + { + "epoch": 1.0697744360902255, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 22785 + }, + { + "epoch": 1.0698345864661654, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 22786 + }, + { + "epoch": 1.0698947368421052, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 22787 + }, + { + "epoch": 1.069954887218045, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 22788 + }, + { + "epoch": 1.070015037593985, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 22789 + }, + { + "epoch": 1.0700751879699248, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1829, + "step": 22790 + }, + { + "epoch": 1.0701353383458647, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 22791 + }, + { + "epoch": 1.0701954887218046, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 22792 + }, + { + "epoch": 1.0702556390977445, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 22793 + }, + { + "epoch": 1.070315789473684, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 22794 + }, + { + "epoch": 1.070375939849624, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 22795 + }, + { + "epoch": 1.0704360902255639, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 22796 + }, + { + "epoch": 1.0704962406015037, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 22797 + }, + { + "epoch": 1.0705563909774436, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 22798 + }, + { + "epoch": 1.0706165413533835, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 22799 + }, + { + "epoch": 1.0706766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 22800 + }, + { + "epoch": 1.0707368421052632, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 22801 + }, + { + "epoch": 1.070796992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 22802 + }, + { + "epoch": 1.070857142857143, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 22803 + }, + { + "epoch": 1.0709172932330828, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 22804 + }, + { + "epoch": 1.0709774436090225, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 22805 + }, + { + "epoch": 1.0710375939849623, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 22806 + }, + { + "epoch": 1.0710977443609022, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 22807 + }, + { + "epoch": 1.071157894736842, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 22808 + }, + { + "epoch": 1.071218045112782, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 22809 + }, + { + "epoch": 1.0712781954887218, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 22810 + }, + { + "epoch": 1.0713383458646617, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 22811 + }, + { + "epoch": 1.0713984962406016, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 22812 + }, + { + "epoch": 1.0714586466165414, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 22813 + }, + { + "epoch": 1.071518796992481, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 22814 + }, + { + "epoch": 1.071578947368421, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 22815 + }, + { + "epoch": 1.0716390977443608, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 22816 + }, + { + "epoch": 1.0716992481203007, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 22817 + }, + { + "epoch": 1.0717593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 22818 + }, + { + "epoch": 1.0718195488721804, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 22819 + }, + { + "epoch": 1.0718796992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 22820 + }, + { + "epoch": 1.0719398496240602, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 22821 + }, + { + "epoch": 1.072, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 22822 + }, + { + "epoch": 1.07206015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 22823 + }, + { + "epoch": 1.0721203007518798, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 22824 + }, + { + "epoch": 1.0721804511278195, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 22825 + }, + { + "epoch": 1.0722406015037593, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 22826 + }, + { + "epoch": 1.0723007518796992, + "grad_norm": 1.46875, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 22827 + }, + { + "epoch": 1.072360902255639, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 22828 + }, + { + "epoch": 1.072421052631579, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 22829 + }, + { + "epoch": 1.0724812030075188, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 22830 + }, + { + "epoch": 1.0725413533834587, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 22831 + }, + { + "epoch": 1.0726015037593986, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 22832 + }, + { + "epoch": 1.0726616541353384, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 22833 + }, + { + "epoch": 1.0727218045112783, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 22834 + }, + { + "epoch": 1.072781954887218, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 22835 + }, + { + "epoch": 1.0728421052631578, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1786, + "step": 22836 + }, + { + "epoch": 1.0729022556390977, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 22837 + }, + { + "epoch": 1.0729624060150376, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 22838 + }, + { + "epoch": 1.0730225563909774, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 22839 + }, + { + "epoch": 1.0730827067669173, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 22840 + }, + { + "epoch": 1.0731428571428572, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 22841 + }, + { + "epoch": 1.073203007518797, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 22842 + }, + { + "epoch": 1.073263157894737, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 22843 + }, + { + "epoch": 1.0733233082706768, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 22844 + }, + { + "epoch": 1.0733834586466164, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 22845 + }, + { + "epoch": 1.0734436090225563, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 22846 + }, + { + "epoch": 1.0735037593984962, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 22847 + }, + { + "epoch": 1.073563909774436, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 22848 + }, + { + "epoch": 1.073624060150376, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 22849 + }, + { + "epoch": 1.0736842105263158, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 22850 + }, + { + "epoch": 1.0737443609022557, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 22851 + }, + { + "epoch": 1.0738045112781955, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 22852 + }, + { + "epoch": 1.0738646616541354, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 22853 + }, + { + "epoch": 1.0739248120300753, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 22854 + }, + { + "epoch": 1.0739849624060152, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 22855 + }, + { + "epoch": 1.0740451127819548, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 22856 + }, + { + "epoch": 1.0741052631578947, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 22857 + }, + { + "epoch": 1.0741654135338345, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 22858 + }, + { + "epoch": 1.0742255639097744, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 22859 + }, + { + "epoch": 1.0742857142857143, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 22860 + }, + { + "epoch": 1.0743458646616542, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 22861 + }, + { + "epoch": 1.074406015037594, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 22862 + }, + { + "epoch": 1.074466165413534, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 22863 + }, + { + "epoch": 1.0745263157894738, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 22864 + }, + { + "epoch": 1.0745864661654134, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 22865 + }, + { + "epoch": 1.0746466165413533, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 22866 + }, + { + "epoch": 1.0747067669172932, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 22867 + }, + { + "epoch": 1.074766917293233, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 22868 + }, + { + "epoch": 1.074827067669173, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 22869 + }, + { + "epoch": 1.0748872180451128, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 22870 + }, + { + "epoch": 1.0749473684210527, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 22871 + }, + { + "epoch": 1.0750075187969925, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 22872 + }, + { + "epoch": 1.0750676691729324, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 22873 + }, + { + "epoch": 1.0751278195488723, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 22874 + }, + { + "epoch": 1.0751879699248121, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 22875 + }, + { + "epoch": 1.0752481203007518, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 22876 + }, + { + "epoch": 1.0753082706766917, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 22877 + }, + { + "epoch": 1.0753684210526315, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 22878 + }, + { + "epoch": 1.0754285714285714, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1414, + "step": 22879 + }, + { + "epoch": 1.0754887218045113, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 22880 + }, + { + "epoch": 1.0755488721804511, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 22881 + }, + { + "epoch": 1.075609022556391, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 22882 + }, + { + "epoch": 1.0756691729323309, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 22883 + }, + { + "epoch": 1.0757293233082708, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2065, + "step": 22884 + }, + { + "epoch": 1.0757894736842106, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 22885 + }, + { + "epoch": 1.0758496240601503, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 22886 + }, + { + "epoch": 1.0759097744360901, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 22887 + }, + { + "epoch": 1.07596992481203, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 22888 + }, + { + "epoch": 1.07603007518797, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 22889 + }, + { + "epoch": 1.0760902255639098, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 22890 + }, + { + "epoch": 1.0761503759398496, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 22891 + }, + { + "epoch": 1.0762105263157895, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 22892 + }, + { + "epoch": 1.0762706766917294, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 22893 + }, + { + "epoch": 1.0763308270676692, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 22894 + }, + { + "epoch": 1.0763909774436091, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 22895 + }, + { + "epoch": 1.0764511278195488, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 22896 + }, + { + "epoch": 1.0765112781954886, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2043, + "step": 22897 + }, + { + "epoch": 1.0765714285714285, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 22898 + }, + { + "epoch": 1.0766315789473684, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 22899 + }, + { + "epoch": 1.0766917293233083, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 22900 + }, + { + "epoch": 1.0767518796992481, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 22901 + }, + { + "epoch": 1.076812030075188, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1124, + "step": 22902 + }, + { + "epoch": 1.0768721804511279, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 22903 + }, + { + "epoch": 1.0769323308270677, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 22904 + }, + { + "epoch": 1.0769924812030076, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 22905 + }, + { + "epoch": 1.0770526315789475, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 22906 + }, + { + "epoch": 1.0771127819548871, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 22907 + }, + { + "epoch": 1.077172932330827, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 22908 + }, + { + "epoch": 1.0772330827067669, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 22909 + }, + { + "epoch": 1.0772932330827067, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 22910 + }, + { + "epoch": 1.0773533834586466, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 22911 + }, + { + "epoch": 1.0774135338345865, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 22912 + }, + { + "epoch": 1.0774736842105264, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 22913 + }, + { + "epoch": 1.0775338345864662, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 22914 + }, + { + "epoch": 1.077593984962406, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 22915 + }, + { + "epoch": 1.0776541353383458, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 22916 + }, + { + "epoch": 1.0777142857142856, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 22917 + }, + { + "epoch": 1.0777744360902255, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 22918 + }, + { + "epoch": 1.0778345864661654, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 22919 + }, + { + "epoch": 1.0778947368421052, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 22920 + }, + { + "epoch": 1.077954887218045, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 22921 + }, + { + "epoch": 1.078015037593985, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 22922 + }, + { + "epoch": 1.0780751879699249, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 22923 + }, + { + "epoch": 1.0781353383458647, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 22924 + }, + { + "epoch": 1.0781954887218046, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 22925 + }, + { + "epoch": 1.0782556390977445, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 22926 + }, + { + "epoch": 1.0783157894736841, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 22927 + }, + { + "epoch": 1.078375939849624, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 22928 + }, + { + "epoch": 1.0784360902255639, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 22929 + }, + { + "epoch": 1.0784962406015037, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 22930 + }, + { + "epoch": 1.0785563909774436, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 22931 + }, + { + "epoch": 1.0786165413533835, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 22932 + }, + { + "epoch": 1.0786766917293233, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 22933 + }, + { + "epoch": 1.0787368421052632, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 22934 + }, + { + "epoch": 1.078796992481203, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 22935 + }, + { + "epoch": 1.078857142857143, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 22936 + }, + { + "epoch": 1.0789172932330826, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 22937 + }, + { + "epoch": 1.0789774436090225, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 22938 + }, + { + "epoch": 1.0790375939849624, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 22939 + }, + { + "epoch": 1.0790977443609022, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 22940 + }, + { + "epoch": 1.079157894736842, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 22941 + }, + { + "epoch": 1.079218045112782, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 22942 + }, + { + "epoch": 1.0792781954887218, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 22943 + }, + { + "epoch": 1.0793383458646617, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 22944 + }, + { + "epoch": 1.0793984962406016, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 22945 + }, + { + "epoch": 1.0794586466165415, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 22946 + }, + { + "epoch": 1.079518796992481, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 22947 + }, + { + "epoch": 1.079578947368421, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 22948 + }, + { + "epoch": 1.0796390977443608, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 22949 + }, + { + "epoch": 1.0796992481203007, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 22950 + }, + { + "epoch": 1.0797593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 22951 + }, + { + "epoch": 1.0798195488721805, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 22952 + }, + { + "epoch": 1.0798796992481203, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 22953 + }, + { + "epoch": 1.0799398496240602, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 22954 + }, + { + "epoch": 1.08, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 22955 + }, + { + "epoch": 1.08006015037594, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 22956 + }, + { + "epoch": 1.0801203007518798, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 22957 + }, + { + "epoch": 1.0801804511278195, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 22958 + }, + { + "epoch": 1.0802406015037593, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 22959 + }, + { + "epoch": 1.0803007518796992, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 22960 + }, + { + "epoch": 1.080360902255639, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 22961 + }, + { + "epoch": 1.080421052631579, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 22962 + }, + { + "epoch": 1.0804812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 22963 + }, + { + "epoch": 1.0805413533834587, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 22964 + }, + { + "epoch": 1.0806015037593986, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 22965 + }, + { + "epoch": 1.0806616541353384, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 22966 + }, + { + "epoch": 1.080721804511278, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 22967 + }, + { + "epoch": 1.080781954887218, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 22968 + }, + { + "epoch": 1.0808421052631578, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 22969 + }, + { + "epoch": 1.0809022556390977, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 22970 + }, + { + "epoch": 1.0809624060150376, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 22971 + }, + { + "epoch": 1.0810225563909774, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 22972 + }, + { + "epoch": 1.0810827067669173, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 22973 + }, + { + "epoch": 1.0811428571428572, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 22974 + }, + { + "epoch": 1.081203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 22975 + }, + { + "epoch": 1.081263157894737, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 22976 + }, + { + "epoch": 1.0813233082706768, + "grad_norm": 0.470703125, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 22977 + }, + { + "epoch": 1.0813834586466164, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 22978 + }, + { + "epoch": 1.0814436090225563, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 22979 + }, + { + "epoch": 1.0815037593984962, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 22980 + }, + { + "epoch": 1.081563909774436, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 22981 + }, + { + "epoch": 1.081624060150376, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 22982 + }, + { + "epoch": 1.0816842105263158, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 22983 + }, + { + "epoch": 1.0817443609022557, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 22984 + }, + { + "epoch": 1.0818045112781955, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 22985 + }, + { + "epoch": 1.0818646616541354, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 22986 + }, + { + "epoch": 1.0819248120300753, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 22987 + }, + { + "epoch": 1.081984962406015, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 22988 + }, + { + "epoch": 1.0820451127819548, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 22989 + }, + { + "epoch": 1.0821052631578947, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 22990 + }, + { + "epoch": 1.0821654135338346, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 22991 + }, + { + "epoch": 1.0822255639097744, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 22992 + }, + { + "epoch": 1.0822857142857143, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.3658, + "step": 22993 + }, + { + "epoch": 1.0823458646616542, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 22994 + }, + { + "epoch": 1.082406015037594, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 22995 + }, + { + "epoch": 1.082466165413534, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 22996 + }, + { + "epoch": 1.0825263157894738, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 22997 + }, + { + "epoch": 1.0825864661654134, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 22998 + }, + { + "epoch": 1.0826466165413533, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 22999 + }, + { + "epoch": 1.0827067669172932, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 23000 + }, + { + "epoch": 1.082766917293233, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 23001 + }, + { + "epoch": 1.082827067669173, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 23002 + }, + { + "epoch": 1.0828872180451128, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 23003 + }, + { + "epoch": 1.0829473684210527, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1696, + "step": 23004 + }, + { + "epoch": 1.0830075187969925, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 23005 + }, + { + "epoch": 1.0830676691729324, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 23006 + }, + { + "epoch": 1.0831278195488723, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 23007 + }, + { + "epoch": 1.0831879699248121, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 23008 + }, + { + "epoch": 1.0832481203007518, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 23009 + }, + { + "epoch": 1.0833082706766917, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 23010 + }, + { + "epoch": 1.0833684210526315, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 23011 + }, + { + "epoch": 1.0834285714285714, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 23012 + }, + { + "epoch": 1.0834887218045113, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 23013 + }, + { + "epoch": 1.0835488721804512, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 23014 + }, + { + "epoch": 1.083609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 23015 + }, + { + "epoch": 1.083669172932331, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 23016 + }, + { + "epoch": 1.0837293233082708, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1467, + "step": 23017 + }, + { + "epoch": 1.0837894736842104, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 23018 + }, + { + "epoch": 1.0838496240601503, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 23019 + }, + { + "epoch": 1.0839097744360902, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 23020 + }, + { + "epoch": 1.08396992481203, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 23021 + }, + { + "epoch": 1.08403007518797, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 23022 + }, + { + "epoch": 1.0840902255639098, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 23023 + }, + { + "epoch": 1.0841503759398496, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 23024 + }, + { + "epoch": 1.0842105263157895, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 23025 + }, + { + "epoch": 1.0842706766917294, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 23026 + }, + { + "epoch": 1.0843308270676693, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 23027 + }, + { + "epoch": 1.0843909774436091, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2295, + "step": 23028 + }, + { + "epoch": 1.0844511278195488, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 23029 + }, + { + "epoch": 1.0845112781954886, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 23030 + }, + { + "epoch": 1.0845714285714285, + "grad_norm": 1.65625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 23031 + }, + { + "epoch": 1.0846315789473684, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 23032 + }, + { + "epoch": 1.0846917293233083, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 23033 + }, + { + "epoch": 1.0847518796992481, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 23034 + }, + { + "epoch": 1.084812030075188, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2116, + "step": 23035 + }, + { + "epoch": 1.0848721804511279, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 23036 + }, + { + "epoch": 1.0849323308270677, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 23037 + }, + { + "epoch": 1.0849924812030076, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 23038 + }, + { + "epoch": 1.0850526315789473, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 23039 + }, + { + "epoch": 1.0851127819548871, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 23040 + }, + { + "epoch": 1.085172932330827, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 23041 + }, + { + "epoch": 1.0852330827067669, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 23042 + }, + { + "epoch": 1.0852932330827068, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 23043 + }, + { + "epoch": 1.0853533834586466, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 23044 + }, + { + "epoch": 1.0854135338345865, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 23045 + }, + { + "epoch": 1.0854736842105264, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 23046 + }, + { + "epoch": 1.0855338345864662, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 23047 + }, + { + "epoch": 1.0855939849624061, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 23048 + }, + { + "epoch": 1.0856541353383458, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 23049 + }, + { + "epoch": 1.0857142857142856, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 23050 + }, + { + "epoch": 1.0857744360902255, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 23051 + }, + { + "epoch": 1.0858345864661654, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 23052 + }, + { + "epoch": 1.0858947368421052, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 23053 + }, + { + "epoch": 1.0859548872180451, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 23054 + }, + { + "epoch": 1.086015037593985, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 23055 + }, + { + "epoch": 1.0860751879699249, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 23056 + }, + { + "epoch": 1.0861353383458647, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 23057 + }, + { + "epoch": 1.0861954887218046, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1932, + "step": 23058 + }, + { + "epoch": 1.0862556390977445, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 23059 + }, + { + "epoch": 1.0863157894736841, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 23060 + }, + { + "epoch": 1.086375939849624, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 23061 + }, + { + "epoch": 1.0864360902255639, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 23062 + }, + { + "epoch": 1.0864962406015037, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 23063 + }, + { + "epoch": 1.0865563909774436, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 23064 + }, + { + "epoch": 1.0866165413533835, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 23065 + }, + { + "epoch": 1.0866766917293234, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 23066 + }, + { + "epoch": 1.0867368421052632, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 23067 + }, + { + "epoch": 1.086796992481203, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 23068 + }, + { + "epoch": 1.0868571428571427, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 23069 + }, + { + "epoch": 1.0869172932330826, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 23070 + }, + { + "epoch": 1.0869774436090225, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 23071 + }, + { + "epoch": 1.0870375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 23072 + }, + { + "epoch": 1.0870977443609022, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 23073 + }, + { + "epoch": 1.087157894736842, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 23074 + }, + { + "epoch": 1.087218045112782, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 23075 + }, + { + "epoch": 1.0872781954887218, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2003, + "step": 23076 + }, + { + "epoch": 1.0873383458646617, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2007, + "step": 23077 + }, + { + "epoch": 1.0873984962406016, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 23078 + }, + { + "epoch": 1.0874586466165415, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 23079 + }, + { + "epoch": 1.087518796992481, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 23080 + }, + { + "epoch": 1.087578947368421, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1891, + "step": 23081 + }, + { + "epoch": 1.0876390977443608, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 23082 + }, + { + "epoch": 1.0876992481203007, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 23083 + }, + { + "epoch": 1.0877593984962406, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 23084 + }, + { + "epoch": 1.0878195488721805, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 23085 + }, + { + "epoch": 1.0878796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 23086 + }, + { + "epoch": 1.0879398496240602, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 23087 + }, + { + "epoch": 1.088, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 23088 + }, + { + "epoch": 1.08806015037594, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 23089 + }, + { + "epoch": 1.0881203007518796, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 23090 + }, + { + "epoch": 1.0881804511278195, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.186, + "step": 23091 + }, + { + "epoch": 1.0882406015037593, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 23092 + }, + { + "epoch": 1.0883007518796992, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 23093 + }, + { + "epoch": 1.088360902255639, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 23094 + }, + { + "epoch": 1.088421052631579, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 23095 + }, + { + "epoch": 1.0884812030075188, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 23096 + }, + { + "epoch": 1.0885413533834587, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 23097 + }, + { + "epoch": 1.0886015037593986, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 23098 + }, + { + "epoch": 1.0886616541353384, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 23099 + }, + { + "epoch": 1.088721804511278, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 23100 + }, + { + "epoch": 1.088781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 23101 + }, + { + "epoch": 1.0888421052631578, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 23102 + }, + { + "epoch": 1.0889022556390977, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 23103 + }, + { + "epoch": 1.0889624060150376, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 23104 + }, + { + "epoch": 1.0890225563909774, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 23105 + }, + { + "epoch": 1.0890827067669173, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 23106 + }, + { + "epoch": 1.0891428571428572, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 23107 + }, + { + "epoch": 1.089203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 23108 + }, + { + "epoch": 1.089263157894737, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 23109 + }, + { + "epoch": 1.0893233082706768, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 23110 + }, + { + "epoch": 1.0893834586466165, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 23111 + }, + { + "epoch": 1.0894436090225563, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 23112 + }, + { + "epoch": 1.0895037593984962, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 23113 + }, + { + "epoch": 1.089563909774436, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 23114 + }, + { + "epoch": 1.089624060150376, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 23115 + }, + { + "epoch": 1.0896842105263158, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 23116 + }, + { + "epoch": 1.0897443609022557, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 23117 + }, + { + "epoch": 1.0898045112781956, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 23118 + }, + { + "epoch": 1.0898646616541354, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 23119 + }, + { + "epoch": 1.089924812030075, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 23120 + }, + { + "epoch": 1.089984962406015, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 23121 + }, + { + "epoch": 1.0900451127819548, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 23122 + }, + { + "epoch": 1.0901052631578947, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 23123 + }, + { + "epoch": 1.0901654135338346, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 23124 + }, + { + "epoch": 1.0902255639097744, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 23125 + }, + { + "epoch": 1.0902857142857143, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 23126 + }, + { + "epoch": 1.0903458646616542, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2005, + "step": 23127 + }, + { + "epoch": 1.090406015037594, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 23128 + }, + { + "epoch": 1.090466165413534, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 23129 + }, + { + "epoch": 1.0905263157894738, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 23130 + }, + { + "epoch": 1.0905864661654134, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 23131 + }, + { + "epoch": 1.0906466165413533, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 23132 + }, + { + "epoch": 1.0907067669172932, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 23133 + }, + { + "epoch": 1.090766917293233, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 23134 + }, + { + "epoch": 1.090827067669173, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 23135 + }, + { + "epoch": 1.0908872180451128, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 23136 + }, + { + "epoch": 1.0909473684210527, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2465, + "step": 23137 + }, + { + "epoch": 1.0910075187969925, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 23138 + }, + { + "epoch": 1.0910676691729324, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 23139 + }, + { + "epoch": 1.0911278195488723, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 23140 + }, + { + "epoch": 1.091187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 23141 + }, + { + "epoch": 1.0912481203007518, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 23142 + }, + { + "epoch": 1.0913082706766917, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 23143 + }, + { + "epoch": 1.0913684210526315, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 23144 + }, + { + "epoch": 1.0914285714285714, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 23145 + }, + { + "epoch": 1.0914887218045113, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 23146 + }, + { + "epoch": 1.0915488721804512, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 23147 + }, + { + "epoch": 1.091609022556391, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 23148 + }, + { + "epoch": 1.091669172932331, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 23149 + }, + { + "epoch": 1.0917293233082708, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 23150 + }, + { + "epoch": 1.0917894736842104, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 23151 + }, + { + "epoch": 1.0918496240601503, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 23152 + }, + { + "epoch": 1.0919097744360902, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 23153 + }, + { + "epoch": 1.09196992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 23154 + }, + { + "epoch": 1.09203007518797, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 23155 + }, + { + "epoch": 1.0920902255639098, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 23156 + }, + { + "epoch": 1.0921503759398496, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 23157 + }, + { + "epoch": 1.0922105263157895, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 23158 + }, + { + "epoch": 1.0922706766917294, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 23159 + }, + { + "epoch": 1.0923308270676693, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 23160 + }, + { + "epoch": 1.0923909774436091, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 23161 + }, + { + "epoch": 1.0924511278195488, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 23162 + }, + { + "epoch": 1.0925112781954887, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 23163 + }, + { + "epoch": 1.0925714285714285, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 23164 + }, + { + "epoch": 1.0926315789473684, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 23165 + }, + { + "epoch": 1.0926917293233083, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 23166 + }, + { + "epoch": 1.0927518796992481, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 23167 + }, + { + "epoch": 1.092812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 23168 + }, + { + "epoch": 1.0928721804511279, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 23169 + }, + { + "epoch": 1.0929323308270678, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 23170 + }, + { + "epoch": 1.0929924812030074, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 23171 + }, + { + "epoch": 1.0930526315789473, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 23172 + }, + { + "epoch": 1.0931127819548871, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 23173 + }, + { + "epoch": 1.093172932330827, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 23174 + }, + { + "epoch": 1.093233082706767, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 23175 + }, + { + "epoch": 1.0932932330827068, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 23176 + }, + { + "epoch": 1.0933533834586466, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 23177 + }, + { + "epoch": 1.0934135338345865, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 23178 + }, + { + "epoch": 1.0934736842105264, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 23179 + }, + { + "epoch": 1.0935338345864662, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 23180 + }, + { + "epoch": 1.0935939849624061, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 23181 + }, + { + "epoch": 1.0936541353383458, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2368, + "step": 23182 + }, + { + "epoch": 1.0937142857142856, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 23183 + }, + { + "epoch": 1.0937744360902255, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 23184 + }, + { + "epoch": 1.0938345864661654, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 23185 + }, + { + "epoch": 1.0938947368421053, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 23186 + }, + { + "epoch": 1.0939548872180451, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 23187 + }, + { + "epoch": 1.094015037593985, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 23188 + }, + { + "epoch": 1.0940751879699249, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1646, + "step": 23189 + }, + { + "epoch": 1.0941353383458647, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 23190 + }, + { + "epoch": 1.0941954887218046, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 23191 + }, + { + "epoch": 1.0942556390977443, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 23192 + }, + { + "epoch": 1.0943157894736841, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 23193 + }, + { + "epoch": 1.094375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 23194 + }, + { + "epoch": 1.0944360902255639, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 23195 + }, + { + "epoch": 1.0944962406015037, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 23196 + }, + { + "epoch": 1.0945563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 23197 + }, + { + "epoch": 1.0946165413533835, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 23198 + }, + { + "epoch": 1.0946766917293234, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 23199 + }, + { + "epoch": 1.0947368421052632, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 23200 + }, + { + "epoch": 1.094796992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 23201 + }, + { + "epoch": 1.0948571428571428, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 23202 + }, + { + "epoch": 1.0949172932330826, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 23203 + }, + { + "epoch": 1.0949774436090225, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 23204 + }, + { + "epoch": 1.0950375939849624, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 23205 + }, + { + "epoch": 1.0950977443609022, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 23206 + }, + { + "epoch": 1.095157894736842, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 23207 + }, + { + "epoch": 1.095218045112782, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1667, + "step": 23208 + }, + { + "epoch": 1.0952781954887219, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 23209 + }, + { + "epoch": 1.0953383458646617, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 23210 + }, + { + "epoch": 1.0953984962406016, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 23211 + }, + { + "epoch": 1.0954586466165415, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1719, + "step": 23212 + }, + { + "epoch": 1.0955187969924811, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2018, + "step": 23213 + }, + { + "epoch": 1.095578947368421, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 23214 + }, + { + "epoch": 1.0956390977443609, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 23215 + }, + { + "epoch": 1.0956992481203007, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 23216 + }, + { + "epoch": 1.0957593984962406, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 23217 + }, + { + "epoch": 1.0958195488721805, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 23218 + }, + { + "epoch": 1.0958796992481203, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 23219 + }, + { + "epoch": 1.0959398496240602, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 23220 + }, + { + "epoch": 1.096, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 23221 + }, + { + "epoch": 1.0960601503759397, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 23222 + }, + { + "epoch": 1.0961203007518796, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 23223 + }, + { + "epoch": 1.0961804511278195, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 23224 + }, + { + "epoch": 1.0962406015037593, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 23225 + }, + { + "epoch": 1.0963007518796992, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 23226 + }, + { + "epoch": 1.096360902255639, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 23227 + }, + { + "epoch": 1.096421052631579, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 23228 + }, + { + "epoch": 1.0964812030075188, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 23229 + }, + { + "epoch": 1.0965413533834587, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 23230 + }, + { + "epoch": 1.0966015037593986, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 23231 + }, + { + "epoch": 1.0966616541353384, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 23232 + }, + { + "epoch": 1.096721804511278, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 23233 + }, + { + "epoch": 1.096781954887218, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 23234 + }, + { + "epoch": 1.0968421052631578, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 23235 + }, + { + "epoch": 1.0969022556390977, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 23236 + }, + { + "epoch": 1.0969624060150376, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 23237 + }, + { + "epoch": 1.0970225563909775, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 23238 + }, + { + "epoch": 1.0970827067669173, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 23239 + }, + { + "epoch": 1.0971428571428572, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 23240 + }, + { + "epoch": 1.097203007518797, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1623, + "step": 23241 + }, + { + "epoch": 1.097263157894737, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2848, + "step": 23242 + }, + { + "epoch": 1.0973233082706766, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 23243 + }, + { + "epoch": 1.0973834586466165, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 23244 + }, + { + "epoch": 1.0974436090225563, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 23245 + }, + { + "epoch": 1.0975037593984962, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 23246 + }, + { + "epoch": 1.097563909774436, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 23247 + }, + { + "epoch": 1.097624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 23248 + }, + { + "epoch": 1.0976842105263158, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 23249 + }, + { + "epoch": 1.0977443609022557, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 23250 + }, + { + "epoch": 1.0978045112781956, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 23251 + }, + { + "epoch": 1.0978646616541354, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 23252 + }, + { + "epoch": 1.097924812030075, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 23253 + }, + { + "epoch": 1.097984962406015, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 23254 + }, + { + "epoch": 1.0980451127819548, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 23255 + }, + { + "epoch": 1.0981052631578947, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 23256 + }, + { + "epoch": 1.0981654135338346, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 23257 + }, + { + "epoch": 1.0982255639097744, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1716, + "step": 23258 + }, + { + "epoch": 1.0982857142857143, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 23259 + }, + { + "epoch": 1.0983458646616542, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 23260 + }, + { + "epoch": 1.098406015037594, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 23261 + }, + { + "epoch": 1.098466165413534, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 23262 + }, + { + "epoch": 1.0985263157894738, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 23263 + }, + { + "epoch": 1.0985864661654134, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 23264 + }, + { + "epoch": 1.0986466165413533, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 23265 + }, + { + "epoch": 1.0987067669172932, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 23266 + }, + { + "epoch": 1.098766917293233, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 23267 + }, + { + "epoch": 1.098827067669173, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 23268 + }, + { + "epoch": 1.0988872180451128, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 23269 + }, + { + "epoch": 1.0989473684210527, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 23270 + }, + { + "epoch": 1.0990075187969925, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2278, + "step": 23271 + }, + { + "epoch": 1.0990676691729324, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.2958, + "step": 23272 + }, + { + "epoch": 1.099127819548872, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 23273 + }, + { + "epoch": 1.099187969924812, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 23274 + }, + { + "epoch": 1.0992481203007518, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 23275 + }, + { + "epoch": 1.0993082706766917, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 23276 + }, + { + "epoch": 1.0993684210526316, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 23277 + }, + { + "epoch": 1.0994285714285714, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 23278 + }, + { + "epoch": 1.0994887218045113, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 23279 + }, + { + "epoch": 1.0995488721804512, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 23280 + }, + { + "epoch": 1.099609022556391, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 23281 + }, + { + "epoch": 1.099669172932331, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 23282 + }, + { + "epoch": 1.0997293233082708, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 23283 + }, + { + "epoch": 1.0997894736842104, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 23284 + }, + { + "epoch": 1.0998496240601503, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 23285 + }, + { + "epoch": 1.0999097744360902, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 23286 + }, + { + "epoch": 1.09996992481203, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 23287 + }, + { + "epoch": 1.10003007518797, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1429, + "step": 23288 + }, + { + "epoch": 1.1000902255639098, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 23289 + }, + { + "epoch": 1.1001503759398497, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 23290 + }, + { + "epoch": 1.1002105263157895, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 23291 + }, + { + "epoch": 1.1002706766917294, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 23292 + }, + { + "epoch": 1.1003308270676693, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1585, + "step": 23293 + }, + { + "epoch": 1.100390977443609, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 23294 + }, + { + "epoch": 1.1004511278195488, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 23295 + }, + { + "epoch": 1.1005112781954887, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 23296 + }, + { + "epoch": 1.1005714285714285, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 23297 + }, + { + "epoch": 1.1006315789473684, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 23298 + }, + { + "epoch": 1.1006917293233083, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 23299 + }, + { + "epoch": 1.1007518796992481, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 23300 + }, + { + "epoch": 1.100812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 23301 + }, + { + "epoch": 1.100872180451128, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 23302 + }, + { + "epoch": 1.1009323308270678, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1446, + "step": 23303 + }, + { + "epoch": 1.1009924812030074, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 23304 + }, + { + "epoch": 1.1010526315789473, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 23305 + }, + { + "epoch": 1.1011127819548872, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 23306 + }, + { + "epoch": 1.101172932330827, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 23307 + }, + { + "epoch": 1.101233082706767, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 23308 + }, + { + "epoch": 1.1012932330827068, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 23309 + }, + { + "epoch": 1.1013533834586466, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 23310 + }, + { + "epoch": 1.1014135338345865, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 23311 + }, + { + "epoch": 1.1014736842105264, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 23312 + }, + { + "epoch": 1.1015338345864663, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 23313 + }, + { + "epoch": 1.1015939849624061, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 23314 + }, + { + "epoch": 1.1016541353383458, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 23315 + }, + { + "epoch": 1.1017142857142856, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 23316 + }, + { + "epoch": 1.1017744360902255, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 23317 + }, + { + "epoch": 1.1018345864661654, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 23318 + }, + { + "epoch": 1.1018947368421053, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 23319 + }, + { + "epoch": 1.1019548872180451, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 23320 + }, + { + "epoch": 1.102015037593985, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 23321 + }, + { + "epoch": 1.1020751879699249, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 23322 + }, + { + "epoch": 1.1021353383458647, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 23323 + }, + { + "epoch": 1.1021954887218044, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 23324 + }, + { + "epoch": 1.1022556390977443, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 23325 + }, + { + "epoch": 1.1023157894736841, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 23326 + }, + { + "epoch": 1.102375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 23327 + }, + { + "epoch": 1.1024360902255639, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 23328 + }, + { + "epoch": 1.1024962406015038, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 23329 + }, + { + "epoch": 1.1025563909774436, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2152, + "step": 23330 + }, + { + "epoch": 1.1026165413533835, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 23331 + }, + { + "epoch": 1.1026766917293234, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 23332 + }, + { + "epoch": 1.1027368421052632, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 23333 + }, + { + "epoch": 1.102796992481203, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 23334 + }, + { + "epoch": 1.1028571428571428, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 23335 + }, + { + "epoch": 1.1029172932330826, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1636, + "step": 23336 + }, + { + "epoch": 1.1029774436090225, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 23337 + }, + { + "epoch": 1.1030375939849624, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 23338 + }, + { + "epoch": 1.1030977443609022, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2258, + "step": 23339 + }, + { + "epoch": 1.1031578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 23340 + }, + { + "epoch": 1.103218045112782, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 23341 + }, + { + "epoch": 1.1032781954887219, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.3322, + "step": 23342 + }, + { + "epoch": 1.1033383458646617, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 23343 + }, + { + "epoch": 1.1033984962406016, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 23344 + }, + { + "epoch": 1.1034586466165412, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 23345 + }, + { + "epoch": 1.1035187969924811, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 23346 + }, + { + "epoch": 1.103578947368421, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 23347 + }, + { + "epoch": 1.1036390977443609, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 23348 + }, + { + "epoch": 1.1036992481203007, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 23349 + }, + { + "epoch": 1.1037593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 23350 + }, + { + "epoch": 1.1038195488721805, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 23351 + }, + { + "epoch": 1.1038796992481203, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 23352 + }, + { + "epoch": 1.1039398496240602, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 23353 + }, + { + "epoch": 1.104, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 23354 + }, + { + "epoch": 1.1040601503759397, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 23355 + }, + { + "epoch": 1.1041203007518796, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 23356 + }, + { + "epoch": 1.1041804511278195, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 23357 + }, + { + "epoch": 1.1042406015037594, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 23358 + }, + { + "epoch": 1.1043007518796992, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 23359 + }, + { + "epoch": 1.104360902255639, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 23360 + }, + { + "epoch": 1.104421052631579, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 23361 + }, + { + "epoch": 1.1044812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 23362 + }, + { + "epoch": 1.1045413533834587, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 23363 + }, + { + "epoch": 1.1046015037593986, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 23364 + }, + { + "epoch": 1.1046616541353385, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 23365 + }, + { + "epoch": 1.104721804511278, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 23366 + }, + { + "epoch": 1.104781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 23367 + }, + { + "epoch": 1.1048421052631578, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 23368 + }, + { + "epoch": 1.1049022556390977, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 23369 + }, + { + "epoch": 1.1049624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 23370 + }, + { + "epoch": 1.1050225563909775, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 23371 + }, + { + "epoch": 1.1050827067669173, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1747, + "step": 23372 + }, + { + "epoch": 1.1051428571428572, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 23373 + }, + { + "epoch": 1.105203007518797, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 23374 + }, + { + "epoch": 1.1052631578947367, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 23375 + }, + { + "epoch": 1.1053233082706766, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 23376 + }, + { + "epoch": 1.1053834586466165, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 23377 + }, + { + "epoch": 1.1054436090225563, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 23378 + }, + { + "epoch": 1.1055037593984962, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 23379 + }, + { + "epoch": 1.105563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 23380 + }, + { + "epoch": 1.105624060150376, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 23381 + }, + { + "epoch": 1.1056842105263158, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 23382 + }, + { + "epoch": 1.1057443609022557, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 23383 + }, + { + "epoch": 1.1058045112781956, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 23384 + }, + { + "epoch": 1.1058646616541354, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 23385 + }, + { + "epoch": 1.105924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 23386 + }, + { + "epoch": 1.105984962406015, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 23387 + }, + { + "epoch": 1.1060451127819548, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 23388 + }, + { + "epoch": 1.1061052631578947, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 23389 + }, + { + "epoch": 1.1061654135338346, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 23390 + }, + { + "epoch": 1.1062255639097744, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 23391 + }, + { + "epoch": 1.1062857142857143, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 23392 + }, + { + "epoch": 1.1063458646616542, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 23393 + }, + { + "epoch": 1.106406015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 23394 + }, + { + "epoch": 1.106466165413534, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 23395 + }, + { + "epoch": 1.1065263157894736, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1708, + "step": 23396 + }, + { + "epoch": 1.1065864661654135, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 23397 + }, + { + "epoch": 1.1066466165413533, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1932, + "step": 23398 + }, + { + "epoch": 1.1067067669172932, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 23399 + }, + { + "epoch": 1.106766917293233, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 23400 + }, + { + "epoch": 1.106827067669173, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 23401 + }, + { + "epoch": 1.1068872180451128, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 23402 + }, + { + "epoch": 1.1069473684210527, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 23403 + }, + { + "epoch": 1.1070075187969926, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 23404 + }, + { + "epoch": 1.1070676691729324, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 23405 + }, + { + "epoch": 1.107127819548872, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 23406 + }, + { + "epoch": 1.107187969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2708, + "step": 23407 + }, + { + "epoch": 1.1072481203007518, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 23408 + }, + { + "epoch": 1.1073082706766917, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 23409 + }, + { + "epoch": 1.1073684210526316, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 23410 + }, + { + "epoch": 1.1074285714285714, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 23411 + }, + { + "epoch": 1.1074887218045113, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 23412 + }, + { + "epoch": 1.1075488721804512, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 23413 + }, + { + "epoch": 1.107609022556391, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 23414 + }, + { + "epoch": 1.107669172932331, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 23415 + }, + { + "epoch": 1.1077293233082708, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 23416 + }, + { + "epoch": 1.1077894736842104, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 23417 + }, + { + "epoch": 1.1078496240601503, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 23418 + }, + { + "epoch": 1.1079097744360902, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 23419 + }, + { + "epoch": 1.10796992481203, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 23420 + }, + { + "epoch": 1.10803007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 23421 + }, + { + "epoch": 1.1080902255639098, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 23422 + }, + { + "epoch": 1.1081503759398497, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 23423 + }, + { + "epoch": 1.1082105263157895, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 23424 + }, + { + "epoch": 1.1082706766917294, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 23425 + }, + { + "epoch": 1.108330827067669, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 23426 + }, + { + "epoch": 1.108390977443609, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 23427 + }, + { + "epoch": 1.1084511278195488, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 23428 + }, + { + "epoch": 1.1085112781954887, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 23429 + }, + { + "epoch": 1.1085714285714285, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 23430 + }, + { + "epoch": 1.1086315789473684, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 23431 + }, + { + "epoch": 1.1086917293233083, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 23432 + }, + { + "epoch": 1.1087518796992482, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1758, + "step": 23433 + }, + { + "epoch": 1.108812030075188, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 23434 + }, + { + "epoch": 1.108872180451128, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 23435 + }, + { + "epoch": 1.1089323308270678, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 23436 + }, + { + "epoch": 1.1089924812030074, + "grad_norm": 1.421875, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 23437 + }, + { + "epoch": 1.1090526315789473, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 23438 + }, + { + "epoch": 1.1091127819548872, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 23439 + }, + { + "epoch": 1.109172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 23440 + }, + { + "epoch": 1.109233082706767, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 23441 + }, + { + "epoch": 1.1092932330827068, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 23442 + }, + { + "epoch": 1.1093533834586466, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 23443 + }, + { + "epoch": 1.1094135338345865, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 23444 + }, + { + "epoch": 1.1094736842105264, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 23445 + }, + { + "epoch": 1.1095338345864663, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 23446 + }, + { + "epoch": 1.1095939849624061, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 23447 + }, + { + "epoch": 1.1096541353383458, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2917, + "step": 23448 + }, + { + "epoch": 1.1097142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 23449 + }, + { + "epoch": 1.1097744360902255, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 23450 + }, + { + "epoch": 1.1098345864661654, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 23451 + }, + { + "epoch": 1.1098947368421053, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 23452 + }, + { + "epoch": 1.1099548872180451, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 23453 + }, + { + "epoch": 1.110015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 23454 + }, + { + "epoch": 1.1100751879699249, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1254, + "step": 23455 + }, + { + "epoch": 1.1101353383458648, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 23456 + }, + { + "epoch": 1.1101954887218044, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 23457 + }, + { + "epoch": 1.1102556390977443, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 23458 + }, + { + "epoch": 1.1103157894736841, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 23459 + }, + { + "epoch": 1.110375939849624, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 23460 + }, + { + "epoch": 1.1104360902255639, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 23461 + }, + { + "epoch": 1.1104962406015038, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 23462 + }, + { + "epoch": 1.1105563909774436, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 23463 + }, + { + "epoch": 1.1106165413533835, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 23464 + }, + { + "epoch": 1.1106766917293234, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 23465 + }, + { + "epoch": 1.1107368421052632, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 23466 + }, + { + "epoch": 1.1107969924812031, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 23467 + }, + { + "epoch": 1.1108571428571428, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 23468 + }, + { + "epoch": 1.1109172932330826, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 23469 + }, + { + "epoch": 1.1109774436090225, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1962, + "step": 23470 + }, + { + "epoch": 1.1110375939849624, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 23471 + }, + { + "epoch": 1.1110977443609023, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 23472 + }, + { + "epoch": 1.1111578947368421, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 23473 + }, + { + "epoch": 1.111218045112782, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 23474 + }, + { + "epoch": 1.1112781954887219, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 23475 + }, + { + "epoch": 1.1113383458646617, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 23476 + }, + { + "epoch": 1.1113984962406016, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 23477 + }, + { + "epoch": 1.1114586466165413, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9217, + "step": 23478 + }, + { + "epoch": 1.1115187969924811, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 23479 + }, + { + "epoch": 1.111578947368421, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 23480 + }, + { + "epoch": 1.1116390977443609, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2103, + "step": 23481 + }, + { + "epoch": 1.1116992481203007, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 23482 + }, + { + "epoch": 1.1117593984962406, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 23483 + }, + { + "epoch": 1.1118195488721805, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2068, + "step": 23484 + }, + { + "epoch": 1.1118796992481204, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1864, + "step": 23485 + }, + { + "epoch": 1.1119398496240602, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 23486 + }, + { + "epoch": 1.112, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 23487 + }, + { + "epoch": 1.1120601503759397, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 23488 + }, + { + "epoch": 1.1121203007518796, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 23489 + }, + { + "epoch": 1.1121804511278195, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 23490 + }, + { + "epoch": 1.1122406015037594, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 23491 + }, + { + "epoch": 1.1123007518796992, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 23492 + }, + { + "epoch": 1.112360902255639, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2718, + "step": 23493 + }, + { + "epoch": 1.112421052631579, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1177, + "step": 23494 + }, + { + "epoch": 1.1124812030075188, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 23495 + }, + { + "epoch": 1.1125413533834587, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 23496 + }, + { + "epoch": 1.1126015037593986, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 23497 + }, + { + "epoch": 1.1126616541353385, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 23498 + }, + { + "epoch": 1.1127218045112781, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 23499 + }, + { + "epoch": 1.112781954887218, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 23500 + }, + { + "epoch": 1.1128421052631579, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 23501 + }, + { + "epoch": 1.1129022556390977, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 23502 + }, + { + "epoch": 1.1129624060150376, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 23503 + }, + { + "epoch": 1.1130225563909775, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 23504 + }, + { + "epoch": 1.1130827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 23505 + }, + { + "epoch": 1.1131428571428572, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 23506 + }, + { + "epoch": 1.113203007518797, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 23507 + }, + { + "epoch": 1.1132631578947367, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 23508 + }, + { + "epoch": 1.1133233082706766, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1773, + "step": 23509 + }, + { + "epoch": 1.1133834586466165, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 23510 + }, + { + "epoch": 1.1134436090225563, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 23511 + }, + { + "epoch": 1.1135037593984962, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 23512 + }, + { + "epoch": 1.113563909774436, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 23513 + }, + { + "epoch": 1.113624060150376, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 23514 + }, + { + "epoch": 1.1136842105263158, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 23515 + }, + { + "epoch": 1.1137443609022557, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 23516 + }, + { + "epoch": 1.1138045112781956, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 23517 + }, + { + "epoch": 1.1138646616541354, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 23518 + }, + { + "epoch": 1.113924812030075, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 23519 + }, + { + "epoch": 1.113984962406015, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 23520 + }, + { + "epoch": 1.1140451127819548, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 23521 + }, + { + "epoch": 1.1141052631578947, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 23522 + }, + { + "epoch": 1.1141654135338346, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 23523 + }, + { + "epoch": 1.1142255639097745, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 23524 + }, + { + "epoch": 1.1142857142857143, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 23525 + }, + { + "epoch": 1.1143458646616542, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 23526 + }, + { + "epoch": 1.114406015037594, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 23527 + }, + { + "epoch": 1.114466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 23528 + }, + { + "epoch": 1.1145263157894736, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.3095, + "step": 23529 + }, + { + "epoch": 1.1145864661654135, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 23530 + }, + { + "epoch": 1.1146466165413533, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 23531 + }, + { + "epoch": 1.1147067669172932, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 23532 + }, + { + "epoch": 1.114766917293233, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 23533 + }, + { + "epoch": 1.114827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 23534 + }, + { + "epoch": 1.1148872180451128, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 23535 + }, + { + "epoch": 1.1149473684210527, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 23536 + }, + { + "epoch": 1.1150075187969926, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 23537 + }, + { + "epoch": 1.1150676691729324, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 23538 + }, + { + "epoch": 1.115127819548872, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 23539 + }, + { + "epoch": 1.115187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 23540 + }, + { + "epoch": 1.1152481203007518, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 23541 + }, + { + "epoch": 1.1153082706766917, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 23542 + }, + { + "epoch": 1.1153684210526316, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 23543 + }, + { + "epoch": 1.1154285714285714, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2179, + "step": 23544 + }, + { + "epoch": 1.1154887218045113, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 23545 + }, + { + "epoch": 1.1155488721804512, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 23546 + }, + { + "epoch": 1.115609022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 23547 + }, + { + "epoch": 1.115669172932331, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 23548 + }, + { + "epoch": 1.1157293233082708, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 23549 + }, + { + "epoch": 1.1157894736842104, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 23550 + }, + { + "epoch": 1.1158496240601503, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 23551 + }, + { + "epoch": 1.1159097744360902, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 23552 + }, + { + "epoch": 1.11596992481203, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 23553 + }, + { + "epoch": 1.11603007518797, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 23554 + }, + { + "epoch": 1.1160902255639098, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 23555 + }, + { + "epoch": 1.1161503759398497, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 23556 + }, + { + "epoch": 1.1162105263157895, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 23557 + }, + { + "epoch": 1.1162706766917294, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 23558 + }, + { + "epoch": 1.116330827067669, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 23559 + }, + { + "epoch": 1.116390977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 23560 + }, + { + "epoch": 1.1164511278195488, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 23561 + }, + { + "epoch": 1.1165112781954887, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 23562 + }, + { + "epoch": 1.1165714285714285, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 23563 + }, + { + "epoch": 1.1166315789473684, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 23564 + }, + { + "epoch": 1.1166917293233083, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 23565 + }, + { + "epoch": 1.1167518796992482, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 23566 + }, + { + "epoch": 1.116812030075188, + "grad_norm": 0.455078125, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 23567 + }, + { + "epoch": 1.116872180451128, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 23568 + }, + { + "epoch": 1.1169323308270678, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 23569 + }, + { + "epoch": 1.1169924812030074, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 23570 + }, + { + "epoch": 1.1170526315789473, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 23571 + }, + { + "epoch": 1.1171127819548872, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 23572 + }, + { + "epoch": 1.117172932330827, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 23573 + }, + { + "epoch": 1.117233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 23574 + }, + { + "epoch": 1.1172932330827068, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 23575 + }, + { + "epoch": 1.1173533834586467, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 23576 + }, + { + "epoch": 1.1174135338345865, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 23577 + }, + { + "epoch": 1.1174736842105264, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 23578 + }, + { + "epoch": 1.1175338345864663, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 23579 + }, + { + "epoch": 1.117593984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 23580 + }, + { + "epoch": 1.1176541353383458, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 23581 + }, + { + "epoch": 1.1177142857142857, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 23582 + }, + { + "epoch": 1.1177744360902255, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 23583 + }, + { + "epoch": 1.1178345864661654, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 23584 + }, + { + "epoch": 1.1178947368421053, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1614, + "step": 23585 + }, + { + "epoch": 1.1179548872180451, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 23586 + }, + { + "epoch": 1.118015037593985, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 23587 + }, + { + "epoch": 1.1180751879699249, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 23588 + }, + { + "epoch": 1.1181353383458648, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1809, + "step": 23589 + }, + { + "epoch": 1.1181954887218044, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 23590 + }, + { + "epoch": 1.1182556390977443, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 23591 + }, + { + "epoch": 1.1183157894736842, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 23592 + }, + { + "epoch": 1.118375939849624, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 23593 + }, + { + "epoch": 1.118436090225564, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 23594 + }, + { + "epoch": 1.1184962406015038, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 23595 + }, + { + "epoch": 1.1185563909774436, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 23596 + }, + { + "epoch": 1.1186165413533835, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 23597 + }, + { + "epoch": 1.1186766917293234, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 23598 + }, + { + "epoch": 1.1187368421052633, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 23599 + }, + { + "epoch": 1.1187969924812031, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 23600 + }, + { + "epoch": 1.1188571428571428, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 23601 + }, + { + "epoch": 1.1189172932330826, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 23602 + }, + { + "epoch": 1.1189774436090225, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 23603 + }, + { + "epoch": 1.1190375939849624, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 23604 + }, + { + "epoch": 1.1190977443609023, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 23605 + }, + { + "epoch": 1.1191578947368421, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 23606 + }, + { + "epoch": 1.119218045112782, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 23607 + }, + { + "epoch": 1.1192781954887219, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 23608 + }, + { + "epoch": 1.1193383458646617, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 23609 + }, + { + "epoch": 1.1193984962406014, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 23610 + }, + { + "epoch": 1.1194586466165413, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 23611 + }, + { + "epoch": 1.1195187969924811, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9393, + "step": 23612 + }, + { + "epoch": 1.119578947368421, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 23613 + }, + { + "epoch": 1.1196390977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 23614 + }, + { + "epoch": 1.1196992481203007, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 23615 + }, + { + "epoch": 1.1197593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 23616 + }, + { + "epoch": 1.1198195488721805, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 23617 + }, + { + "epoch": 1.1198796992481204, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 23618 + }, + { + "epoch": 1.1199398496240602, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 23619 + }, + { + "epoch": 1.12, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 23620 + }, + { + "epoch": 1.1200601503759398, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 23621 + }, + { + "epoch": 1.1201203007518796, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 23622 + }, + { + "epoch": 1.1201804511278195, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 23623 + }, + { + "epoch": 1.1202406015037594, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 23624 + }, + { + "epoch": 1.1203007518796992, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 23625 + }, + { + "epoch": 1.1203609022556391, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 23626 + }, + { + "epoch": 1.120421052631579, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 23627 + }, + { + "epoch": 1.1204812030075189, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9071, + "step": 23628 + }, + { + "epoch": 1.1205413533834587, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 23629 + }, + { + "epoch": 1.1206015037593986, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 23630 + }, + { + "epoch": 1.1206616541353382, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 23631 + }, + { + "epoch": 1.1207218045112781, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 23632 + }, + { + "epoch": 1.120781954887218, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 23633 + }, + { + "epoch": 1.1208421052631579, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 23634 + }, + { + "epoch": 1.1209022556390977, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 23635 + }, + { + "epoch": 1.1209624060150376, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 23636 + }, + { + "epoch": 1.1210225563909775, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 23637 + }, + { + "epoch": 1.1210827067669173, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 23638 + }, + { + "epoch": 1.1211428571428572, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 23639 + }, + { + "epoch": 1.121203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 23640 + }, + { + "epoch": 1.1212631578947367, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2034, + "step": 23641 + }, + { + "epoch": 1.1213233082706766, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 23642 + }, + { + "epoch": 1.1213834586466165, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 23643 + }, + { + "epoch": 1.1214436090225564, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 23644 + }, + { + "epoch": 1.1215037593984962, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 23645 + }, + { + "epoch": 1.121563909774436, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 23646 + }, + { + "epoch": 1.121624060150376, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 23647 + }, + { + "epoch": 1.1216842105263158, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 23648 + }, + { + "epoch": 1.1217443609022557, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 23649 + }, + { + "epoch": 1.1218045112781956, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 23650 + }, + { + "epoch": 1.1218646616541355, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 23651 + }, + { + "epoch": 1.121924812030075, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 23652 + }, + { + "epoch": 1.121984962406015, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 23653 + }, + { + "epoch": 1.1220451127819548, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 23654 + }, + { + "epoch": 1.1221052631578947, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 23655 + }, + { + "epoch": 1.1221654135338346, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 23656 + }, + { + "epoch": 1.1222255639097745, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 23657 + }, + { + "epoch": 1.1222857142857143, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 23658 + }, + { + "epoch": 1.1223458646616542, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2318, + "step": 23659 + }, + { + "epoch": 1.122406015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 23660 + }, + { + "epoch": 1.1224661654135337, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 23661 + }, + { + "epoch": 1.1225263157894736, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 23662 + }, + { + "epoch": 1.1225864661654135, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 23663 + }, + { + "epoch": 1.1226466165413533, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 23664 + }, + { + "epoch": 1.1227067669172932, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 23665 + }, + { + "epoch": 1.122766917293233, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 23666 + }, + { + "epoch": 1.122827067669173, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 23667 + }, + { + "epoch": 1.1228872180451128, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 23668 + }, + { + "epoch": 1.1229473684210527, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 23669 + }, + { + "epoch": 1.1230075187969926, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 23670 + }, + { + "epoch": 1.1230676691729324, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 23671 + }, + { + "epoch": 1.123127819548872, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 23672 + }, + { + "epoch": 1.123187969924812, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 23673 + }, + { + "epoch": 1.1232481203007518, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 23674 + }, + { + "epoch": 1.1233082706766917, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 23675 + }, + { + "epoch": 1.1233684210526316, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 23676 + }, + { + "epoch": 1.1234285714285714, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 23677 + }, + { + "epoch": 1.1234887218045113, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 23678 + }, + { + "epoch": 1.1235488721804512, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 23679 + }, + { + "epoch": 1.123609022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 23680 + }, + { + "epoch": 1.123669172932331, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 23681 + }, + { + "epoch": 1.1237293233082706, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 23682 + }, + { + "epoch": 1.1237894736842104, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 23683 + }, + { + "epoch": 1.1238496240601503, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 23684 + }, + { + "epoch": 1.1239097744360902, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 23685 + }, + { + "epoch": 1.12396992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 23686 + }, + { + "epoch": 1.12403007518797, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 23687 + }, + { + "epoch": 1.1240902255639098, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.147, + "step": 23688 + }, + { + "epoch": 1.1241503759398497, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 23689 + }, + { + "epoch": 1.1242105263157895, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 23690 + }, + { + "epoch": 1.1242706766917294, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 23691 + }, + { + "epoch": 1.124330827067669, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 23692 + }, + { + "epoch": 1.124390977443609, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 23693 + }, + { + "epoch": 1.1244511278195488, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 23694 + }, + { + "epoch": 1.1245112781954887, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 23695 + }, + { + "epoch": 1.1245714285714286, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 23696 + }, + { + "epoch": 1.1246315789473684, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 23697 + }, + { + "epoch": 1.1246917293233083, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 23698 + }, + { + "epoch": 1.1247518796992482, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 23699 + }, + { + "epoch": 1.124812030075188, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 23700 + }, + { + "epoch": 1.124872180451128, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 23701 + }, + { + "epoch": 1.1249323308270678, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 23702 + }, + { + "epoch": 1.1249924812030074, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 23703 + }, + { + "epoch": 1.1250526315789473, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 23704 + }, + { + "epoch": 1.1251127819548872, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 23705 + }, + { + "epoch": 1.125172932330827, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 23706 + }, + { + "epoch": 1.125233082706767, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 23707 + }, + { + "epoch": 1.1252932330827068, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 23708 + }, + { + "epoch": 1.1253533834586467, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 23709 + }, + { + "epoch": 1.1254135338345865, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 23710 + }, + { + "epoch": 1.1254736842105264, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 23711 + }, + { + "epoch": 1.125533834586466, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2351, + "step": 23712 + }, + { + "epoch": 1.125593984962406, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 23713 + }, + { + "epoch": 1.1256541353383458, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 23714 + }, + { + "epoch": 1.1257142857142857, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 23715 + }, + { + "epoch": 1.1257744360902255, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 23716 + }, + { + "epoch": 1.1258345864661654, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1667, + "step": 23717 + }, + { + "epoch": 1.1258947368421053, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2388, + "step": 23718 + }, + { + "epoch": 1.1259548872180452, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 23719 + }, + { + "epoch": 1.126015037593985, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 23720 + }, + { + "epoch": 1.126075187969925, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 23721 + }, + { + "epoch": 1.1261353383458648, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 23722 + }, + { + "epoch": 1.1261954887218044, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 23723 + }, + { + "epoch": 1.1262556390977443, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 23724 + }, + { + "epoch": 1.1263157894736842, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 23725 + }, + { + "epoch": 1.126375939849624, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 23726 + }, + { + "epoch": 1.126436090225564, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 23727 + }, + { + "epoch": 1.1264962406015038, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 23728 + }, + { + "epoch": 1.1265563909774436, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 23729 + }, + { + "epoch": 1.1266165413533835, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 23730 + }, + { + "epoch": 1.1266766917293234, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1668, + "step": 23731 + }, + { + "epoch": 1.126736842105263, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 23732 + }, + { + "epoch": 1.1267969924812031, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 23733 + }, + { + "epoch": 1.1268571428571428, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 23734 + }, + { + "epoch": 1.1269172932330827, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 23735 + }, + { + "epoch": 1.1269774436090225, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 23736 + }, + { + "epoch": 1.1270375939849624, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 23737 + }, + { + "epoch": 1.1270977443609023, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 23738 + }, + { + "epoch": 1.1271578947368421, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 23739 + }, + { + "epoch": 1.127218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 23740 + }, + { + "epoch": 1.1272781954887219, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 23741 + }, + { + "epoch": 1.1273383458646618, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.19, + "step": 23742 + }, + { + "epoch": 1.1273984962406014, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 23743 + }, + { + "epoch": 1.1274586466165413, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 23744 + }, + { + "epoch": 1.1275187969924811, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 23745 + }, + { + "epoch": 1.127578947368421, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 23746 + }, + { + "epoch": 1.1276390977443609, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 23747 + }, + { + "epoch": 1.1276992481203008, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 23748 + }, + { + "epoch": 1.1277593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 23749 + }, + { + "epoch": 1.1278195488721805, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 23750 + }, + { + "epoch": 1.1278796992481204, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 23751 + }, + { + "epoch": 1.1279398496240602, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 23752 + }, + { + "epoch": 1.1280000000000001, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 23753 + }, + { + "epoch": 1.1280601503759398, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 23754 + }, + { + "epoch": 1.1281203007518796, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 23755 + }, + { + "epoch": 1.1281804511278195, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 23756 + }, + { + "epoch": 1.1282406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 23757 + }, + { + "epoch": 1.1283007518796992, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 23758 + }, + { + "epoch": 1.1283609022556391, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 23759 + }, + { + "epoch": 1.128421052631579, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 23760 + }, + { + "epoch": 1.1284812030075189, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 23761 + }, + { + "epoch": 1.1285413533834587, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 23762 + }, + { + "epoch": 1.1286015037593984, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 23763 + }, + { + "epoch": 1.1286616541353383, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 23764 + }, + { + "epoch": 1.1287218045112781, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 23765 + }, + { + "epoch": 1.128781954887218, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 23766 + }, + { + "epoch": 1.1288421052631579, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 23767 + }, + { + "epoch": 1.1289022556390977, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 23768 + }, + { + "epoch": 1.1289624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 23769 + }, + { + "epoch": 1.1290225563909775, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 23770 + }, + { + "epoch": 1.1290827067669174, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 23771 + }, + { + "epoch": 1.1291428571428572, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 23772 + }, + { + "epoch": 1.129203007518797, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 23773 + }, + { + "epoch": 1.1292631578947367, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 23774 + }, + { + "epoch": 1.1293233082706766, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 23775 + }, + { + "epoch": 1.1293834586466165, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 23776 + }, + { + "epoch": 1.1294436090225564, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 23777 + }, + { + "epoch": 1.1295037593984962, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 23778 + }, + { + "epoch": 1.129563909774436, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 23779 + }, + { + "epoch": 1.129624060150376, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 23780 + }, + { + "epoch": 1.1296842105263158, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 23781 + }, + { + "epoch": 1.1297443609022557, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 23782 + }, + { + "epoch": 1.1298045112781954, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 23783 + }, + { + "epoch": 1.1298646616541355, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 23784 + }, + { + "epoch": 1.129924812030075, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 23785 + }, + { + "epoch": 1.129984962406015, + "grad_norm": 1.40625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 23786 + }, + { + "epoch": 1.1300451127819549, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 23787 + }, + { + "epoch": 1.1301052631578947, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 23788 + }, + { + "epoch": 1.1301654135338346, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 23789 + }, + { + "epoch": 1.1302255639097745, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 23790 + }, + { + "epoch": 1.1302857142857143, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 23791 + }, + { + "epoch": 1.1303458646616542, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 23792 + }, + { + "epoch": 1.130406015037594, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 23793 + }, + { + "epoch": 1.1304661654135337, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 23794 + }, + { + "epoch": 1.1305263157894736, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 23795 + }, + { + "epoch": 1.1305864661654135, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 23796 + }, + { + "epoch": 1.1306466165413533, + "grad_norm": 1.3984375, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 23797 + }, + { + "epoch": 1.1307067669172932, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 23798 + }, + { + "epoch": 1.130766917293233, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 23799 + }, + { + "epoch": 1.130827067669173, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 23800 + }, + { + "epoch": 1.1308872180451128, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 23801 + }, + { + "epoch": 1.1309473684210527, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 23802 + }, + { + "epoch": 1.1310075187969926, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 23803 + }, + { + "epoch": 1.1310676691729324, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 23804 + }, + { + "epoch": 1.131127819548872, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 23805 + }, + { + "epoch": 1.131187969924812, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 23806 + }, + { + "epoch": 1.1312481203007518, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 23807 + }, + { + "epoch": 1.1313082706766917, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 23808 + }, + { + "epoch": 1.1313684210526316, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 23809 + }, + { + "epoch": 1.1314285714285715, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 23810 + }, + { + "epoch": 1.1314887218045113, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 23811 + }, + { + "epoch": 1.1315488721804512, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 23812 + }, + { + "epoch": 1.131609022556391, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 23813 + }, + { + "epoch": 1.1316691729323307, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 23814 + }, + { + "epoch": 1.1317293233082706, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 23815 + }, + { + "epoch": 1.1317894736842105, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 23816 + }, + { + "epoch": 1.1318496240601503, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 23817 + }, + { + "epoch": 1.1319097744360902, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 23818 + }, + { + "epoch": 1.13196992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 23819 + }, + { + "epoch": 1.13203007518797, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 23820 + }, + { + "epoch": 1.1320902255639098, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 23821 + }, + { + "epoch": 1.1321503759398497, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 23822 + }, + { + "epoch": 1.1322105263157896, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 23823 + }, + { + "epoch": 1.1322706766917294, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 23824 + }, + { + "epoch": 1.132330827067669, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1934, + "step": 23825 + }, + { + "epoch": 1.132390977443609, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2211, + "step": 23826 + }, + { + "epoch": 1.1324511278195488, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 23827 + }, + { + "epoch": 1.1325112781954887, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 23828 + }, + { + "epoch": 1.1325714285714286, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 23829 + }, + { + "epoch": 1.1326315789473684, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 23830 + }, + { + "epoch": 1.1326917293233083, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 23831 + }, + { + "epoch": 1.1327518796992482, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 23832 + }, + { + "epoch": 1.132812030075188, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 23833 + }, + { + "epoch": 1.1328721804511277, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 23834 + }, + { + "epoch": 1.1329323308270678, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 23835 + }, + { + "epoch": 1.1329924812030074, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 23836 + }, + { + "epoch": 1.1330526315789473, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 23837 + }, + { + "epoch": 1.1331127819548872, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 23838 + }, + { + "epoch": 1.133172932330827, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 23839 + }, + { + "epoch": 1.133233082706767, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 23840 + }, + { + "epoch": 1.1332932330827068, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 23841 + }, + { + "epoch": 1.1333533834586467, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 23842 + }, + { + "epoch": 1.1334135338345865, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 23843 + }, + { + "epoch": 1.1334736842105264, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 23844 + }, + { + "epoch": 1.133533834586466, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 23845 + }, + { + "epoch": 1.133593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 23846 + }, + { + "epoch": 1.1336541353383458, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 23847 + }, + { + "epoch": 1.1337142857142857, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 23848 + }, + { + "epoch": 1.1337744360902255, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 23849 + }, + { + "epoch": 1.1338345864661654, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 23850 + }, + { + "epoch": 1.1338947368421053, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 23851 + }, + { + "epoch": 1.1339548872180452, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 23852 + }, + { + "epoch": 1.134015037593985, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 23853 + }, + { + "epoch": 1.134075187969925, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 23854 + }, + { + "epoch": 1.1341353383458648, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 23855 + }, + { + "epoch": 1.1341954887218044, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 23856 + }, + { + "epoch": 1.1342556390977443, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2842, + "step": 23857 + }, + { + "epoch": 1.1343157894736842, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 23858 + }, + { + "epoch": 1.134375939849624, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 23859 + }, + { + "epoch": 1.134436090225564, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 23860 + }, + { + "epoch": 1.1344962406015038, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1688, + "step": 23861 + }, + { + "epoch": 1.1345563909774437, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 23862 + }, + { + "epoch": 1.1346165413533835, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 23863 + }, + { + "epoch": 1.1346766917293234, + "grad_norm": 1.625, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 23864 + }, + { + "epoch": 1.134736842105263, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 23865 + }, + { + "epoch": 1.134796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1567, + "step": 23866 + }, + { + "epoch": 1.1348571428571428, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 23867 + }, + { + "epoch": 1.1349172932330827, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 23868 + }, + { + "epoch": 1.1349774436090225, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 23869 + }, + { + "epoch": 1.1350375939849624, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 23870 + }, + { + "epoch": 1.1350977443609023, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 23871 + }, + { + "epoch": 1.1351578947368421, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 23872 + }, + { + "epoch": 1.135218045112782, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1776, + "step": 23873 + }, + { + "epoch": 1.1352781954887219, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 23874 + }, + { + "epoch": 1.1353383458646618, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 23875 + }, + { + "epoch": 1.1353984962406014, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 23876 + }, + { + "epoch": 1.1354586466165413, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 23877 + }, + { + "epoch": 1.1355187969924811, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 23878 + }, + { + "epoch": 1.135578947368421, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 23879 + }, + { + "epoch": 1.135639097744361, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 23880 + }, + { + "epoch": 1.1356992481203008, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 23881 + }, + { + "epoch": 1.1357593984962406, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 23882 + }, + { + "epoch": 1.1358195488721805, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 23883 + }, + { + "epoch": 1.1358796992481204, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9139, + "step": 23884 + }, + { + "epoch": 1.13593984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 23885 + }, + { + "epoch": 1.1360000000000001, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 23886 + }, + { + "epoch": 1.1360601503759398, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 23887 + }, + { + "epoch": 1.1361203007518796, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 23888 + }, + { + "epoch": 1.1361804511278195, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 23889 + }, + { + "epoch": 1.1362406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 23890 + }, + { + "epoch": 1.1363007518796993, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 23891 + }, + { + "epoch": 1.1363609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 23892 + }, + { + "epoch": 1.136421052631579, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 23893 + }, + { + "epoch": 1.1364812030075189, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 23894 + }, + { + "epoch": 1.1365413533834587, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 23895 + }, + { + "epoch": 1.1366015037593984, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 23896 + }, + { + "epoch": 1.1366616541353383, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 23897 + }, + { + "epoch": 1.1367218045112781, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 23898 + }, + { + "epoch": 1.136781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 23899 + }, + { + "epoch": 1.1368421052631579, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 23900 + }, + { + "epoch": 1.1369022556390977, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 23901 + }, + { + "epoch": 1.1369624060150376, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 23902 + }, + { + "epoch": 1.1370225563909775, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 23903 + }, + { + "epoch": 1.1370827067669174, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 23904 + }, + { + "epoch": 1.1371428571428572, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 23905 + }, + { + "epoch": 1.137203007518797, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 23906 + }, + { + "epoch": 1.1372631578947368, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 23907 + }, + { + "epoch": 1.1373233082706766, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9504, + "step": 23908 + }, + { + "epoch": 1.1373834586466165, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 23909 + }, + { + "epoch": 1.1374436090225564, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 23910 + }, + { + "epoch": 1.1375037593984962, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 23911 + }, + { + "epoch": 1.137563909774436, + "grad_norm": 1.40625, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 23912 + }, + { + "epoch": 1.137624060150376, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 23913 + }, + { + "epoch": 1.1376842105263159, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.2105, + "step": 23914 + }, + { + "epoch": 1.1377443609022557, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 23915 + }, + { + "epoch": 1.1378045112781954, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 23916 + }, + { + "epoch": 1.1378646616541352, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 23917 + }, + { + "epoch": 1.1379248120300751, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 23918 + }, + { + "epoch": 1.137984962406015, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1429, + "step": 23919 + }, + { + "epoch": 1.1380451127819549, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 23920 + }, + { + "epoch": 1.1381052631578947, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 23921 + }, + { + "epoch": 1.1381654135338346, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 23922 + }, + { + "epoch": 1.1382255639097745, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 23923 + }, + { + "epoch": 1.1382857142857143, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 23924 + }, + { + "epoch": 1.1383458646616542, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 23925 + }, + { + "epoch": 1.138406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 23926 + }, + { + "epoch": 1.1384661654135337, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 23927 + }, + { + "epoch": 1.1385263157894736, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 23928 + }, + { + "epoch": 1.1385864661654135, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 23929 + }, + { + "epoch": 1.1386466165413534, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 23930 + }, + { + "epoch": 1.1387067669172932, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 23931 + }, + { + "epoch": 1.138766917293233, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 23932 + }, + { + "epoch": 1.138827067669173, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 23933 + }, + { + "epoch": 1.1388872180451128, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 23934 + }, + { + "epoch": 1.1389473684210527, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 23935 + }, + { + "epoch": 1.1390075187969924, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1632, + "step": 23936 + }, + { + "epoch": 1.1390676691729325, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 23937 + }, + { + "epoch": 1.139127819548872, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 23938 + }, + { + "epoch": 1.139187969924812, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2796, + "step": 23939 + }, + { + "epoch": 1.1392481203007518, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 23940 + }, + { + "epoch": 1.1393082706766917, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 23941 + }, + { + "epoch": 1.1393684210526316, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 23942 + }, + { + "epoch": 1.1394285714285715, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 23943 + }, + { + "epoch": 1.1394887218045113, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 23944 + }, + { + "epoch": 1.1395488721804512, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 23945 + }, + { + "epoch": 1.139609022556391, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 23946 + }, + { + "epoch": 1.1396691729323307, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 23947 + }, + { + "epoch": 1.1397293233082706, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 23948 + }, + { + "epoch": 1.1397894736842105, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 23949 + }, + { + "epoch": 1.1398496240601503, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 23950 + }, + { + "epoch": 1.1399097744360902, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 23951 + }, + { + "epoch": 1.13996992481203, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.2064, + "step": 23952 + }, + { + "epoch": 1.14003007518797, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 23953 + }, + { + "epoch": 1.1400902255639098, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 23954 + }, + { + "epoch": 1.1401503759398497, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 23955 + }, + { + "epoch": 1.1402105263157896, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 23956 + }, + { + "epoch": 1.1402706766917294, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 23957 + }, + { + "epoch": 1.140330827067669, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 23958 + }, + { + "epoch": 1.140390977443609, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 23959 + }, + { + "epoch": 1.1404511278195488, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 23960 + }, + { + "epoch": 1.1405112781954887, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 23961 + }, + { + "epoch": 1.1405714285714286, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 23962 + }, + { + "epoch": 1.1406315789473684, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 23963 + }, + { + "epoch": 1.1406917293233083, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 23964 + }, + { + "epoch": 1.1407518796992482, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 23965 + }, + { + "epoch": 1.140812030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 23966 + }, + { + "epoch": 1.1408721804511277, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 23967 + }, + { + "epoch": 1.1409323308270676, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 23968 + }, + { + "epoch": 1.1409924812030074, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 23969 + }, + { + "epoch": 1.1410526315789473, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 23970 + }, + { + "epoch": 1.1411127819548872, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 23971 + }, + { + "epoch": 1.141172932330827, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 23972 + }, + { + "epoch": 1.141233082706767, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 23973 + }, + { + "epoch": 1.1412932330827068, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 23974 + }, + { + "epoch": 1.1413533834586467, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 23975 + }, + { + "epoch": 1.1414135338345865, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 23976 + }, + { + "epoch": 1.1414736842105264, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 23977 + }, + { + "epoch": 1.141533834586466, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.8949, + "step": 23978 + }, + { + "epoch": 1.141593984962406, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 23979 + }, + { + "epoch": 1.1416541353383458, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 23980 + }, + { + "epoch": 1.1417142857142857, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 23981 + }, + { + "epoch": 1.1417744360902256, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 23982 + }, + { + "epoch": 1.1418345864661654, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 23983 + }, + { + "epoch": 1.1418947368421053, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 23984 + }, + { + "epoch": 1.1419548872180452, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 23985 + }, + { + "epoch": 1.142015037593985, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 23986 + }, + { + "epoch": 1.1420751879699247, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 23987 + }, + { + "epoch": 1.1421353383458648, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 23988 + }, + { + "epoch": 1.1421954887218044, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 23989 + }, + { + "epoch": 1.1422556390977443, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 23990 + }, + { + "epoch": 1.1423157894736842, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 23991 + }, + { + "epoch": 1.142375939849624, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 23992 + }, + { + "epoch": 1.142436090225564, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 23993 + }, + { + "epoch": 1.1424962406015038, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 23994 + }, + { + "epoch": 1.1425563909774437, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 23995 + }, + { + "epoch": 1.1426165413533835, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 23996 + }, + { + "epoch": 1.1426766917293234, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 23997 + }, + { + "epoch": 1.142736842105263, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 23998 + }, + { + "epoch": 1.142796992481203, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 23999 + }, + { + "epoch": 1.1428571428571428, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 24000 + }, + { + "epoch": 1.1429172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 24001 + }, + { + "epoch": 1.1429774436090225, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 24002 + }, + { + "epoch": 1.1430375939849624, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 24003 + }, + { + "epoch": 1.1430977443609023, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 24004 + }, + { + "epoch": 1.1431578947368422, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 24005 + }, + { + "epoch": 1.143218045112782, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 24006 + }, + { + "epoch": 1.143278195488722, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 24007 + }, + { + "epoch": 1.1433383458646618, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 24008 + }, + { + "epoch": 1.1433984962406014, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 24009 + }, + { + "epoch": 1.1434586466165413, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 24010 + }, + { + "epoch": 1.1435187969924812, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 24011 + }, + { + "epoch": 1.143578947368421, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 24012 + }, + { + "epoch": 1.143639097744361, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 24013 + }, + { + "epoch": 1.1436992481203008, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 24014 + }, + { + "epoch": 1.1437593984962406, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 24015 + }, + { + "epoch": 1.1438195488721805, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.297, + "step": 24016 + }, + { + "epoch": 1.1438796992481204, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 24017 + }, + { + "epoch": 1.14393984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 24018 + }, + { + "epoch": 1.144, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 24019 + }, + { + "epoch": 1.1440601503759398, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 24020 + }, + { + "epoch": 1.1441203007518796, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1832, + "step": 24021 + }, + { + "epoch": 1.1441804511278195, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 24022 + }, + { + "epoch": 1.1442406015037594, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 24023 + }, + { + "epoch": 1.1443007518796993, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 24024 + }, + { + "epoch": 1.1443609022556391, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.2156, + "step": 24025 + }, + { + "epoch": 1.144421052631579, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 24026 + }, + { + "epoch": 1.1444812030075189, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 24027 + }, + { + "epoch": 1.1445413533834587, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 24028 + }, + { + "epoch": 1.1446015037593984, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 24029 + }, + { + "epoch": 1.1446616541353383, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 24030 + }, + { + "epoch": 1.1447218045112781, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 24031 + }, + { + "epoch": 1.144781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 24032 + }, + { + "epoch": 1.1448421052631579, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 24033 + }, + { + "epoch": 1.1449022556390978, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 24034 + }, + { + "epoch": 1.1449624060150376, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 24035 + }, + { + "epoch": 1.1450225563909775, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 24036 + }, + { + "epoch": 1.1450827067669174, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 24037 + }, + { + "epoch": 1.145142857142857, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 24038 + }, + { + "epoch": 1.1452030075187971, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 24039 + }, + { + "epoch": 1.1452631578947368, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 24040 + }, + { + "epoch": 1.1453233082706766, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 24041 + }, + { + "epoch": 1.1453834586466165, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 24042 + }, + { + "epoch": 1.1454436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 24043 + }, + { + "epoch": 1.1455037593984962, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 24044 + }, + { + "epoch": 1.1455639097744361, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 24045 + }, + { + "epoch": 1.145624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 24046 + }, + { + "epoch": 1.1456842105263159, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 24047 + }, + { + "epoch": 1.1457443609022557, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 24048 + }, + { + "epoch": 1.1458045112781954, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 24049 + }, + { + "epoch": 1.1458646616541353, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.176, + "step": 24050 + }, + { + "epoch": 1.1459248120300751, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 24051 + }, + { + "epoch": 1.145984962406015, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 24052 + }, + { + "epoch": 1.1460451127819549, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 24053 + }, + { + "epoch": 1.1461052631578947, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 24054 + }, + { + "epoch": 1.1461654135338346, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 24055 + }, + { + "epoch": 1.1462255639097745, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 24056 + }, + { + "epoch": 1.1462857142857144, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 24057 + }, + { + "epoch": 1.1463458646616542, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 24058 + }, + { + "epoch": 1.146406015037594, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 24059 + }, + { + "epoch": 1.1464661654135337, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 24060 + }, + { + "epoch": 1.1465263157894736, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 24061 + }, + { + "epoch": 1.1465864661654135, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 24062 + }, + { + "epoch": 1.1466466165413534, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 24063 + }, + { + "epoch": 1.1467067669172932, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 24064 + }, + { + "epoch": 1.146766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 24065 + }, + { + "epoch": 1.146827067669173, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 24066 + }, + { + "epoch": 1.1468872180451128, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 24067 + }, + { + "epoch": 1.1469473684210527, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 24068 + }, + { + "epoch": 1.1470075187969924, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 24069 + }, + { + "epoch": 1.1470676691729322, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1418, + "step": 24070 + }, + { + "epoch": 1.147127819548872, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 24071 + }, + { + "epoch": 1.147187969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 24072 + }, + { + "epoch": 1.1472481203007519, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 24073 + }, + { + "epoch": 1.1473082706766917, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 24074 + }, + { + "epoch": 1.1473684210526316, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 24075 + }, + { + "epoch": 1.1474285714285715, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 24076 + }, + { + "epoch": 1.1474887218045113, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 24077 + }, + { + "epoch": 1.1475488721804512, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 24078 + }, + { + "epoch": 1.147609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 24079 + }, + { + "epoch": 1.1476691729323307, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 24080 + }, + { + "epoch": 1.1477293233082706, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 24081 + }, + { + "epoch": 1.1477894736842105, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 24082 + }, + { + "epoch": 1.1478496240601503, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 24083 + }, + { + "epoch": 1.1479097744360902, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 24084 + }, + { + "epoch": 1.14796992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 24085 + }, + { + "epoch": 1.14803007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 24086 + }, + { + "epoch": 1.1480902255639098, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 24087 + }, + { + "epoch": 1.1481503759398497, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 24088 + }, + { + "epoch": 1.1482105263157896, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 24089 + }, + { + "epoch": 1.1482706766917294, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 24090 + }, + { + "epoch": 1.148330827067669, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 24091 + }, + { + "epoch": 1.148390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 24092 + }, + { + "epoch": 1.1484511278195488, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 24093 + }, + { + "epoch": 1.1485112781954887, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 24094 + }, + { + "epoch": 1.1485714285714286, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 24095 + }, + { + "epoch": 1.1486315789473684, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 24096 + }, + { + "epoch": 1.1486917293233083, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 24097 + }, + { + "epoch": 1.1487518796992482, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 24098 + }, + { + "epoch": 1.148812030075188, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 24099 + }, + { + "epoch": 1.1488721804511277, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 24100 + }, + { + "epoch": 1.1489323308270676, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 24101 + }, + { + "epoch": 1.1489924812030075, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 24102 + }, + { + "epoch": 1.1490526315789473, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 24103 + }, + { + "epoch": 1.1491127819548872, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 24104 + }, + { + "epoch": 1.149172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 24105 + }, + { + "epoch": 1.149233082706767, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 24106 + }, + { + "epoch": 1.1492932330827068, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 24107 + }, + { + "epoch": 1.1493533834586467, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 24108 + }, + { + "epoch": 1.1494135338345866, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 24109 + }, + { + "epoch": 1.1494736842105264, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 24110 + }, + { + "epoch": 1.149533834586466, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 24111 + }, + { + "epoch": 1.149593984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 24112 + }, + { + "epoch": 1.1496541353383458, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 24113 + }, + { + "epoch": 1.1497142857142857, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 24114 + }, + { + "epoch": 1.1497744360902256, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 24115 + }, + { + "epoch": 1.1498345864661654, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 24116 + }, + { + "epoch": 1.1498947368421053, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 24117 + }, + { + "epoch": 1.1499548872180452, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 24118 + }, + { + "epoch": 1.150015037593985, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 24119 + }, + { + "epoch": 1.1500751879699247, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 24120 + }, + { + "epoch": 1.1501353383458646, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 24121 + }, + { + "epoch": 1.1501954887218044, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 24122 + }, + { + "epoch": 1.1502556390977443, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 24123 + }, + { + "epoch": 1.1503157894736842, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 24124 + }, + { + "epoch": 1.150375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 24125 + }, + { + "epoch": 1.150436090225564, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 24126 + }, + { + "epoch": 1.1504962406015038, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 24127 + }, + { + "epoch": 1.1505563909774437, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 24128 + }, + { + "epoch": 1.1506165413533835, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 24129 + }, + { + "epoch": 1.1506766917293234, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 24130 + }, + { + "epoch": 1.150736842105263, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 24131 + }, + { + "epoch": 1.150796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 24132 + }, + { + "epoch": 1.1508571428571428, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 24133 + }, + { + "epoch": 1.1509172932330827, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 24134 + }, + { + "epoch": 1.1509774436090225, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 24135 + }, + { + "epoch": 1.1510375939849624, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 24136 + }, + { + "epoch": 1.1510977443609023, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 24137 + }, + { + "epoch": 1.1511578947368422, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 24138 + }, + { + "epoch": 1.151218045112782, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 24139 + }, + { + "epoch": 1.151278195488722, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1518, + "step": 24140 + }, + { + "epoch": 1.1513383458646618, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 24141 + }, + { + "epoch": 1.1513984962406014, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 24142 + }, + { + "epoch": 1.1514586466165413, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 24143 + }, + { + "epoch": 1.1515187969924812, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 24144 + }, + { + "epoch": 1.151578947368421, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 24145 + }, + { + "epoch": 1.151639097744361, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 24146 + }, + { + "epoch": 1.1516992481203008, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 24147 + }, + { + "epoch": 1.1517593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 24148 + }, + { + "epoch": 1.1518195488721805, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 24149 + }, + { + "epoch": 1.1518796992481204, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 24150 + }, + { + "epoch": 1.15193984962406, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 24151 + }, + { + "epoch": 1.152, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 24152 + }, + { + "epoch": 1.1520601503759398, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 24153 + }, + { + "epoch": 1.1521203007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 24154 + }, + { + "epoch": 1.1521804511278195, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 24155 + }, + { + "epoch": 1.1522406015037594, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.147, + "step": 24156 + }, + { + "epoch": 1.1523007518796993, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 24157 + }, + { + "epoch": 1.1523609022556391, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 24158 + }, + { + "epoch": 1.152421052631579, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 24159 + }, + { + "epoch": 1.1524812030075189, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 24160 + }, + { + "epoch": 1.1525413533834588, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 24161 + }, + { + "epoch": 1.1526015037593984, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 24162 + }, + { + "epoch": 1.1526616541353383, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 24163 + }, + { + "epoch": 1.1527218045112781, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 24164 + }, + { + "epoch": 1.152781954887218, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 24165 + }, + { + "epoch": 1.152842105263158, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 24166 + }, + { + "epoch": 1.1529022556390978, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 24167 + }, + { + "epoch": 1.1529624060150376, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 24168 + }, + { + "epoch": 1.1530225563909775, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 24169 + }, + { + "epoch": 1.1530827067669174, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 24170 + }, + { + "epoch": 1.153142857142857, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 24171 + }, + { + "epoch": 1.153203007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 24172 + }, + { + "epoch": 1.1532631578947368, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 24173 + }, + { + "epoch": 1.1533233082706766, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 24174 + }, + { + "epoch": 1.1533834586466165, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 24175 + }, + { + "epoch": 1.1534436090225564, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 24176 + }, + { + "epoch": 1.1535037593984963, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 24177 + }, + { + "epoch": 1.1535639097744361, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 24178 + }, + { + "epoch": 1.153624060150376, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 24179 + }, + { + "epoch": 1.1536842105263159, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 24180 + }, + { + "epoch": 1.1537443609022557, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2025, + "step": 24181 + }, + { + "epoch": 1.1538045112781954, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 24182 + }, + { + "epoch": 1.1538646616541353, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.2352, + "step": 24183 + }, + { + "epoch": 1.1539248120300751, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 24184 + }, + { + "epoch": 1.153984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 24185 + }, + { + "epoch": 1.1540451127819549, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1412, + "step": 24186 + }, + { + "epoch": 1.1541052631578947, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 24187 + }, + { + "epoch": 1.1541654135338346, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 24188 + }, + { + "epoch": 1.1542255639097745, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 24189 + }, + { + "epoch": 1.1542857142857144, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 24190 + }, + { + "epoch": 1.1543458646616542, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 24191 + }, + { + "epoch": 1.154406015037594, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 24192 + }, + { + "epoch": 1.1544661654135338, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 24193 + }, + { + "epoch": 1.1545263157894736, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 24194 + }, + { + "epoch": 1.1545864661654135, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 24195 + }, + { + "epoch": 1.1546466165413534, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 24196 + }, + { + "epoch": 1.1547067669172932, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.305, + "step": 24197 + }, + { + "epoch": 1.154766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 24198 + }, + { + "epoch": 1.154827067669173, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 24199 + }, + { + "epoch": 1.1548872180451129, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 24200 + }, + { + "epoch": 1.1549473684210527, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 24201 + }, + { + "epoch": 1.1550075187969924, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 24202 + }, + { + "epoch": 1.1550676691729322, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 24203 + }, + { + "epoch": 1.1551278195488721, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1704, + "step": 24204 + }, + { + "epoch": 1.155187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 24205 + }, + { + "epoch": 1.1552481203007519, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 24206 + }, + { + "epoch": 1.1553082706766917, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 24207 + }, + { + "epoch": 1.1553684210526316, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 24208 + }, + { + "epoch": 1.1554285714285715, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 24209 + }, + { + "epoch": 1.1554887218045113, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 24210 + }, + { + "epoch": 1.1555488721804512, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 24211 + }, + { + "epoch": 1.155609022556391, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 24212 + }, + { + "epoch": 1.1556691729323307, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 24213 + }, + { + "epoch": 1.1557293233082706, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 24214 + }, + { + "epoch": 1.1557894736842105, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 24215 + }, + { + "epoch": 1.1558496240601503, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 24216 + }, + { + "epoch": 1.1559097744360902, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 24217 + }, + { + "epoch": 1.15596992481203, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 24218 + }, + { + "epoch": 1.15603007518797, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 24219 + }, + { + "epoch": 1.1560902255639098, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 24220 + }, + { + "epoch": 1.1561503759398497, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 24221 + }, + { + "epoch": 1.1562105263157894, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 24222 + }, + { + "epoch": 1.1562706766917292, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 24223 + }, + { + "epoch": 1.156330827067669, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 24224 + }, + { + "epoch": 1.156390977443609, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 24225 + }, + { + "epoch": 1.1564511278195488, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 24226 + }, + { + "epoch": 1.1565112781954887, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 24227 + }, + { + "epoch": 1.1565714285714286, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 24228 + }, + { + "epoch": 1.1566315789473685, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 24229 + }, + { + "epoch": 1.1566917293233083, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 24230 + }, + { + "epoch": 1.1567518796992482, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 24231 + }, + { + "epoch": 1.156812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 24232 + }, + { + "epoch": 1.1568721804511277, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 24233 + }, + { + "epoch": 1.1569323308270676, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 24234 + }, + { + "epoch": 1.1569924812030075, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 24235 + }, + { + "epoch": 1.1570526315789473, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 24236 + }, + { + "epoch": 1.1571127819548872, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 24237 + }, + { + "epoch": 1.157172932330827, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 24238 + }, + { + "epoch": 1.157233082706767, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 24239 + }, + { + "epoch": 1.1572932330827068, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 24240 + }, + { + "epoch": 1.1573533834586467, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 24241 + }, + { + "epoch": 1.1574135338345866, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 24242 + }, + { + "epoch": 1.1574736842105264, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 24243 + }, + { + "epoch": 1.157533834586466, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 24244 + }, + { + "epoch": 1.157593984962406, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 24245 + }, + { + "epoch": 1.1576541353383458, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 24246 + }, + { + "epoch": 1.1577142857142857, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 24247 + }, + { + "epoch": 1.1577744360902256, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 24248 + }, + { + "epoch": 1.1578345864661654, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 24249 + }, + { + "epoch": 1.1578947368421053, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 24250 + }, + { + "epoch": 1.1579548872180452, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 24251 + }, + { + "epoch": 1.158015037593985, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 24252 + }, + { + "epoch": 1.1580751879699247, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 24253 + }, + { + "epoch": 1.1581353383458646, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 24254 + }, + { + "epoch": 1.1581954887218044, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.2866, + "step": 24255 + }, + { + "epoch": 1.1582556390977443, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 24256 + }, + { + "epoch": 1.1583157894736842, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 24257 + }, + { + "epoch": 1.158375939849624, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 24258 + }, + { + "epoch": 1.158436090225564, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 24259 + }, + { + "epoch": 1.1584962406015038, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 24260 + }, + { + "epoch": 1.1585563909774437, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 24261 + }, + { + "epoch": 1.1586165413533835, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 24262 + }, + { + "epoch": 1.1586766917293234, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 24263 + }, + { + "epoch": 1.158736842105263, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 24264 + }, + { + "epoch": 1.158796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 24265 + }, + { + "epoch": 1.1588571428571428, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 24266 + }, + { + "epoch": 1.1589172932330827, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 24267 + }, + { + "epoch": 1.1589774436090226, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 24268 + }, + { + "epoch": 1.1590375939849624, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 24269 + }, + { + "epoch": 1.1590977443609023, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 24270 + }, + { + "epoch": 1.1591578947368422, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 24271 + }, + { + "epoch": 1.159218045112782, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 24272 + }, + { + "epoch": 1.1592781954887217, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 24273 + }, + { + "epoch": 1.1593383458646618, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 24274 + }, + { + "epoch": 1.1593984962406014, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 24275 + }, + { + "epoch": 1.1594586466165413, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 24276 + }, + { + "epoch": 1.1595187969924812, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 24277 + }, + { + "epoch": 1.159578947368421, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 24278 + }, + { + "epoch": 1.159639097744361, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 24279 + }, + { + "epoch": 1.1596992481203008, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 24280 + }, + { + "epoch": 1.1597593984962407, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 24281 + }, + { + "epoch": 1.1598195488721805, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 24282 + }, + { + "epoch": 1.1598796992481204, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 24283 + }, + { + "epoch": 1.15993984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 24284 + }, + { + "epoch": 1.16, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 24285 + }, + { + "epoch": 1.1600601503759398, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 24286 + }, + { + "epoch": 1.1601203007518797, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 24287 + }, + { + "epoch": 1.1601804511278195, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 24288 + }, + { + "epoch": 1.1602406015037594, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 24289 + }, + { + "epoch": 1.1603007518796993, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 24290 + }, + { + "epoch": 1.1603609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 24291 + }, + { + "epoch": 1.160421052631579, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 24292 + }, + { + "epoch": 1.160481203007519, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 24293 + }, + { + "epoch": 1.1605413533834588, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 24294 + }, + { + "epoch": 1.1606015037593984, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 24295 + }, + { + "epoch": 1.1606616541353383, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 24296 + }, + { + "epoch": 1.1607218045112782, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 24297 + }, + { + "epoch": 1.160781954887218, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 24298 + }, + { + "epoch": 1.160842105263158, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 24299 + }, + { + "epoch": 1.1609022556390978, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 24300 + }, + { + "epoch": 1.1609624060150376, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 24301 + }, + { + "epoch": 1.1610225563909775, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 24302 + }, + { + "epoch": 1.1610827067669174, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 24303 + }, + { + "epoch": 1.161142857142857, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 24304 + }, + { + "epoch": 1.161203007518797, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 24305 + }, + { + "epoch": 1.1612631578947368, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 24306 + }, + { + "epoch": 1.1613233082706766, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 24307 + }, + { + "epoch": 1.1613834586466165, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 24308 + }, + { + "epoch": 1.1614436090225564, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 24309 + }, + { + "epoch": 1.1615037593984963, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 24310 + }, + { + "epoch": 1.1615639097744361, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 24311 + }, + { + "epoch": 1.161624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 24312 + }, + { + "epoch": 1.1616842105263159, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 24313 + }, + { + "epoch": 1.1617443609022557, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 24314 + }, + { + "epoch": 1.1618045112781954, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 24315 + }, + { + "epoch": 1.1618646616541353, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 24316 + }, + { + "epoch": 1.1619248120300751, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 24317 + }, + { + "epoch": 1.161984962406015, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 24318 + }, + { + "epoch": 1.1620451127819549, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 24319 + }, + { + "epoch": 1.1621052631578948, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 24320 + }, + { + "epoch": 1.1621654135338346, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 24321 + }, + { + "epoch": 1.1622255639097745, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 24322 + }, + { + "epoch": 1.1622857142857144, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 24323 + }, + { + "epoch": 1.162345864661654, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 24324 + }, + { + "epoch": 1.162406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 24325 + }, + { + "epoch": 1.1624661654135338, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 24326 + }, + { + "epoch": 1.1625263157894736, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 24327 + }, + { + "epoch": 1.1625864661654135, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 24328 + }, + { + "epoch": 1.1626466165413534, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 24329 + }, + { + "epoch": 1.1627067669172932, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 24330 + }, + { + "epoch": 1.1627669172932331, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 24331 + }, + { + "epoch": 1.162827067669173, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 24332 + }, + { + "epoch": 1.1628872180451129, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 24333 + }, + { + "epoch": 1.1629473684210527, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 24334 + }, + { + "epoch": 1.1630075187969924, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 24335 + }, + { + "epoch": 1.1630676691729323, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 24336 + }, + { + "epoch": 1.1631278195488721, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 24337 + }, + { + "epoch": 1.163187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 24338 + }, + { + "epoch": 1.1632481203007519, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 24339 + }, + { + "epoch": 1.1633082706766917, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 24340 + }, + { + "epoch": 1.1633684210526316, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 24341 + }, + { + "epoch": 1.1634285714285715, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 24342 + }, + { + "epoch": 1.1634887218045113, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 24343 + }, + { + "epoch": 1.1635488721804512, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 24344 + }, + { + "epoch": 1.163609022556391, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 24345 + }, + { + "epoch": 1.1636691729323307, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 24346 + }, + { + "epoch": 1.1637293233082706, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 24347 + }, + { + "epoch": 1.1637894736842105, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 24348 + }, + { + "epoch": 1.1638496240601504, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 24349 + }, + { + "epoch": 1.1639097744360902, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2527, + "step": 24350 + }, + { + "epoch": 1.16396992481203, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 24351 + }, + { + "epoch": 1.16403007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 24352 + }, + { + "epoch": 1.1640902255639098, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 24353 + }, + { + "epoch": 1.1641503759398497, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 24354 + }, + { + "epoch": 1.1642105263157894, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 24355 + }, + { + "epoch": 1.1642706766917292, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 24356 + }, + { + "epoch": 1.164330827067669, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 24357 + }, + { + "epoch": 1.164390977443609, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 24358 + }, + { + "epoch": 1.1644511278195488, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 24359 + }, + { + "epoch": 1.1645112781954887, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2417, + "step": 24360 + }, + { + "epoch": 1.1645714285714286, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 24361 + }, + { + "epoch": 1.1646315789473685, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 24362 + }, + { + "epoch": 1.1646917293233083, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 24363 + }, + { + "epoch": 1.1647518796992482, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 24364 + }, + { + "epoch": 1.164812030075188, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 24365 + }, + { + "epoch": 1.1648721804511277, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 24366 + }, + { + "epoch": 1.1649323308270676, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 24367 + }, + { + "epoch": 1.1649924812030075, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 24368 + }, + { + "epoch": 1.1650526315789473, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 24369 + }, + { + "epoch": 1.1651127819548872, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 24370 + }, + { + "epoch": 1.165172932330827, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 24371 + }, + { + "epoch": 1.165233082706767, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 24372 + }, + { + "epoch": 1.1652932330827068, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 24373 + }, + { + "epoch": 1.1653533834586467, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 24374 + }, + { + "epoch": 1.1654135338345863, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 24375 + }, + { + "epoch": 1.1654736842105264, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 24376 + }, + { + "epoch": 1.165533834586466, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 24377 + }, + { + "epoch": 1.165593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 24378 + }, + { + "epoch": 1.1656541353383458, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 24379 + }, + { + "epoch": 1.1657142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1581, + "step": 24380 + }, + { + "epoch": 1.1657744360902256, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 24381 + }, + { + "epoch": 1.1658345864661654, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 24382 + }, + { + "epoch": 1.1658947368421053, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1751, + "step": 24383 + }, + { + "epoch": 1.1659548872180452, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 24384 + }, + { + "epoch": 1.166015037593985, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 24385 + }, + { + "epoch": 1.1660751879699247, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 24386 + }, + { + "epoch": 1.1661353383458646, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 24387 + }, + { + "epoch": 1.1661954887218045, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 24388 + }, + { + "epoch": 1.1662556390977443, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 24389 + }, + { + "epoch": 1.1663157894736842, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 24390 + }, + { + "epoch": 1.166375939849624, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 24391 + }, + { + "epoch": 1.166436090225564, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 24392 + }, + { + "epoch": 1.1664962406015038, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 24393 + }, + { + "epoch": 1.1665563909774437, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 24394 + }, + { + "epoch": 1.1666165413533836, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 24395 + }, + { + "epoch": 1.1666766917293234, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 24396 + }, + { + "epoch": 1.166736842105263, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 24397 + }, + { + "epoch": 1.166796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 24398 + }, + { + "epoch": 1.1668571428571428, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 24399 + }, + { + "epoch": 1.1669172932330827, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 24400 + }, + { + "epoch": 1.1669774436090226, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 24401 + }, + { + "epoch": 1.1670375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 24402 + }, + { + "epoch": 1.1670977443609023, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 24403 + }, + { + "epoch": 1.1671578947368422, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 24404 + }, + { + "epoch": 1.167218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 24405 + }, + { + "epoch": 1.1672781954887217, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 24406 + }, + { + "epoch": 1.1673383458646616, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 24407 + }, + { + "epoch": 1.1673984962406014, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 24408 + }, + { + "epoch": 1.1674586466165413, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 24409 + }, + { + "epoch": 1.1675187969924812, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 24410 + }, + { + "epoch": 1.167578947368421, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 24411 + }, + { + "epoch": 1.167639097744361, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 24412 + }, + { + "epoch": 1.1676992481203008, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 24413 + }, + { + "epoch": 1.1677593984962407, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 24414 + }, + { + "epoch": 1.1678195488721805, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 24415 + }, + { + "epoch": 1.1678796992481204, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 24416 + }, + { + "epoch": 1.16793984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 24417 + }, + { + "epoch": 1.168, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 24418 + }, + { + "epoch": 1.1680601503759398, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 24419 + }, + { + "epoch": 1.1681203007518797, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 24420 + }, + { + "epoch": 1.1681804511278195, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 24421 + }, + { + "epoch": 1.1682406015037594, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 24422 + }, + { + "epoch": 1.1683007518796993, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 24423 + }, + { + "epoch": 1.1683609022556392, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 24424 + }, + { + "epoch": 1.168421052631579, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 24425 + }, + { + "epoch": 1.1684812030075187, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 24426 + }, + { + "epoch": 1.1685413533834588, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0994, + "step": 24427 + }, + { + "epoch": 1.1686015037593984, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 24428 + }, + { + "epoch": 1.1686616541353383, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 24429 + }, + { + "epoch": 1.1687218045112782, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 24430 + }, + { + "epoch": 1.168781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 24431 + }, + { + "epoch": 1.168842105263158, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 24432 + }, + { + "epoch": 1.1689022556390978, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 24433 + }, + { + "epoch": 1.1689624060150376, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2113, + "step": 24434 + }, + { + "epoch": 1.1690225563909775, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 24435 + }, + { + "epoch": 1.1690827067669174, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 24436 + }, + { + "epoch": 1.169142857142857, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 24437 + }, + { + "epoch": 1.169203007518797, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 24438 + }, + { + "epoch": 1.1692631578947368, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 24439 + }, + { + "epoch": 1.1693233082706767, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 24440 + }, + { + "epoch": 1.1693834586466165, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1807, + "step": 24441 + }, + { + "epoch": 1.1694436090225564, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 24442 + }, + { + "epoch": 1.1695037593984963, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 24443 + }, + { + "epoch": 1.1695639097744361, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 24444 + }, + { + "epoch": 1.169624060150376, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2315, + "step": 24445 + }, + { + "epoch": 1.1696842105263159, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 24446 + }, + { + "epoch": 1.1697443609022558, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 24447 + }, + { + "epoch": 1.1698045112781954, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 24448 + }, + { + "epoch": 1.1698646616541353, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 24449 + }, + { + "epoch": 1.1699248120300751, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 24450 + }, + { + "epoch": 1.169984962406015, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 24451 + }, + { + "epoch": 1.1700451127819549, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 24452 + }, + { + "epoch": 1.1701052631578948, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 24453 + }, + { + "epoch": 1.1701654135338346, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 24454 + }, + { + "epoch": 1.1702255639097745, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 24455 + }, + { + "epoch": 1.1702857142857144, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 24456 + }, + { + "epoch": 1.170345864661654, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 24457 + }, + { + "epoch": 1.170406015037594, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 24458 + }, + { + "epoch": 1.1704661654135338, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 24459 + }, + { + "epoch": 1.1705263157894736, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 24460 + }, + { + "epoch": 1.1705864661654135, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 24461 + }, + { + "epoch": 1.1706466165413534, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 24462 + }, + { + "epoch": 1.1707067669172933, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 24463 + }, + { + "epoch": 1.1707669172932331, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 24464 + }, + { + "epoch": 1.170827067669173, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 24465 + }, + { + "epoch": 1.1708872180451129, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 24466 + }, + { + "epoch": 1.1709473684210527, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 24467 + }, + { + "epoch": 1.1710075187969924, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 24468 + }, + { + "epoch": 1.1710676691729323, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 24469 + }, + { + "epoch": 1.1711278195488721, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 24470 + }, + { + "epoch": 1.171187969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 24471 + }, + { + "epoch": 1.1712481203007519, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 24472 + }, + { + "epoch": 1.1713082706766917, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 24473 + }, + { + "epoch": 1.1713684210526316, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 24474 + }, + { + "epoch": 1.1714285714285715, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 24475 + }, + { + "epoch": 1.1714887218045114, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 24476 + }, + { + "epoch": 1.171548872180451, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 24477 + }, + { + "epoch": 1.171609022556391, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 24478 + }, + { + "epoch": 1.1716691729323307, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 24479 + }, + { + "epoch": 1.1717293233082706, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 24480 + }, + { + "epoch": 1.1717894736842105, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 24481 + }, + { + "epoch": 1.1718496240601504, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 24482 + }, + { + "epoch": 1.1719097744360902, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 24483 + }, + { + "epoch": 1.17196992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 24484 + }, + { + "epoch": 1.17203007518797, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2147, + "step": 24485 + }, + { + "epoch": 1.1720902255639098, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 24486 + }, + { + "epoch": 1.1721503759398497, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 24487 + }, + { + "epoch": 1.1722105263157894, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 24488 + }, + { + "epoch": 1.1722706766917292, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 24489 + }, + { + "epoch": 1.1723308270676691, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 24490 + }, + { + "epoch": 1.172390977443609, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 24491 + }, + { + "epoch": 1.1724511278195489, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 24492 + }, + { + "epoch": 1.1725112781954887, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 24493 + }, + { + "epoch": 1.1725714285714286, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 24494 + }, + { + "epoch": 1.1726315789473685, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 24495 + }, + { + "epoch": 1.1726917293233083, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 24496 + }, + { + "epoch": 1.1727518796992482, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 24497 + }, + { + "epoch": 1.172812030075188, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 24498 + }, + { + "epoch": 1.1728721804511277, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2489, + "step": 24499 + }, + { + "epoch": 1.1729323308270676, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 24500 + }, + { + "epoch": 1.1729924812030075, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 24501 + }, + { + "epoch": 1.1730526315789473, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 24502 + }, + { + "epoch": 1.1731127819548872, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.198, + "step": 24503 + }, + { + "epoch": 1.173172932330827, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 24504 + }, + { + "epoch": 1.173233082706767, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 24505 + }, + { + "epoch": 1.1732932330827068, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 24506 + }, + { + "epoch": 1.1733533834586467, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 24507 + }, + { + "epoch": 1.1734135338345864, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 24508 + }, + { + "epoch": 1.1734736842105262, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 24509 + }, + { + "epoch": 1.173533834586466, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1667, + "step": 24510 + }, + { + "epoch": 1.173593984962406, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 24511 + }, + { + "epoch": 1.1736541353383458, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 24512 + }, + { + "epoch": 1.1737142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 24513 + }, + { + "epoch": 1.1737744360902256, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 24514 + }, + { + "epoch": 1.1738345864661655, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 24515 + }, + { + "epoch": 1.1738947368421053, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 24516 + }, + { + "epoch": 1.1739548872180452, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 24517 + }, + { + "epoch": 1.174015037593985, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 24518 + }, + { + "epoch": 1.1740751879699247, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 24519 + }, + { + "epoch": 1.1741353383458646, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 24520 + }, + { + "epoch": 1.1741954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 24521 + }, + { + "epoch": 1.1742556390977443, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 24522 + }, + { + "epoch": 1.1743157894736842, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 24523 + }, + { + "epoch": 1.174375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 24524 + }, + { + "epoch": 1.174436090225564, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 24525 + }, + { + "epoch": 1.1744962406015038, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 24526 + }, + { + "epoch": 1.1745563909774437, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 24527 + }, + { + "epoch": 1.1746165413533833, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 24528 + }, + { + "epoch": 1.1746766917293234, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 24529 + }, + { + "epoch": 1.174736842105263, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 24530 + }, + { + "epoch": 1.174796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 24531 + }, + { + "epoch": 1.1748571428571428, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 24532 + }, + { + "epoch": 1.1749172932330827, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 24533 + }, + { + "epoch": 1.1749774436090226, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 24534 + }, + { + "epoch": 1.1750375939849624, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 24535 + }, + { + "epoch": 1.1750977443609023, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 24536 + }, + { + "epoch": 1.1751578947368422, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 24537 + }, + { + "epoch": 1.175218045112782, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 24538 + }, + { + "epoch": 1.1752781954887217, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 24539 + }, + { + "epoch": 1.1753383458646616, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 24540 + }, + { + "epoch": 1.1753984962406014, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 24541 + }, + { + "epoch": 1.1754586466165413, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 24542 + }, + { + "epoch": 1.1755187969924812, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 24543 + }, + { + "epoch": 1.175578947368421, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 24544 + }, + { + "epoch": 1.175639097744361, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 24545 + }, + { + "epoch": 1.1756992481203008, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 24546 + }, + { + "epoch": 1.1757593984962407, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 24547 + }, + { + "epoch": 1.1758195488721805, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 24548 + }, + { + "epoch": 1.1758796992481204, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 24549 + }, + { + "epoch": 1.17593984962406, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 24550 + }, + { + "epoch": 1.176, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 24551 + }, + { + "epoch": 1.1760601503759398, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 24552 + }, + { + "epoch": 1.1761203007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 24553 + }, + { + "epoch": 1.1761804511278195, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 24554 + }, + { + "epoch": 1.1762406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 24555 + }, + { + "epoch": 1.1763007518796993, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 24556 + }, + { + "epoch": 1.1763609022556392, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 24557 + }, + { + "epoch": 1.176421052631579, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 24558 + }, + { + "epoch": 1.1764812030075187, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 24559 + }, + { + "epoch": 1.1765413533834586, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 24560 + }, + { + "epoch": 1.1766015037593984, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 24561 + }, + { + "epoch": 1.1766616541353383, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 24562 + }, + { + "epoch": 1.1767218045112782, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 24563 + }, + { + "epoch": 1.176781954887218, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 24564 + }, + { + "epoch": 1.176842105263158, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 24565 + }, + { + "epoch": 1.1769022556390978, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 24566 + }, + { + "epoch": 1.1769624060150377, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 24567 + }, + { + "epoch": 1.1770225563909775, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 24568 + }, + { + "epoch": 1.1770827067669174, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 24569 + }, + { + "epoch": 1.177142857142857, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 24570 + }, + { + "epoch": 1.177203007518797, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 24571 + }, + { + "epoch": 1.1772631578947368, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 24572 + }, + { + "epoch": 1.1773233082706767, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 24573 + }, + { + "epoch": 1.1773834586466165, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 24574 + }, + { + "epoch": 1.1774436090225564, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 24575 + }, + { + "epoch": 1.1775037593984963, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 24576 + }, + { + "epoch": 1.1775639097744361, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 24577 + }, + { + "epoch": 1.177624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 24578 + }, + { + "epoch": 1.1776842105263157, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 24579 + }, + { + "epoch": 1.1777443609022558, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 24580 + }, + { + "epoch": 1.1778045112781954, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 24581 + }, + { + "epoch": 1.1778646616541353, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 24582 + }, + { + "epoch": 1.1779248120300752, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 24583 + }, + { + "epoch": 1.177984962406015, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 24584 + }, + { + "epoch": 1.178045112781955, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 24585 + }, + { + "epoch": 1.1781052631578948, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 24586 + }, + { + "epoch": 1.1781654135338346, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 24587 + }, + { + "epoch": 1.1782255639097745, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 24588 + }, + { + "epoch": 1.1782857142857144, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 24589 + }, + { + "epoch": 1.178345864661654, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1768, + "step": 24590 + }, + { + "epoch": 1.178406015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 24591 + }, + { + "epoch": 1.1784661654135338, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 24592 + }, + { + "epoch": 1.1785263157894736, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 24593 + }, + { + "epoch": 1.1785864661654135, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 24594 + }, + { + "epoch": 1.1786466165413534, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 24595 + }, + { + "epoch": 1.1787067669172933, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 24596 + }, + { + "epoch": 1.1787669172932331, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 24597 + }, + { + "epoch": 1.178827067669173, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 24598 + }, + { + "epoch": 1.1788872180451129, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 24599 + }, + { + "epoch": 1.1789473684210527, + "grad_norm": 1.5390625, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 24600 + }, + { + "epoch": 1.1790075187969924, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 24601 + }, + { + "epoch": 1.1790676691729323, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 24602 + }, + { + "epoch": 1.1791278195488721, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 24603 + }, + { + "epoch": 1.179187969924812, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1167, + "step": 24604 + }, + { + "epoch": 1.1792481203007519, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 24605 + }, + { + "epoch": 1.1793082706766918, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 24606 + }, + { + "epoch": 1.1793684210526316, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 24607 + }, + { + "epoch": 1.1794285714285715, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 24608 + }, + { + "epoch": 1.1794887218045114, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 24609 + }, + { + "epoch": 1.179548872180451, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 24610 + }, + { + "epoch": 1.1796090225563909, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1954, + "step": 24611 + }, + { + "epoch": 1.1796691729323308, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 24612 + }, + { + "epoch": 1.1797293233082706, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 24613 + }, + { + "epoch": 1.1797894736842105, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 24614 + }, + { + "epoch": 1.1798496240601504, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 24615 + }, + { + "epoch": 1.1799097744360902, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 24616 + }, + { + "epoch": 1.1799699248120301, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 24617 + }, + { + "epoch": 1.18003007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 24618 + }, + { + "epoch": 1.1800902255639099, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 24619 + }, + { + "epoch": 1.1801503759398497, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 24620 + }, + { + "epoch": 1.1802105263157894, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1334, + "step": 24621 + }, + { + "epoch": 1.1802706766917292, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2322, + "step": 24622 + }, + { + "epoch": 1.1803308270676691, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 24623 + }, + { + "epoch": 1.180390977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 24624 + }, + { + "epoch": 1.1804511278195489, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 24625 + }, + { + "epoch": 1.1805112781954887, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 24626 + }, + { + "epoch": 1.1805714285714286, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 24627 + }, + { + "epoch": 1.1806315789473685, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 24628 + }, + { + "epoch": 1.1806917293233083, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 24629 + }, + { + "epoch": 1.180751879699248, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 24630 + }, + { + "epoch": 1.180812030075188, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 24631 + }, + { + "epoch": 1.1808721804511277, + "grad_norm": 1.390625, + "learning_rate": 3e-05, + "loss": 1.2146, + "step": 24632 + }, + { + "epoch": 1.1809323308270676, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 24633 + }, + { + "epoch": 1.1809924812030075, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 24634 + }, + { + "epoch": 1.1810526315789474, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 24635 + }, + { + "epoch": 1.1811127819548872, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 24636 + }, + { + "epoch": 1.181172932330827, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 24637 + }, + { + "epoch": 1.181233082706767, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 24638 + }, + { + "epoch": 1.1812932330827068, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 24639 + }, + { + "epoch": 1.1813533834586467, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 24640 + }, + { + "epoch": 1.1814135338345864, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 24641 + }, + { + "epoch": 1.1814736842105262, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 24642 + }, + { + "epoch": 1.181533834586466, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 24643 + }, + { + "epoch": 1.181593984962406, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.2297, + "step": 24644 + }, + { + "epoch": 1.1816541353383458, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2298, + "step": 24645 + }, + { + "epoch": 1.1817142857142857, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 24646 + }, + { + "epoch": 1.1817744360902256, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 24647 + }, + { + "epoch": 1.1818345864661655, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1482, + "step": 24648 + }, + { + "epoch": 1.1818947368421053, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 24649 + }, + { + "epoch": 1.1819548872180452, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 24650 + }, + { + "epoch": 1.182015037593985, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.3083, + "step": 24651 + }, + { + "epoch": 1.1820751879699247, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 24652 + }, + { + "epoch": 1.1821353383458646, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 24653 + }, + { + "epoch": 1.1821954887218045, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 24654 + }, + { + "epoch": 1.1822556390977443, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 24655 + }, + { + "epoch": 1.1823157894736842, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 24656 + }, + { + "epoch": 1.182375939849624, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 24657 + }, + { + "epoch": 1.182436090225564, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1959, + "step": 24658 + }, + { + "epoch": 1.1824962406015038, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 24659 + }, + { + "epoch": 1.1825563909774437, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 24660 + }, + { + "epoch": 1.1826165413533833, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 24661 + }, + { + "epoch": 1.1826766917293232, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 24662 + }, + { + "epoch": 1.182736842105263, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 24663 + }, + { + "epoch": 1.182796992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 24664 + }, + { + "epoch": 1.1828571428571428, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 24665 + }, + { + "epoch": 1.1829172932330827, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 24666 + }, + { + "epoch": 1.1829774436090226, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 24667 + }, + { + "epoch": 1.1830375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 24668 + }, + { + "epoch": 1.1830977443609023, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 24669 + }, + { + "epoch": 1.1831578947368422, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 24670 + }, + { + "epoch": 1.183218045112782, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 24671 + }, + { + "epoch": 1.1832781954887217, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 24672 + }, + { + "epoch": 1.1833383458646616, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 24673 + }, + { + "epoch": 1.1833984962406014, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 24674 + }, + { + "epoch": 1.1834586466165413, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 24675 + }, + { + "epoch": 1.1835187969924812, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 24676 + }, + { + "epoch": 1.183578947368421, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 24677 + }, + { + "epoch": 1.183639097744361, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 24678 + }, + { + "epoch": 1.1836992481203008, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 24679 + }, + { + "epoch": 1.1837593984962407, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 24680 + }, + { + "epoch": 1.1838195488721803, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 24681 + }, + { + "epoch": 1.1838796992481204, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 24682 + }, + { + "epoch": 1.18393984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 24683 + }, + { + "epoch": 1.184, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 24684 + }, + { + "epoch": 1.1840601503759398, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 24685 + }, + { + "epoch": 1.1841203007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 24686 + }, + { + "epoch": 1.1841804511278196, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 24687 + }, + { + "epoch": 1.1842406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 24688 + }, + { + "epoch": 1.1843007518796993, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 24689 + }, + { + "epoch": 1.1843609022556392, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 24690 + }, + { + "epoch": 1.184421052631579, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 24691 + }, + { + "epoch": 1.1844812030075187, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 24692 + }, + { + "epoch": 1.1845413533834586, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 24693 + }, + { + "epoch": 1.1846015037593984, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 24694 + }, + { + "epoch": 1.1846616541353383, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1908, + "step": 24695 + }, + { + "epoch": 1.1847218045112782, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 24696 + }, + { + "epoch": 1.184781954887218, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 24697 + }, + { + "epoch": 1.184842105263158, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 24698 + }, + { + "epoch": 1.1849022556390978, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 24699 + }, + { + "epoch": 1.1849624060150377, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 24700 + }, + { + "epoch": 1.1850225563909775, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 24701 + }, + { + "epoch": 1.1850827067669174, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 24702 + }, + { + "epoch": 1.185142857142857, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 24703 + }, + { + "epoch": 1.185203007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 24704 + }, + { + "epoch": 1.1852631578947368, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 24705 + }, + { + "epoch": 1.1853233082706767, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 24706 + }, + { + "epoch": 1.1853834586466165, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 24707 + }, + { + "epoch": 1.1854436090225564, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 24708 + }, + { + "epoch": 1.1855037593984963, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 24709 + }, + { + "epoch": 1.1855639097744362, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 24710 + }, + { + "epoch": 1.185624060150376, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 24711 + }, + { + "epoch": 1.1856842105263157, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 24712 + }, + { + "epoch": 1.1857443609022555, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 24713 + }, + { + "epoch": 1.1858045112781954, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 24714 + }, + { + "epoch": 1.1858646616541353, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 24715 + }, + { + "epoch": 1.1859248120300752, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 24716 + }, + { + "epoch": 1.185984962406015, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 24717 + }, + { + "epoch": 1.186045112781955, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 24718 + }, + { + "epoch": 1.1861052631578948, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 24719 + }, + { + "epoch": 1.1861654135338346, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9133, + "step": 24720 + }, + { + "epoch": 1.1862255639097745, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 24721 + }, + { + "epoch": 1.1862857142857144, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 24722 + }, + { + "epoch": 1.186345864661654, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 24723 + }, + { + "epoch": 1.186406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 24724 + }, + { + "epoch": 1.1864661654135338, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 24725 + }, + { + "epoch": 1.1865263157894737, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 24726 + }, + { + "epoch": 1.1865864661654135, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 24727 + }, + { + "epoch": 1.1866466165413534, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 24728 + }, + { + "epoch": 1.1867067669172933, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 24729 + }, + { + "epoch": 1.1867669172932331, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 24730 + }, + { + "epoch": 1.186827067669173, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 24731 + }, + { + "epoch": 1.1868872180451127, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 24732 + }, + { + "epoch": 1.1869473684210528, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 24733 + }, + { + "epoch": 1.1870075187969924, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 24734 + }, + { + "epoch": 1.1870676691729323, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 24735 + }, + { + "epoch": 1.1871278195488721, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 24736 + }, + { + "epoch": 1.187187969924812, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 24737 + }, + { + "epoch": 1.1872481203007519, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 24738 + }, + { + "epoch": 1.1873082706766918, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 24739 + }, + { + "epoch": 1.1873684210526316, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 24740 + }, + { + "epoch": 1.1874285714285715, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 24741 + }, + { + "epoch": 1.1874887218045114, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 24742 + }, + { + "epoch": 1.187548872180451, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 24743 + }, + { + "epoch": 1.187609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 24744 + }, + { + "epoch": 1.1876691729323308, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 24745 + }, + { + "epoch": 1.1877293233082706, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 24746 + }, + { + "epoch": 1.1877894736842105, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 24747 + }, + { + "epoch": 1.1878496240601504, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 24748 + }, + { + "epoch": 1.1879097744360902, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 24749 + }, + { + "epoch": 1.1879699248120301, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 24750 + }, + { + "epoch": 1.18803007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 24751 + }, + { + "epoch": 1.1880902255639099, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 24752 + }, + { + "epoch": 1.1881503759398497, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 24753 + }, + { + "epoch": 1.1882105263157894, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 24754 + }, + { + "epoch": 1.1882706766917293, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 24755 + }, + { + "epoch": 1.1883308270676691, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 24756 + }, + { + "epoch": 1.188390977443609, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 24757 + }, + { + "epoch": 1.1884511278195489, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 24758 + }, + { + "epoch": 1.1885112781954887, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 24759 + }, + { + "epoch": 1.1885714285714286, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2365, + "step": 24760 + }, + { + "epoch": 1.1886315789473685, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 24761 + }, + { + "epoch": 1.1886917293233084, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 24762 + }, + { + "epoch": 1.188751879699248, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 24763 + }, + { + "epoch": 1.1888120300751879, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 24764 + }, + { + "epoch": 1.1888721804511277, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 24765 + }, + { + "epoch": 1.1889323308270676, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 24766 + }, + { + "epoch": 1.1889924812030075, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 24767 + }, + { + "epoch": 1.1890526315789474, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 24768 + }, + { + "epoch": 1.1891127819548872, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 24769 + }, + { + "epoch": 1.189172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 24770 + }, + { + "epoch": 1.189233082706767, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 24771 + }, + { + "epoch": 1.1892932330827068, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 24772 + }, + { + "epoch": 1.1893533834586467, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 24773 + }, + { + "epoch": 1.1894135338345864, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 24774 + }, + { + "epoch": 1.1894736842105262, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 24775 + }, + { + "epoch": 1.189533834586466, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 24776 + }, + { + "epoch": 1.189593984962406, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 24777 + }, + { + "epoch": 1.1896541353383459, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 24778 + }, + { + "epoch": 1.1897142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 24779 + }, + { + "epoch": 1.1897744360902256, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 24780 + }, + { + "epoch": 1.1898345864661655, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 24781 + }, + { + "epoch": 1.1898947368421053, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 24782 + }, + { + "epoch": 1.1899548872180452, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 24783 + }, + { + "epoch": 1.190015037593985, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 24784 + }, + { + "epoch": 1.1900751879699247, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 24785 + }, + { + "epoch": 1.1901353383458646, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 24786 + }, + { + "epoch": 1.1901954887218045, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 24787 + }, + { + "epoch": 1.1902556390977443, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 24788 + }, + { + "epoch": 1.1903157894736842, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 24789 + }, + { + "epoch": 1.190375939849624, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 24790 + }, + { + "epoch": 1.190436090225564, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 24791 + }, + { + "epoch": 1.1904962406015038, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 24792 + }, + { + "epoch": 1.1905563909774437, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 24793 + }, + { + "epoch": 1.1906165413533834, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 24794 + }, + { + "epoch": 1.1906766917293232, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 24795 + }, + { + "epoch": 1.190736842105263, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.3699, + "step": 24796 + }, + { + "epoch": 1.190796992481203, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 24797 + }, + { + "epoch": 1.1908571428571428, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 24798 + }, + { + "epoch": 1.1909172932330827, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 24799 + }, + { + "epoch": 1.1909774436090226, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 24800 + }, + { + "epoch": 1.1910375939849625, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 24801 + }, + { + "epoch": 1.1910977443609023, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 24802 + }, + { + "epoch": 1.1911578947368422, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 24803 + }, + { + "epoch": 1.191218045112782, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 24804 + }, + { + "epoch": 1.1912781954887217, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 24805 + }, + { + "epoch": 1.1913383458646616, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 24806 + }, + { + "epoch": 1.1913984962406015, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 24807 + }, + { + "epoch": 1.1914586466165413, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1889, + "step": 24808 + }, + { + "epoch": 1.1915187969924812, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.376, + "step": 24809 + }, + { + "epoch": 1.191578947368421, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 24810 + }, + { + "epoch": 1.191639097744361, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 24811 + }, + { + "epoch": 1.1916992481203008, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 24812 + }, + { + "epoch": 1.1917593984962407, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 24813 + }, + { + "epoch": 1.1918195488721803, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 24814 + }, + { + "epoch": 1.1918796992481202, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 24815 + }, + { + "epoch": 1.19193984962406, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 24816 + }, + { + "epoch": 1.192, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 24817 + }, + { + "epoch": 1.1920601503759398, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 24818 + }, + { + "epoch": 1.1921203007518797, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 24819 + }, + { + "epoch": 1.1921804511278196, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 24820 + }, + { + "epoch": 1.1922406015037594, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 24821 + }, + { + "epoch": 1.1923007518796993, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 24822 + }, + { + "epoch": 1.1923609022556392, + "grad_norm": 1.8984375, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 24823 + }, + { + "epoch": 1.192421052631579, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 24824 + }, + { + "epoch": 1.1924812030075187, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1124, + "step": 24825 + }, + { + "epoch": 1.1925413533834586, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 24826 + }, + { + "epoch": 1.1926015037593984, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 24827 + }, + { + "epoch": 1.1926616541353383, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 24828 + }, + { + "epoch": 1.1927218045112782, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 24829 + }, + { + "epoch": 1.192781954887218, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 24830 + }, + { + "epoch": 1.192842105263158, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 24831 + }, + { + "epoch": 1.1929022556390978, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 24832 + }, + { + "epoch": 1.1929624060150377, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 24833 + }, + { + "epoch": 1.1930225563909775, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 24834 + }, + { + "epoch": 1.1930827067669174, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 24835 + }, + { + "epoch": 1.193142857142857, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2558, + "step": 24836 + }, + { + "epoch": 1.193203007518797, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 24837 + }, + { + "epoch": 1.1932631578947368, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 24838 + }, + { + "epoch": 1.1933233082706767, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 24839 + }, + { + "epoch": 1.1933834586466165, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 24840 + }, + { + "epoch": 1.1934436090225564, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 24841 + }, + { + "epoch": 1.1935037593984963, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 24842 + }, + { + "epoch": 1.1935639097744362, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 24843 + }, + { + "epoch": 1.193624060150376, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2189, + "step": 24844 + }, + { + "epoch": 1.1936842105263157, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 24845 + }, + { + "epoch": 1.1937443609022556, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 24846 + }, + { + "epoch": 1.1938045112781954, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 24847 + }, + { + "epoch": 1.1938646616541353, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 24848 + }, + { + "epoch": 1.1939248120300752, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 24849 + }, + { + "epoch": 1.193984962406015, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 24850 + }, + { + "epoch": 1.194045112781955, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 24851 + }, + { + "epoch": 1.1941052631578948, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 24852 + }, + { + "epoch": 1.1941654135338347, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 24853 + }, + { + "epoch": 1.1942255639097745, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 24854 + }, + { + "epoch": 1.1942857142857144, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 24855 + }, + { + "epoch": 1.194345864661654, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 24856 + }, + { + "epoch": 1.194406015037594, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 24857 + }, + { + "epoch": 1.1944661654135338, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 24858 + }, + { + "epoch": 1.1945263157894737, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 24859 + }, + { + "epoch": 1.1945864661654135, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 24860 + }, + { + "epoch": 1.1946466165413534, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 24861 + }, + { + "epoch": 1.1947067669172933, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 24862 + }, + { + "epoch": 1.1947669172932331, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 24863 + }, + { + "epoch": 1.194827067669173, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 24864 + }, + { + "epoch": 1.1948872180451127, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 24865 + }, + { + "epoch": 1.1949473684210525, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 24866 + }, + { + "epoch": 1.1950075187969924, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 24867 + }, + { + "epoch": 1.1950676691729323, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 24868 + }, + { + "epoch": 1.1951278195488722, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 24869 + }, + { + "epoch": 1.195187969924812, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 24870 + }, + { + "epoch": 1.195248120300752, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 24871 + }, + { + "epoch": 1.1953082706766918, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 24872 + }, + { + "epoch": 1.1953684210526316, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 24873 + }, + { + "epoch": 1.1954285714285715, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 24874 + }, + { + "epoch": 1.1954887218045114, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 24875 + }, + { + "epoch": 1.195548872180451, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 24876 + }, + { + "epoch": 1.195609022556391, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 24877 + }, + { + "epoch": 1.1956691729323308, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 24878 + }, + { + "epoch": 1.1957293233082706, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 24879 + }, + { + "epoch": 1.1957894736842105, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1847, + "step": 24880 + }, + { + "epoch": 1.1958496240601504, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 24881 + }, + { + "epoch": 1.1959097744360903, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 24882 + }, + { + "epoch": 1.1959699248120301, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 24883 + }, + { + "epoch": 1.19603007518797, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 24884 + }, + { + "epoch": 1.1960902255639099, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 24885 + }, + { + "epoch": 1.1961503759398497, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 24886 + }, + { + "epoch": 1.1962105263157894, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 24887 + }, + { + "epoch": 1.1962706766917293, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 24888 + }, + { + "epoch": 1.1963308270676691, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 24889 + }, + { + "epoch": 1.196390977443609, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 24890 + }, + { + "epoch": 1.1964511278195489, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 24891 + }, + { + "epoch": 1.1965112781954887, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 24892 + }, + { + "epoch": 1.1965714285714286, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 24893 + }, + { + "epoch": 1.1966315789473685, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 24894 + }, + { + "epoch": 1.1966917293233084, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 24895 + }, + { + "epoch": 1.196751879699248, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 24896 + }, + { + "epoch": 1.1968120300751879, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 24897 + }, + { + "epoch": 1.1968721804511278, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 24898 + }, + { + "epoch": 1.1969323308270676, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 24899 + }, + { + "epoch": 1.1969924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 24900 + }, + { + "epoch": 1.1970526315789474, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 24901 + }, + { + "epoch": 1.1971127819548872, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2325, + "step": 24902 + }, + { + "epoch": 1.197172932330827, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 24903 + }, + { + "epoch": 1.197233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 24904 + }, + { + "epoch": 1.1972932330827069, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 24905 + }, + { + "epoch": 1.1973533834586467, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 24906 + }, + { + "epoch": 1.1974135338345864, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 24907 + }, + { + "epoch": 1.1974736842105262, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 24908 + }, + { + "epoch": 1.1975338345864661, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 24909 + }, + { + "epoch": 1.197593984962406, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 24910 + }, + { + "epoch": 1.1976541353383459, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 24911 + }, + { + "epoch": 1.1977142857142857, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 24912 + }, + { + "epoch": 1.1977744360902256, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 24913 + }, + { + "epoch": 1.1978345864661655, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 24914 + }, + { + "epoch": 1.1978947368421053, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 24915 + }, + { + "epoch": 1.197954887218045, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1429, + "step": 24916 + }, + { + "epoch": 1.1980150375939849, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 24917 + }, + { + "epoch": 1.1980751879699247, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 24918 + }, + { + "epoch": 1.1981353383458646, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 24919 + }, + { + "epoch": 1.1981954887218045, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 24920 + }, + { + "epoch": 1.1982556390977444, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 24921 + }, + { + "epoch": 1.1983157894736842, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 24922 + }, + { + "epoch": 1.198375939849624, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 24923 + }, + { + "epoch": 1.198436090225564, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 24924 + }, + { + "epoch": 1.1984962406015038, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 24925 + }, + { + "epoch": 1.1985563909774437, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1757, + "step": 24926 + }, + { + "epoch": 1.1986165413533834, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 24927 + }, + { + "epoch": 1.1986766917293232, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 24928 + }, + { + "epoch": 1.198736842105263, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 24929 + }, + { + "epoch": 1.198796992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 24930 + }, + { + "epoch": 1.1988571428571428, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 24931 + }, + { + "epoch": 1.1989172932330827, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 24932 + }, + { + "epoch": 1.1989774436090226, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 24933 + }, + { + "epoch": 1.1990375939849625, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1767, + "step": 24934 + }, + { + "epoch": 1.1990977443609023, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 24935 + }, + { + "epoch": 1.1991578947368422, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 24936 + }, + { + "epoch": 1.199218045112782, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 24937 + }, + { + "epoch": 1.1992781954887217, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 24938 + }, + { + "epoch": 1.1993383458646616, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 24939 + }, + { + "epoch": 1.1993984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 24940 + }, + { + "epoch": 1.1994586466165413, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 24941 + }, + { + "epoch": 1.1995187969924812, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 24942 + }, + { + "epoch": 1.199578947368421, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 24943 + }, + { + "epoch": 1.199639097744361, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 24944 + }, + { + "epoch": 1.1996992481203008, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 24945 + }, + { + "epoch": 1.1997593984962407, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 24946 + }, + { + "epoch": 1.1998195488721803, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 24947 + }, + { + "epoch": 1.1998796992481202, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 24948 + }, + { + "epoch": 1.19993984962406, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 24949 + }, + { + "epoch": 1.2, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 24950 + }, + { + "epoch": 1.2000601503759398, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2392, + "step": 24951 + }, + { + "epoch": 1.2001203007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 24952 + }, + { + "epoch": 1.2001804511278196, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2084, + "step": 24953 + }, + { + "epoch": 1.2002406015037594, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 24954 + }, + { + "epoch": 1.2003007518796993, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 24955 + }, + { + "epoch": 1.2003609022556392, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 24956 + }, + { + "epoch": 1.200421052631579, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 24957 + }, + { + "epoch": 1.2004812030075187, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 24958 + }, + { + "epoch": 1.2005413533834586, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 24959 + }, + { + "epoch": 1.2006015037593984, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 24960 + }, + { + "epoch": 1.2006616541353383, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 24961 + }, + { + "epoch": 1.2007218045112782, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 24962 + }, + { + "epoch": 1.200781954887218, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 24963 + }, + { + "epoch": 1.200842105263158, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 24964 + }, + { + "epoch": 1.2009022556390978, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 24965 + }, + { + "epoch": 1.2009624060150377, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 24966 + }, + { + "epoch": 1.2010225563909773, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 24967 + }, + { + "epoch": 1.2010827067669172, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 24968 + }, + { + "epoch": 1.201142857142857, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 24969 + }, + { + "epoch": 1.201203007518797, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 24970 + }, + { + "epoch": 1.2012631578947368, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 24971 + }, + { + "epoch": 1.2013233082706767, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 24972 + }, + { + "epoch": 1.2013834586466166, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 24973 + }, + { + "epoch": 1.2014436090225564, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 24974 + }, + { + "epoch": 1.2015037593984963, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 24975 + }, + { + "epoch": 1.2015639097744362, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 24976 + }, + { + "epoch": 1.201624060150376, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 24977 + }, + { + "epoch": 1.2016842105263157, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 24978 + }, + { + "epoch": 1.2017443609022556, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 24979 + }, + { + "epoch": 1.2018045112781954, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1763, + "step": 24980 + }, + { + "epoch": 1.2018646616541353, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 24981 + }, + { + "epoch": 1.2019248120300752, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 24982 + }, + { + "epoch": 1.201984962406015, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 24983 + }, + { + "epoch": 1.202045112781955, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 24984 + }, + { + "epoch": 1.2021052631578948, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 24985 + }, + { + "epoch": 1.2021654135338347, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 24986 + }, + { + "epoch": 1.2022255639097745, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 24987 + }, + { + "epoch": 1.2022857142857144, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 24988 + }, + { + "epoch": 1.202345864661654, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 24989 + }, + { + "epoch": 1.202406015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 24990 + }, + { + "epoch": 1.2024661654135338, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 24991 + }, + { + "epoch": 1.2025263157894737, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 24992 + }, + { + "epoch": 1.2025864661654135, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 24993 + }, + { + "epoch": 1.2026466165413534, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 24994 + }, + { + "epoch": 1.2027067669172933, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 24995 + }, + { + "epoch": 1.2027669172932332, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 24996 + }, + { + "epoch": 1.202827067669173, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 24997 + }, + { + "epoch": 1.2028872180451127, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 24998 + }, + { + "epoch": 1.2029473684210525, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 24999 + }, + { + "epoch": 1.2030075187969924, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9001, + "step": 25000 + }, + { + "epoch": 1.2030676691729323, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2586, + "step": 25001 + }, + { + "epoch": 1.2031278195488722, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 25002 + }, + { + "epoch": 1.203187969924812, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 25003 + }, + { + "epoch": 1.203248120300752, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 25004 + }, + { + "epoch": 1.2033082706766918, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 25005 + }, + { + "epoch": 1.2033684210526316, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 25006 + }, + { + "epoch": 1.2034285714285715, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 25007 + }, + { + "epoch": 1.2034887218045114, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2719, + "step": 25008 + }, + { + "epoch": 1.203548872180451, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 25009 + }, + { + "epoch": 1.203609022556391, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 25010 + }, + { + "epoch": 1.2036691729323308, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 25011 + }, + { + "epoch": 1.2037293233082706, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 25012 + }, + { + "epoch": 1.2037894736842105, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 25013 + }, + { + "epoch": 1.2038496240601504, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1494, + "step": 25014 + }, + { + "epoch": 1.2039097744360903, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 25015 + }, + { + "epoch": 1.2039699248120301, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 25016 + }, + { + "epoch": 1.20403007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2195, + "step": 25017 + }, + { + "epoch": 1.2040902255639097, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 25018 + }, + { + "epoch": 1.2041503759398497, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1638, + "step": 25019 + }, + { + "epoch": 1.2042105263157894, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 25020 + }, + { + "epoch": 1.2042706766917293, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 25021 + }, + { + "epoch": 1.2043308270676691, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 25022 + }, + { + "epoch": 1.204390977443609, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 25023 + }, + { + "epoch": 1.2044511278195489, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 25024 + }, + { + "epoch": 1.2045112781954888, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 25025 + }, + { + "epoch": 1.2045714285714286, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 25026 + }, + { + "epoch": 1.2046315789473685, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 25027 + }, + { + "epoch": 1.2046917293233084, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 25028 + }, + { + "epoch": 1.204751879699248, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 25029 + }, + { + "epoch": 1.204812030075188, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 25030 + }, + { + "epoch": 1.2048721804511278, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 25031 + }, + { + "epoch": 1.2049323308270676, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2003, + "step": 25032 + }, + { + "epoch": 1.2049924812030075, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 25033 + }, + { + "epoch": 1.2050526315789474, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2536, + "step": 25034 + }, + { + "epoch": 1.2051127819548872, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 25035 + }, + { + "epoch": 1.2051729323308271, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 25036 + }, + { + "epoch": 1.205233082706767, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 25037 + }, + { + "epoch": 1.2052932330827069, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 25038 + }, + { + "epoch": 1.2053533834586467, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 25039 + }, + { + "epoch": 1.2054135338345864, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 25040 + }, + { + "epoch": 1.2054736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 25041 + }, + { + "epoch": 1.2055338345864661, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 25042 + }, + { + "epoch": 1.205593984962406, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 25043 + }, + { + "epoch": 1.2056541353383459, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 25044 + }, + { + "epoch": 1.2057142857142857, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 25045 + }, + { + "epoch": 1.2057744360902256, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 25046 + }, + { + "epoch": 1.2058345864661655, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 25047 + }, + { + "epoch": 1.2058947368421054, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 25048 + }, + { + "epoch": 1.205954887218045, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 25049 + }, + { + "epoch": 1.2060150375939849, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 25050 + }, + { + "epoch": 1.2060751879699247, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 25051 + }, + { + "epoch": 1.2061353383458646, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 25052 + }, + { + "epoch": 1.2061954887218045, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 25053 + }, + { + "epoch": 1.2062556390977444, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 25054 + }, + { + "epoch": 1.2063157894736842, + "grad_norm": 1.9375, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 25055 + }, + { + "epoch": 1.206375939849624, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 25056 + }, + { + "epoch": 1.206436090225564, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 25057 + }, + { + "epoch": 1.2064962406015038, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 25058 + }, + { + "epoch": 1.2065563909774437, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 25059 + }, + { + "epoch": 1.2066165413533834, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 25060 + }, + { + "epoch": 1.2066766917293232, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 25061 + }, + { + "epoch": 1.206736842105263, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 25062 + }, + { + "epoch": 1.206796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 25063 + }, + { + "epoch": 1.2068571428571429, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 25064 + }, + { + "epoch": 1.2069172932330827, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 25065 + }, + { + "epoch": 1.2069774436090226, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 25066 + }, + { + "epoch": 1.2070375939849625, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 25067 + }, + { + "epoch": 1.2070977443609023, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1725, + "step": 25068 + }, + { + "epoch": 1.207157894736842, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 25069 + }, + { + "epoch": 1.207218045112782, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 25070 + }, + { + "epoch": 1.2072781954887217, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 25071 + }, + { + "epoch": 1.2073383458646616, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.275, + "step": 25072 + }, + { + "epoch": 1.2073984962406015, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 25073 + }, + { + "epoch": 1.2074586466165413, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 25074 + }, + { + "epoch": 1.2075187969924812, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 25075 + }, + { + "epoch": 1.207578947368421, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 25076 + }, + { + "epoch": 1.207639097744361, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 25077 + }, + { + "epoch": 1.2076992481203008, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 25078 + }, + { + "epoch": 1.2077593984962407, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 25079 + }, + { + "epoch": 1.2078195488721803, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 25080 + }, + { + "epoch": 1.2078796992481202, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 25081 + }, + { + "epoch": 1.20793984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 25082 + }, + { + "epoch": 1.208, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 25083 + }, + { + "epoch": 1.2080601503759398, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 25084 + }, + { + "epoch": 1.2081203007518797, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 25085 + }, + { + "epoch": 1.2081804511278196, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 25086 + }, + { + "epoch": 1.2082406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 25087 + }, + { + "epoch": 1.2083007518796993, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 25088 + }, + { + "epoch": 1.2083609022556392, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 25089 + }, + { + "epoch": 1.208421052631579, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 25090 + }, + { + "epoch": 1.2084812030075187, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 25091 + }, + { + "epoch": 1.2085413533834586, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 25092 + }, + { + "epoch": 1.2086015037593985, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 25093 + }, + { + "epoch": 1.2086616541353383, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 25094 + }, + { + "epoch": 1.2087218045112782, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 25095 + }, + { + "epoch": 1.208781954887218, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 25096 + }, + { + "epoch": 1.208842105263158, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 25097 + }, + { + "epoch": 1.2089022556390978, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 25098 + }, + { + "epoch": 1.2089624060150377, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 25099 + }, + { + "epoch": 1.2090225563909773, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 25100 + }, + { + "epoch": 1.2090827067669172, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 25101 + }, + { + "epoch": 1.209142857142857, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 25102 + }, + { + "epoch": 1.209203007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 25103 + }, + { + "epoch": 1.2092631578947368, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.6247, + "step": 25104 + }, + { + "epoch": 1.2093233082706767, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 25105 + }, + { + "epoch": 1.2093834586466166, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 25106 + }, + { + "epoch": 1.2094436090225564, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 25107 + }, + { + "epoch": 1.2095037593984963, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 25108 + }, + { + "epoch": 1.2095639097744362, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 25109 + }, + { + "epoch": 1.209624060150376, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 25110 + }, + { + "epoch": 1.2096842105263157, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 25111 + }, + { + "epoch": 1.2097443609022556, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 25112 + }, + { + "epoch": 1.2098045112781954, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 25113 + }, + { + "epoch": 1.2098646616541353, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 25114 + }, + { + "epoch": 1.2099248120300752, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 25115 + }, + { + "epoch": 1.209984962406015, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 25116 + }, + { + "epoch": 1.210045112781955, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 25117 + }, + { + "epoch": 1.2101052631578948, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 25118 + }, + { + "epoch": 1.2101654135338347, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 25119 + }, + { + "epoch": 1.2102255639097743, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1585, + "step": 25120 + }, + { + "epoch": 1.2102857142857144, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 25121 + }, + { + "epoch": 1.210345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 25122 + }, + { + "epoch": 1.210406015037594, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 25123 + }, + { + "epoch": 1.2104661654135338, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 25124 + }, + { + "epoch": 1.2105263157894737, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 25125 + }, + { + "epoch": 1.2105864661654135, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2694, + "step": 25126 + }, + { + "epoch": 1.2106466165413534, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 25127 + }, + { + "epoch": 1.2107067669172933, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 25128 + }, + { + "epoch": 1.2107669172932332, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 25129 + }, + { + "epoch": 1.210827067669173, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 25130 + }, + { + "epoch": 1.2108872180451127, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 25131 + }, + { + "epoch": 1.2109473684210526, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 25132 + }, + { + "epoch": 1.2110075187969924, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 25133 + }, + { + "epoch": 1.2110676691729323, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 25134 + }, + { + "epoch": 1.2111278195488722, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 25135 + }, + { + "epoch": 1.211187969924812, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 25136 + }, + { + "epoch": 1.211248120300752, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 25137 + }, + { + "epoch": 1.2113082706766918, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 25138 + }, + { + "epoch": 1.2113684210526316, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 25139 + }, + { + "epoch": 1.2114285714285715, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 25140 + }, + { + "epoch": 1.2114887218045114, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 25141 + }, + { + "epoch": 1.211548872180451, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 25142 + }, + { + "epoch": 1.211609022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 25143 + }, + { + "epoch": 1.2116691729323308, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 25144 + }, + { + "epoch": 1.2117293233082707, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 25145 + }, + { + "epoch": 1.2117894736842105, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 25146 + }, + { + "epoch": 1.2118496240601504, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 25147 + }, + { + "epoch": 1.2119097744360903, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 25148 + }, + { + "epoch": 1.2119699248120301, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 25149 + }, + { + "epoch": 1.21203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 25150 + }, + { + "epoch": 1.2120902255639097, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1945, + "step": 25151 + }, + { + "epoch": 1.2121503759398495, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 25152 + }, + { + "epoch": 1.2122105263157894, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 25153 + }, + { + "epoch": 1.2122706766917293, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 25154 + }, + { + "epoch": 1.2123308270676691, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1545, + "step": 25155 + }, + { + "epoch": 1.212390977443609, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2396, + "step": 25156 + }, + { + "epoch": 1.212451127819549, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 25157 + }, + { + "epoch": 1.2125112781954888, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 25158 + }, + { + "epoch": 1.2125714285714286, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 25159 + }, + { + "epoch": 1.2126315789473685, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2762, + "step": 25160 + }, + { + "epoch": 1.2126917293233084, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 25161 + }, + { + "epoch": 1.212751879699248, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 25162 + }, + { + "epoch": 1.212812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 25163 + }, + { + "epoch": 1.2128721804511278, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 25164 + }, + { + "epoch": 1.2129323308270676, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 25165 + }, + { + "epoch": 1.2129924812030075, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 25166 + }, + { + "epoch": 1.2130526315789474, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 25167 + }, + { + "epoch": 1.2131127819548873, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 25168 + }, + { + "epoch": 1.2131729323308271, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 25169 + }, + { + "epoch": 1.213233082706767, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 25170 + }, + { + "epoch": 1.2132932330827066, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 25171 + }, + { + "epoch": 1.2133533834586467, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 25172 + }, + { + "epoch": 1.2134135338345864, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 25173 + }, + { + "epoch": 1.2134736842105263, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 25174 + }, + { + "epoch": 1.2135338345864661, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1764, + "step": 25175 + }, + { + "epoch": 1.213593984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 25176 + }, + { + "epoch": 1.2136541353383459, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 25177 + }, + { + "epoch": 1.2137142857142857, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2057, + "step": 25178 + }, + { + "epoch": 1.2137744360902256, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 25179 + }, + { + "epoch": 1.2138345864661655, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 25180 + }, + { + "epoch": 1.2138947368421054, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 25181 + }, + { + "epoch": 1.213954887218045, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 25182 + }, + { + "epoch": 1.2140150375939849, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 25183 + }, + { + "epoch": 1.2140751879699248, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 25184 + }, + { + "epoch": 1.2141353383458646, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 25185 + }, + { + "epoch": 1.2141954887218045, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 25186 + }, + { + "epoch": 1.2142556390977444, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 25187 + }, + { + "epoch": 1.2143157894736842, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 25188 + }, + { + "epoch": 1.214375939849624, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 25189 + }, + { + "epoch": 1.214436090225564, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.2505, + "step": 25190 + }, + { + "epoch": 1.2144962406015039, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 25191 + }, + { + "epoch": 1.2145563909774437, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 25192 + }, + { + "epoch": 1.2146165413533834, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 25193 + }, + { + "epoch": 1.2146766917293232, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 25194 + }, + { + "epoch": 1.2147368421052631, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 25195 + }, + { + "epoch": 1.214796992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 25196 + }, + { + "epoch": 1.2148571428571429, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 25197 + }, + { + "epoch": 1.2149172932330827, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 25198 + }, + { + "epoch": 1.2149774436090226, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 25199 + }, + { + "epoch": 1.2150375939849625, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 25200 + }, + { + "epoch": 1.2150977443609023, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 25201 + }, + { + "epoch": 1.215157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 25202 + }, + { + "epoch": 1.2152180451127819, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 25203 + }, + { + "epoch": 1.2152781954887217, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 25204 + }, + { + "epoch": 1.2153383458646616, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 25205 + }, + { + "epoch": 1.2153984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 25206 + }, + { + "epoch": 1.2154586466165413, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 25207 + }, + { + "epoch": 1.2155187969924812, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 25208 + }, + { + "epoch": 1.215578947368421, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 25209 + }, + { + "epoch": 1.215639097744361, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 25210 + }, + { + "epoch": 1.2156992481203008, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 25211 + }, + { + "epoch": 1.2157593984962407, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1633, + "step": 25212 + }, + { + "epoch": 1.2158195488721804, + "grad_norm": 1.7265625, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 25213 + }, + { + "epoch": 1.2158796992481202, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 25214 + }, + { + "epoch": 1.21593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 25215 + }, + { + "epoch": 1.216, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 25216 + }, + { + "epoch": 1.2160601503759398, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 25217 + }, + { + "epoch": 1.2161203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 25218 + }, + { + "epoch": 1.2161804511278196, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 25219 + }, + { + "epoch": 1.2162406015037595, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 25220 + }, + { + "epoch": 1.2163007518796993, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 25221 + }, + { + "epoch": 1.216360902255639, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 25222 + }, + { + "epoch": 1.216421052631579, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 25223 + }, + { + "epoch": 1.2164812030075187, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 25224 + }, + { + "epoch": 1.2165413533834586, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 25225 + }, + { + "epoch": 1.2166015037593985, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 25226 + }, + { + "epoch": 1.2166616541353383, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 25227 + }, + { + "epoch": 1.2167218045112782, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1876, + "step": 25228 + }, + { + "epoch": 1.216781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 25229 + }, + { + "epoch": 1.216842105263158, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 25230 + }, + { + "epoch": 1.2169022556390978, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 25231 + }, + { + "epoch": 1.2169624060150377, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 25232 + }, + { + "epoch": 1.2170225563909773, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 25233 + }, + { + "epoch": 1.2170827067669172, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 25234 + }, + { + "epoch": 1.217142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 25235 + }, + { + "epoch": 1.217203007518797, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 25236 + }, + { + "epoch": 1.2172631578947368, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 25237 + }, + { + "epoch": 1.2173233082706767, + "grad_norm": 1.5625, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 25238 + }, + { + "epoch": 1.2173834586466166, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 25239 + }, + { + "epoch": 1.2174436090225564, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 25240 + }, + { + "epoch": 1.2175037593984963, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 25241 + }, + { + "epoch": 1.2175639097744362, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 25242 + }, + { + "epoch": 1.217624060150376, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 25243 + }, + { + "epoch": 1.2176842105263157, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 25244 + }, + { + "epoch": 1.2177443609022556, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 25245 + }, + { + "epoch": 1.2178045112781954, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 25246 + }, + { + "epoch": 1.2178646616541353, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 25247 + }, + { + "epoch": 1.2179248120300752, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 25248 + }, + { + "epoch": 1.217984962406015, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 25249 + }, + { + "epoch": 1.218045112781955, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 25250 + }, + { + "epoch": 1.2181052631578948, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 25251 + }, + { + "epoch": 1.2181654135338347, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 25252 + }, + { + "epoch": 1.2182255639097743, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 25253 + }, + { + "epoch": 1.2182857142857142, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 25254 + }, + { + "epoch": 1.218345864661654, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 25255 + }, + { + "epoch": 1.218406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 25256 + }, + { + "epoch": 1.2184661654135338, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 25257 + }, + { + "epoch": 1.2185263157894737, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 25258 + }, + { + "epoch": 1.2185864661654136, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 25259 + }, + { + "epoch": 1.2186466165413534, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 25260 + }, + { + "epoch": 1.2187067669172933, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 25261 + }, + { + "epoch": 1.2187669172932332, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 25262 + }, + { + "epoch": 1.218827067669173, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 25263 + }, + { + "epoch": 1.2188872180451127, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 25264 + }, + { + "epoch": 1.2189473684210526, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 25265 + }, + { + "epoch": 1.2190075187969924, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 25266 + }, + { + "epoch": 1.2190676691729323, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 25267 + }, + { + "epoch": 1.2191278195488722, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 25268 + }, + { + "epoch": 1.219187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 25269 + }, + { + "epoch": 1.219248120300752, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 25270 + }, + { + "epoch": 1.2193082706766918, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 25271 + }, + { + "epoch": 1.2193684210526317, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 25272 + }, + { + "epoch": 1.2194285714285713, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 25273 + }, + { + "epoch": 1.2194887218045114, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 25274 + }, + { + "epoch": 1.219548872180451, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 25275 + }, + { + "epoch": 1.219609022556391, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 25276 + }, + { + "epoch": 1.2196691729323308, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 25277 + }, + { + "epoch": 1.2197293233082707, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 25278 + }, + { + "epoch": 1.2197894736842105, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.215, + "step": 25279 + }, + { + "epoch": 1.2198496240601504, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 25280 + }, + { + "epoch": 1.2199097744360903, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 25281 + }, + { + "epoch": 1.2199699248120301, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 25282 + }, + { + "epoch": 1.22003007518797, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 25283 + }, + { + "epoch": 1.2200902255639097, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 25284 + }, + { + "epoch": 1.2201503759398495, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 25285 + }, + { + "epoch": 1.2202105263157894, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 25286 + }, + { + "epoch": 1.2202706766917293, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 25287 + }, + { + "epoch": 1.2203308270676692, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 25288 + }, + { + "epoch": 1.220390977443609, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 25289 + }, + { + "epoch": 1.220451127819549, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 25290 + }, + { + "epoch": 1.2205112781954888, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.194, + "step": 25291 + }, + { + "epoch": 1.2205714285714286, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 25292 + }, + { + "epoch": 1.2206315789473685, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 25293 + }, + { + "epoch": 1.2206917293233084, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 25294 + }, + { + "epoch": 1.220751879699248, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 25295 + }, + { + "epoch": 1.220812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 25296 + }, + { + "epoch": 1.2208721804511278, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1496, + "step": 25297 + }, + { + "epoch": 1.2209323308270676, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 25298 + }, + { + "epoch": 1.2209924812030075, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 25299 + }, + { + "epoch": 1.2210526315789474, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 25300 + }, + { + "epoch": 1.2211127819548873, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 25301 + }, + { + "epoch": 1.2211729323308271, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 25302 + }, + { + "epoch": 1.221233082706767, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 25303 + }, + { + "epoch": 1.2212932330827067, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 25304 + }, + { + "epoch": 1.2213533834586465, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 25305 + }, + { + "epoch": 1.2214135338345864, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 25306 + }, + { + "epoch": 1.2214736842105263, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 25307 + }, + { + "epoch": 1.2215338345864661, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 25308 + }, + { + "epoch": 1.221593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 25309 + }, + { + "epoch": 1.2216541353383459, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 25310 + }, + { + "epoch": 1.2217142857142858, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 25311 + }, + { + "epoch": 1.2217744360902256, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 25312 + }, + { + "epoch": 1.2218345864661655, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 25313 + }, + { + "epoch": 1.2218947368421054, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 25314 + }, + { + "epoch": 1.221954887218045, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 25315 + }, + { + "epoch": 1.2220150375939849, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 25316 + }, + { + "epoch": 1.2220751879699248, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 25317 + }, + { + "epoch": 1.2221353383458646, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 25318 + }, + { + "epoch": 1.2221954887218045, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 25319 + }, + { + "epoch": 1.2222556390977444, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 25320 + }, + { + "epoch": 1.2223157894736842, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 25321 + }, + { + "epoch": 1.2223759398496241, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 25322 + }, + { + "epoch": 1.222436090225564, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 25323 + }, + { + "epoch": 1.2224962406015036, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 25324 + }, + { + "epoch": 1.2225563909774437, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 25325 + }, + { + "epoch": 1.2226165413533834, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 25326 + }, + { + "epoch": 1.2226766917293233, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 25327 + }, + { + "epoch": 1.2227368421052631, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 25328 + }, + { + "epoch": 1.222796992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 25329 + }, + { + "epoch": 1.2228571428571429, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 25330 + }, + { + "epoch": 1.2229172932330827, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 25331 + }, + { + "epoch": 1.2229774436090226, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 25332 + }, + { + "epoch": 1.2230375939849625, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 25333 + }, + { + "epoch": 1.2230977443609024, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 25334 + }, + { + "epoch": 1.223157894736842, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 25335 + }, + { + "epoch": 1.2232180451127819, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 25336 + }, + { + "epoch": 1.2232781954887217, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 25337 + }, + { + "epoch": 1.2233383458646616, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 25338 + }, + { + "epoch": 1.2233984962406015, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 25339 + }, + { + "epoch": 1.2234586466165414, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 25340 + }, + { + "epoch": 1.2235187969924812, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 25341 + }, + { + "epoch": 1.223578947368421, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 25342 + }, + { + "epoch": 1.223639097744361, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 25343 + }, + { + "epoch": 1.2236992481203008, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 25344 + }, + { + "epoch": 1.2237593984962407, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 25345 + }, + { + "epoch": 1.2238195488721804, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 25346 + }, + { + "epoch": 1.2238796992481202, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 25347 + }, + { + "epoch": 1.22393984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 25348 + }, + { + "epoch": 1.224, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 25349 + }, + { + "epoch": 1.2240601503759398, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 25350 + }, + { + "epoch": 1.2241203007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 25351 + }, + { + "epoch": 1.2241804511278196, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 25352 + }, + { + "epoch": 1.2242406015037595, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 25353 + }, + { + "epoch": 1.2243007518796993, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 25354 + }, + { + "epoch": 1.224360902255639, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 25355 + }, + { + "epoch": 1.2244210526315789, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 25356 + }, + { + "epoch": 1.2244812030075187, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 25357 + }, + { + "epoch": 1.2245413533834586, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 25358 + }, + { + "epoch": 1.2246015037593985, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 25359 + }, + { + "epoch": 1.2246616541353383, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 25360 + }, + { + "epoch": 1.2247218045112782, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 25361 + }, + { + "epoch": 1.224781954887218, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 25362 + }, + { + "epoch": 1.224842105263158, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1691, + "step": 25363 + }, + { + "epoch": 1.2249022556390978, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 25364 + }, + { + "epoch": 1.2249624060150377, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 25365 + }, + { + "epoch": 1.2250225563909773, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 25366 + }, + { + "epoch": 1.2250827067669172, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 25367 + }, + { + "epoch": 1.225142857142857, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 25368 + }, + { + "epoch": 1.225203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 25369 + }, + { + "epoch": 1.2252631578947368, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 25370 + }, + { + "epoch": 1.2253233082706767, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 25371 + }, + { + "epoch": 1.2253834586466166, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 25372 + }, + { + "epoch": 1.2254436090225564, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 25373 + }, + { + "epoch": 1.2255037593984963, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 25374 + }, + { + "epoch": 1.225563909774436, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 25375 + }, + { + "epoch": 1.225624060150376, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 25376 + }, + { + "epoch": 1.2256842105263157, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 25377 + }, + { + "epoch": 1.2257443609022556, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 25378 + }, + { + "epoch": 1.2258045112781955, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 25379 + }, + { + "epoch": 1.2258646616541353, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 25380 + }, + { + "epoch": 1.2259248120300752, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 25381 + }, + { + "epoch": 1.225984962406015, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 25382 + }, + { + "epoch": 1.226045112781955, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 25383 + }, + { + "epoch": 1.2261052631578948, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 25384 + }, + { + "epoch": 1.2261654135338347, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2378, + "step": 25385 + }, + { + "epoch": 1.2262255639097743, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 25386 + }, + { + "epoch": 1.2262857142857142, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 25387 + }, + { + "epoch": 1.226345864661654, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 25388 + }, + { + "epoch": 1.226406015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 25389 + }, + { + "epoch": 1.2264661654135338, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 25390 + }, + { + "epoch": 1.2265263157894737, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 25391 + }, + { + "epoch": 1.2265864661654136, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 25392 + }, + { + "epoch": 1.2266466165413534, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 25393 + }, + { + "epoch": 1.2267067669172933, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 25394 + }, + { + "epoch": 1.2267669172932332, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 25395 + }, + { + "epoch": 1.226827067669173, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 25396 + }, + { + "epoch": 1.2268872180451127, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 25397 + }, + { + "epoch": 1.2269473684210526, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 25398 + }, + { + "epoch": 1.2270075187969924, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 25399 + }, + { + "epoch": 1.2270676691729323, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 25400 + }, + { + "epoch": 1.2271278195488722, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 25401 + }, + { + "epoch": 1.227187969924812, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 25402 + }, + { + "epoch": 1.227248120300752, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 25403 + }, + { + "epoch": 1.2273082706766918, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 25404 + }, + { + "epoch": 1.2273684210526317, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 25405 + }, + { + "epoch": 1.2274285714285713, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 25406 + }, + { + "epoch": 1.2274887218045112, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 25407 + }, + { + "epoch": 1.227548872180451, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 25408 + }, + { + "epoch": 1.227609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 25409 + }, + { + "epoch": 1.2276691729323308, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 25410 + }, + { + "epoch": 1.2277293233082707, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 25411 + }, + { + "epoch": 1.2277894736842105, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 25412 + }, + { + "epoch": 1.2278496240601504, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 25413 + }, + { + "epoch": 1.2279097744360903, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 25414 + }, + { + "epoch": 1.2279699248120302, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 25415 + }, + { + "epoch": 1.22803007518797, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 25416 + }, + { + "epoch": 1.2280902255639097, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 25417 + }, + { + "epoch": 1.2281503759398495, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 25418 + }, + { + "epoch": 1.2282105263157894, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 25419 + }, + { + "epoch": 1.2282706766917293, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 25420 + }, + { + "epoch": 1.2283308270676692, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 25421 + }, + { + "epoch": 1.228390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 25422 + }, + { + "epoch": 1.228451127819549, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 25423 + }, + { + "epoch": 1.2285112781954888, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 25424 + }, + { + "epoch": 1.2285714285714286, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 25425 + }, + { + "epoch": 1.2286315789473683, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 25426 + }, + { + "epoch": 1.2286917293233084, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 25427 + }, + { + "epoch": 1.228751879699248, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 25428 + }, + { + "epoch": 1.228812030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 25429 + }, + { + "epoch": 1.2288721804511278, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 25430 + }, + { + "epoch": 1.2289323308270677, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 25431 + }, + { + "epoch": 1.2289924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 25432 + }, + { + "epoch": 1.2290526315789474, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 25433 + }, + { + "epoch": 1.2291127819548873, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 25434 + }, + { + "epoch": 1.2291729323308271, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 25435 + }, + { + "epoch": 1.229233082706767, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 25436 + }, + { + "epoch": 1.2292932330827067, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 25437 + }, + { + "epoch": 1.2293533834586465, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 25438 + }, + { + "epoch": 1.2294135338345864, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 25439 + }, + { + "epoch": 1.2294736842105263, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 25440 + }, + { + "epoch": 1.2295338345864661, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 25441 + }, + { + "epoch": 1.229593984962406, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.3659, + "step": 25442 + }, + { + "epoch": 1.2296541353383459, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 25443 + }, + { + "epoch": 1.2297142857142858, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 25444 + }, + { + "epoch": 1.2297744360902256, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 25445 + }, + { + "epoch": 1.2298345864661655, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 25446 + }, + { + "epoch": 1.2298947368421054, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 25447 + }, + { + "epoch": 1.229954887218045, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 25448 + }, + { + "epoch": 1.230015037593985, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 25449 + }, + { + "epoch": 1.2300751879699248, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 25450 + }, + { + "epoch": 1.2301353383458646, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 25451 + }, + { + "epoch": 1.2301954887218045, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 25452 + }, + { + "epoch": 1.2302556390977444, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 25453 + }, + { + "epoch": 1.2303157894736843, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 25454 + }, + { + "epoch": 1.2303759398496241, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 25455 + }, + { + "epoch": 1.230436090225564, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 25456 + }, + { + "epoch": 1.2304962406015036, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 25457 + }, + { + "epoch": 1.2305563909774435, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 25458 + }, + { + "epoch": 1.2306165413533834, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 25459 + }, + { + "epoch": 1.2306766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 25460 + }, + { + "epoch": 1.2307368421052631, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 25461 + }, + { + "epoch": 1.230796992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 25462 + }, + { + "epoch": 1.2308571428571429, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 25463 + }, + { + "epoch": 1.2309172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 25464 + }, + { + "epoch": 1.2309774436090226, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 25465 + }, + { + "epoch": 1.2310375939849625, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 25466 + }, + { + "epoch": 1.2310977443609024, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 25467 + }, + { + "epoch": 1.231157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 25468 + }, + { + "epoch": 1.2312180451127819, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 25469 + }, + { + "epoch": 1.2312781954887217, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1564, + "step": 25470 + }, + { + "epoch": 1.2313383458646616, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 25471 + }, + { + "epoch": 1.2313984962406015, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 25472 + }, + { + "epoch": 1.2314586466165414, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 25473 + }, + { + "epoch": 1.2315187969924812, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 25474 + }, + { + "epoch": 1.231578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 25475 + }, + { + "epoch": 1.231639097744361, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 25476 + }, + { + "epoch": 1.2316992481203006, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 25477 + }, + { + "epoch": 1.2317593984962407, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 25478 + }, + { + "epoch": 1.2318195488721804, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 25479 + }, + { + "epoch": 1.2318796992481202, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 25480 + }, + { + "epoch": 1.2319398496240601, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 25481 + }, + { + "epoch": 1.232, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1749, + "step": 25482 + }, + { + "epoch": 1.2320601503759399, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 25483 + }, + { + "epoch": 1.2321203007518797, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 25484 + }, + { + "epoch": 1.2321804511278196, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 25485 + }, + { + "epoch": 1.2322406015037595, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 25486 + }, + { + "epoch": 1.2323007518796993, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 25487 + }, + { + "epoch": 1.232360902255639, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 25488 + }, + { + "epoch": 1.2324210526315789, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 25489 + }, + { + "epoch": 1.2324812030075187, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2039, + "step": 25490 + }, + { + "epoch": 1.2325413533834586, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 25491 + }, + { + "epoch": 1.2326015037593985, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2085, + "step": 25492 + }, + { + "epoch": 1.2326616541353383, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 25493 + }, + { + "epoch": 1.2327218045112782, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 25494 + }, + { + "epoch": 1.232781954887218, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 25495 + }, + { + "epoch": 1.232842105263158, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 25496 + }, + { + "epoch": 1.2329022556390978, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 25497 + }, + { + "epoch": 1.2329624060150377, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 25498 + }, + { + "epoch": 1.2330225563909774, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 25499 + }, + { + "epoch": 1.2330827067669172, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 25500 + }, + { + "epoch": 1.233142857142857, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 25501 + }, + { + "epoch": 1.233203007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 25502 + }, + { + "epoch": 1.2332631578947368, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 25503 + }, + { + "epoch": 1.2333233082706767, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 25504 + }, + { + "epoch": 1.2333834586466166, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 25505 + }, + { + "epoch": 1.2334436090225565, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2394, + "step": 25506 + }, + { + "epoch": 1.2335037593984963, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 25507 + }, + { + "epoch": 1.233563909774436, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 25508 + }, + { + "epoch": 1.2336240601503758, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 25509 + }, + { + "epoch": 1.2336842105263157, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 25510 + }, + { + "epoch": 1.2337443609022556, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 25511 + }, + { + "epoch": 1.2338045112781955, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 25512 + }, + { + "epoch": 1.2338646616541353, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 25513 + }, + { + "epoch": 1.2339248120300752, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 25514 + }, + { + "epoch": 1.233984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.2297, + "step": 25515 + }, + { + "epoch": 1.234045112781955, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 25516 + }, + { + "epoch": 1.2341052631578948, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 25517 + }, + { + "epoch": 1.2341654135338347, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 25518 + }, + { + "epoch": 1.2342255639097743, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 25519 + }, + { + "epoch": 1.2342857142857142, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 25520 + }, + { + "epoch": 1.234345864661654, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 25521 + }, + { + "epoch": 1.234406015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 25522 + }, + { + "epoch": 1.2344661654135338, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 25523 + }, + { + "epoch": 1.2345263157894737, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 25524 + }, + { + "epoch": 1.2345864661654136, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 25525 + }, + { + "epoch": 1.2346466165413534, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 25526 + }, + { + "epoch": 1.2347067669172933, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 25527 + }, + { + "epoch": 1.2347669172932332, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 25528 + }, + { + "epoch": 1.234827067669173, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 25529 + }, + { + "epoch": 1.2348872180451127, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 25530 + }, + { + "epoch": 1.2349473684210526, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 25531 + }, + { + "epoch": 1.2350075187969924, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 25532 + }, + { + "epoch": 1.2350676691729323, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1716, + "step": 25533 + }, + { + "epoch": 1.2351278195488722, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 25534 + }, + { + "epoch": 1.235187969924812, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 25535 + }, + { + "epoch": 1.235248120300752, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 25536 + }, + { + "epoch": 1.2353082706766918, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 25537 + }, + { + "epoch": 1.2353684210526317, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 25538 + }, + { + "epoch": 1.2354285714285713, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 25539 + }, + { + "epoch": 1.2354887218045112, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 25540 + }, + { + "epoch": 1.235548872180451, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 25541 + }, + { + "epoch": 1.235609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 25542 + }, + { + "epoch": 1.2356691729323308, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 25543 + }, + { + "epoch": 1.2357293233082707, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 25544 + }, + { + "epoch": 1.2357894736842105, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 25545 + }, + { + "epoch": 1.2358496240601504, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 25546 + }, + { + "epoch": 1.2359097744360903, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 25547 + }, + { + "epoch": 1.2359699248120302, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 25548 + }, + { + "epoch": 1.23603007518797, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 25549 + }, + { + "epoch": 1.2360902255639097, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 25550 + }, + { + "epoch": 1.2361503759398496, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 25551 + }, + { + "epoch": 1.2362105263157894, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 25552 + }, + { + "epoch": 1.2362706766917293, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 25553 + }, + { + "epoch": 1.2363308270676692, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 25554 + }, + { + "epoch": 1.236390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 25555 + }, + { + "epoch": 1.236451127819549, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 25556 + }, + { + "epoch": 1.2365112781954888, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 25557 + }, + { + "epoch": 1.2365714285714287, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 25558 + }, + { + "epoch": 1.2366315789473683, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 25559 + }, + { + "epoch": 1.2366917293233082, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 25560 + }, + { + "epoch": 1.236751879699248, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 25561 + }, + { + "epoch": 1.236812030075188, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 25562 + }, + { + "epoch": 1.2368721804511278, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 25563 + }, + { + "epoch": 1.2369323308270677, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 25564 + }, + { + "epoch": 1.2369924812030075, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 25565 + }, + { + "epoch": 1.2370526315789474, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 25566 + }, + { + "epoch": 1.2371127819548873, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 25567 + }, + { + "epoch": 1.2371729323308271, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 25568 + }, + { + "epoch": 1.237233082706767, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 25569 + }, + { + "epoch": 1.2372932330827067, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 25570 + }, + { + "epoch": 1.2373533834586465, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 25571 + }, + { + "epoch": 1.2374135338345864, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 25572 + }, + { + "epoch": 1.2374736842105263, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 25573 + }, + { + "epoch": 1.2375338345864662, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 25574 + }, + { + "epoch": 1.237593984962406, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 25575 + }, + { + "epoch": 1.237654135338346, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 25576 + }, + { + "epoch": 1.2377142857142858, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 25577 + }, + { + "epoch": 1.2377744360902256, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 25578 + }, + { + "epoch": 1.2378345864661655, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 25579 + }, + { + "epoch": 1.2378947368421054, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 25580 + }, + { + "epoch": 1.237954887218045, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 25581 + }, + { + "epoch": 1.238015037593985, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 25582 + }, + { + "epoch": 1.2380751879699248, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 25583 + }, + { + "epoch": 1.2381353383458646, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 25584 + }, + { + "epoch": 1.2381954887218045, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 25585 + }, + { + "epoch": 1.2382556390977444, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 25586 + }, + { + "epoch": 1.2383157894736843, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 25587 + }, + { + "epoch": 1.2383759398496241, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 25588 + }, + { + "epoch": 1.238436090225564, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 25589 + }, + { + "epoch": 1.2384962406015037, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 25590 + }, + { + "epoch": 1.2385563909774435, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 25591 + }, + { + "epoch": 1.2386165413533834, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 25592 + }, + { + "epoch": 1.2386766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 25593 + }, + { + "epoch": 1.2387368421052631, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.225, + "step": 25594 + }, + { + "epoch": 1.238796992481203, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 25595 + }, + { + "epoch": 1.2388571428571429, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 25596 + }, + { + "epoch": 1.2389172932330828, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 25597 + }, + { + "epoch": 1.2389774436090226, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 25598 + }, + { + "epoch": 1.2390375939849625, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 25599 + }, + { + "epoch": 1.2390977443609024, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 25600 + }, + { + "epoch": 1.239157894736842, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 25601 + }, + { + "epoch": 1.2392180451127819, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 25602 + }, + { + "epoch": 1.2392781954887218, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 25603 + }, + { + "epoch": 1.2393383458646616, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 25604 + }, + { + "epoch": 1.2393984962406015, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 25605 + }, + { + "epoch": 1.2394586466165414, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1465, + "step": 25606 + }, + { + "epoch": 1.2395187969924812, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 0.9495, + "step": 25607 + }, + { + "epoch": 1.2395789473684211, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 25608 + }, + { + "epoch": 1.239639097744361, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 25609 + }, + { + "epoch": 1.2396992481203006, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 25610 + }, + { + "epoch": 1.2397593984962405, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 25611 + }, + { + "epoch": 1.2398195488721804, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 25612 + }, + { + "epoch": 1.2398796992481202, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 25613 + }, + { + "epoch": 1.2399398496240601, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 25614 + }, + { + "epoch": 1.24, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 25615 + }, + { + "epoch": 1.2400601503759399, + "grad_norm": 3.375, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 25616 + }, + { + "epoch": 1.2401203007518797, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 25617 + }, + { + "epoch": 1.2401804511278196, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 25618 + }, + { + "epoch": 1.2402406015037595, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 25619 + }, + { + "epoch": 1.2403007518796993, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 25620 + }, + { + "epoch": 1.240360902255639, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 25621 + }, + { + "epoch": 1.2404210526315789, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2349, + "step": 25622 + }, + { + "epoch": 1.2404812030075187, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 25623 + }, + { + "epoch": 1.2405413533834586, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 25624 + }, + { + "epoch": 1.2406015037593985, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 25625 + }, + { + "epoch": 1.2406616541353384, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 25626 + }, + { + "epoch": 1.2407218045112782, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 25627 + }, + { + "epoch": 1.240781954887218, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 25628 + }, + { + "epoch": 1.240842105263158, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 25629 + }, + { + "epoch": 1.2409022556390978, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 25630 + }, + { + "epoch": 1.2409624060150377, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 25631 + }, + { + "epoch": 1.2410225563909774, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 25632 + }, + { + "epoch": 1.2410827067669172, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 25633 + }, + { + "epoch": 1.241142857142857, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 25634 + }, + { + "epoch": 1.241203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 25635 + }, + { + "epoch": 1.2412631578947368, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 25636 + }, + { + "epoch": 1.2413233082706767, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 25637 + }, + { + "epoch": 1.2413834586466166, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.2671, + "step": 25638 + }, + { + "epoch": 1.2414436090225565, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 25639 + }, + { + "epoch": 1.2415037593984963, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 25640 + }, + { + "epoch": 1.241563909774436, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 25641 + }, + { + "epoch": 1.2416240601503759, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 25642 + }, + { + "epoch": 1.2416842105263157, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 25643 + }, + { + "epoch": 1.2417443609022556, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 25644 + }, + { + "epoch": 1.2418045112781955, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 25645 + }, + { + "epoch": 1.2418646616541353, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 25646 + }, + { + "epoch": 1.2419248120300752, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 25647 + }, + { + "epoch": 1.241984962406015, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 25648 + }, + { + "epoch": 1.242045112781955, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2291, + "step": 25649 + }, + { + "epoch": 1.2421052631578948, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 25650 + }, + { + "epoch": 1.2421654135338347, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 25651 + }, + { + "epoch": 1.2422255639097743, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 25652 + }, + { + "epoch": 1.2422857142857142, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 25653 + }, + { + "epoch": 1.242345864661654, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 25654 + }, + { + "epoch": 1.242406015037594, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 25655 + }, + { + "epoch": 1.2424661654135338, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 25656 + }, + { + "epoch": 1.2425263157894737, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 25657 + }, + { + "epoch": 1.2425864661654136, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 25658 + }, + { + "epoch": 1.2426466165413534, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 25659 + }, + { + "epoch": 1.2427067669172933, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 25660 + }, + { + "epoch": 1.242766917293233, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 25661 + }, + { + "epoch": 1.2428270676691728, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 25662 + }, + { + "epoch": 1.2428872180451127, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 25663 + }, + { + "epoch": 1.2429473684210526, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 25664 + }, + { + "epoch": 1.2430075187969924, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 25665 + }, + { + "epoch": 1.2430676691729323, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 25666 + }, + { + "epoch": 1.2431278195488722, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 25667 + }, + { + "epoch": 1.243187969924812, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 25668 + }, + { + "epoch": 1.243248120300752, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 25669 + }, + { + "epoch": 1.2433082706766918, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 25670 + }, + { + "epoch": 1.2433684210526317, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 25671 + }, + { + "epoch": 1.2434285714285713, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 25672 + }, + { + "epoch": 1.2434887218045112, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 25673 + }, + { + "epoch": 1.243548872180451, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 25674 + }, + { + "epoch": 1.243609022556391, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 25675 + }, + { + "epoch": 1.2436691729323308, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 25676 + }, + { + "epoch": 1.2437293233082707, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 25677 + }, + { + "epoch": 1.2437894736842106, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 25678 + }, + { + "epoch": 1.2438496240601504, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 25679 + }, + { + "epoch": 1.2439097744360903, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 25680 + }, + { + "epoch": 1.2439699248120302, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 25681 + }, + { + "epoch": 1.24403007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 25682 + }, + { + "epoch": 1.2440902255639097, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 25683 + }, + { + "epoch": 1.2441503759398496, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2704, + "step": 25684 + }, + { + "epoch": 1.2442105263157894, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 25685 + }, + { + "epoch": 1.2442706766917293, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 25686 + }, + { + "epoch": 1.2443308270676692, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 25687 + }, + { + "epoch": 1.244390977443609, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 25688 + }, + { + "epoch": 1.244451127819549, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 25689 + }, + { + "epoch": 1.2445112781954888, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1545, + "step": 25690 + }, + { + "epoch": 1.2445714285714287, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 25691 + }, + { + "epoch": 1.2446315789473683, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 25692 + }, + { + "epoch": 1.2446917293233082, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 25693 + }, + { + "epoch": 1.244751879699248, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 25694 + }, + { + "epoch": 1.244812030075188, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 25695 + }, + { + "epoch": 1.2448721804511278, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 25696 + }, + { + "epoch": 1.2449323308270677, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 25697 + }, + { + "epoch": 1.2449924812030075, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 25698 + }, + { + "epoch": 1.2450526315789474, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 25699 + }, + { + "epoch": 1.2451127819548873, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 25700 + }, + { + "epoch": 1.2451729323308272, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9333, + "step": 25701 + }, + { + "epoch": 1.245233082706767, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 25702 + }, + { + "epoch": 1.2452932330827067, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 25703 + }, + { + "epoch": 1.2453533834586465, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 25704 + }, + { + "epoch": 1.2454135338345864, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 25705 + }, + { + "epoch": 1.2454736842105263, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 25706 + }, + { + "epoch": 1.2455338345864662, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 25707 + }, + { + "epoch": 1.245593984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 25708 + }, + { + "epoch": 1.245654135338346, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 25709 + }, + { + "epoch": 1.2457142857142858, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 25710 + }, + { + "epoch": 1.2457744360902256, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 25711 + }, + { + "epoch": 1.2458345864661653, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 25712 + }, + { + "epoch": 1.2458947368421052, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 25713 + }, + { + "epoch": 1.245954887218045, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 25714 + }, + { + "epoch": 1.246015037593985, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2327, + "step": 25715 + }, + { + "epoch": 1.2460751879699248, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 25716 + }, + { + "epoch": 1.2461353383458647, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 25717 + }, + { + "epoch": 1.2461954887218045, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 25718 + }, + { + "epoch": 1.2462556390977444, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 25719 + }, + { + "epoch": 1.2463157894736843, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 25720 + }, + { + "epoch": 1.2463759398496241, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 25721 + }, + { + "epoch": 1.246436090225564, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 25722 + }, + { + "epoch": 1.2464962406015037, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 25723 + }, + { + "epoch": 1.2465563909774435, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 25724 + }, + { + "epoch": 1.2466165413533834, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 25725 + }, + { + "epoch": 1.2466766917293233, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 25726 + }, + { + "epoch": 1.2467368421052631, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 25727 + }, + { + "epoch": 1.246796992481203, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 25728 + }, + { + "epoch": 1.2468571428571429, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 25729 + }, + { + "epoch": 1.2469172932330828, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 25730 + }, + { + "epoch": 1.2469774436090226, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 25731 + }, + { + "epoch": 1.2470375939849625, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 25732 + }, + { + "epoch": 1.2470977443609024, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 25733 + }, + { + "epoch": 1.247157894736842, + "grad_norm": 1.53125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 25734 + }, + { + "epoch": 1.247218045112782, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 25735 + }, + { + "epoch": 1.2472781954887218, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 25736 + }, + { + "epoch": 1.2473383458646616, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 25737 + }, + { + "epoch": 1.2473984962406015, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1726, + "step": 25738 + }, + { + "epoch": 1.2474586466165414, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 25739 + }, + { + "epoch": 1.2475187969924812, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 25740 + }, + { + "epoch": 1.2475789473684211, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 25741 + }, + { + "epoch": 1.247639097744361, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 25742 + }, + { + "epoch": 1.2476992481203006, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 25743 + }, + { + "epoch": 1.2477593984962405, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 25744 + }, + { + "epoch": 1.2478195488721804, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 25745 + }, + { + "epoch": 1.2478796992481203, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 25746 + }, + { + "epoch": 1.2479398496240601, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 25747 + }, + { + "epoch": 1.248, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 25748 + }, + { + "epoch": 1.2480601503759399, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 25749 + }, + { + "epoch": 1.2481203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 25750 + }, + { + "epoch": 1.2481804511278196, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 25751 + }, + { + "epoch": 1.2482406015037595, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 25752 + }, + { + "epoch": 1.2483007518796994, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 25753 + }, + { + "epoch": 1.248360902255639, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 25754 + }, + { + "epoch": 1.2484210526315789, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 25755 + }, + { + "epoch": 1.2484812030075187, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 25756 + }, + { + "epoch": 1.2485413533834586, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 25757 + }, + { + "epoch": 1.2486015037593985, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 25758 + }, + { + "epoch": 1.2486616541353384, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 25759 + }, + { + "epoch": 1.2487218045112782, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 25760 + }, + { + "epoch": 1.248781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 25761 + }, + { + "epoch": 1.248842105263158, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 25762 + }, + { + "epoch": 1.2489022556390976, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 25763 + }, + { + "epoch": 1.2489624060150377, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 25764 + }, + { + "epoch": 1.2490225563909774, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2243, + "step": 25765 + }, + { + "epoch": 1.2490827067669172, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 25766 + }, + { + "epoch": 1.249142857142857, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2196, + "step": 25767 + }, + { + "epoch": 1.249203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 25768 + }, + { + "epoch": 1.2492631578947369, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 25769 + }, + { + "epoch": 1.2493233082706767, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 25770 + }, + { + "epoch": 1.2493834586466166, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 25771 + }, + { + "epoch": 1.2494436090225565, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 25772 + }, + { + "epoch": 1.2495037593984963, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 25773 + }, + { + "epoch": 1.249563909774436, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 25774 + }, + { + "epoch": 1.2496240601503759, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 25775 + }, + { + "epoch": 1.2496842105263157, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 25776 + }, + { + "epoch": 1.2497443609022556, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1732, + "step": 25777 + }, + { + "epoch": 1.2498045112781955, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 25778 + }, + { + "epoch": 1.2498646616541353, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 25779 + }, + { + "epoch": 1.2499248120300752, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 25780 + }, + { + "epoch": 1.249984962406015, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 25781 + }, + { + "epoch": 1.250045112781955, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 25782 + }, + { + "epoch": 1.2501052631578946, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 25783 + }, + { + "epoch": 1.2501654135338347, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 25784 + }, + { + "epoch": 1.2502255639097744, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 25785 + }, + { + "epoch": 1.2502857142857142, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 25786 + }, + { + "epoch": 1.250345864661654, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 25787 + }, + { + "epoch": 1.250406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 25788 + }, + { + "epoch": 1.2504661654135338, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 25789 + }, + { + "epoch": 1.2505263157894737, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 25790 + }, + { + "epoch": 1.2505864661654136, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1585, + "step": 25791 + }, + { + "epoch": 1.2506466165413535, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 25792 + }, + { + "epoch": 1.2507067669172933, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 25793 + }, + { + "epoch": 1.250766917293233, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 25794 + }, + { + "epoch": 1.250827067669173, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 25795 + }, + { + "epoch": 1.2508872180451127, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 25796 + }, + { + "epoch": 1.2509473684210526, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 25797 + }, + { + "epoch": 1.2510075187969925, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 25798 + }, + { + "epoch": 1.2510676691729323, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 25799 + }, + { + "epoch": 1.2511278195488722, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 25800 + }, + { + "epoch": 1.251187969924812, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 25801 + }, + { + "epoch": 1.251248120300752, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1847, + "step": 25802 + }, + { + "epoch": 1.2513082706766916, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 25803 + }, + { + "epoch": 1.2513684210526317, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 25804 + }, + { + "epoch": 1.2514285714285713, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 25805 + }, + { + "epoch": 1.2514887218045112, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 25806 + }, + { + "epoch": 1.251548872180451, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 25807 + }, + { + "epoch": 1.251609022556391, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 25808 + }, + { + "epoch": 1.2516691729323308, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 25809 + }, + { + "epoch": 1.2517293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 25810 + }, + { + "epoch": 1.2517894736842106, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 25811 + }, + { + "epoch": 1.2518496240601504, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 25812 + }, + { + "epoch": 1.2519097744360903, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 25813 + }, + { + "epoch": 1.25196992481203, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 25814 + }, + { + "epoch": 1.25203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 25815 + }, + { + "epoch": 1.2520902255639097, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 25816 + }, + { + "epoch": 1.2521503759398496, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 25817 + }, + { + "epoch": 1.2522105263157894, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 25818 + }, + { + "epoch": 1.2522706766917293, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 25819 + }, + { + "epoch": 1.2523308270676692, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 25820 + }, + { + "epoch": 1.252390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 25821 + }, + { + "epoch": 1.252451127819549, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 25822 + }, + { + "epoch": 1.2525112781954888, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 25823 + }, + { + "epoch": 1.2525714285714287, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 25824 + }, + { + "epoch": 1.2526315789473683, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 25825 + }, + { + "epoch": 1.2526917293233082, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 25826 + }, + { + "epoch": 1.252751879699248, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 25827 + }, + { + "epoch": 1.252812030075188, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 25828 + }, + { + "epoch": 1.2528721804511278, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 25829 + }, + { + "epoch": 1.2529323308270677, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 25830 + }, + { + "epoch": 1.2529924812030075, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 25831 + }, + { + "epoch": 1.2530526315789474, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 25832 + }, + { + "epoch": 1.2531127819548873, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 25833 + }, + { + "epoch": 1.253172932330827, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1264, + "step": 25834 + }, + { + "epoch": 1.253233082706767, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 25835 + }, + { + "epoch": 1.2532932330827067, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 25836 + }, + { + "epoch": 1.2533533834586466, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 25837 + }, + { + "epoch": 1.2534135338345864, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 25838 + }, + { + "epoch": 1.2534736842105263, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2073, + "step": 25839 + }, + { + "epoch": 1.2535338345864662, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 25840 + }, + { + "epoch": 1.253593984962406, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 25841 + }, + { + "epoch": 1.253654135338346, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 25842 + }, + { + "epoch": 1.2537142857142858, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 25843 + }, + { + "epoch": 1.2537744360902257, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 25844 + }, + { + "epoch": 1.2538345864661653, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 25845 + }, + { + "epoch": 1.2538947368421054, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 25846 + }, + { + "epoch": 1.253954887218045, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.928, + "step": 25847 + }, + { + "epoch": 1.254015037593985, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 25848 + }, + { + "epoch": 1.2540751879699248, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 25849 + }, + { + "epoch": 1.2541353383458647, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 25850 + }, + { + "epoch": 1.2541954887218045, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 25851 + }, + { + "epoch": 1.2542556390977444, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 25852 + }, + { + "epoch": 1.2543157894736843, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 25853 + }, + { + "epoch": 1.254375939849624, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 25854 + }, + { + "epoch": 1.254436090225564, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 25855 + }, + { + "epoch": 1.2544962406015037, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 25856 + }, + { + "epoch": 1.2545563909774435, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 25857 + }, + { + "epoch": 1.2546165413533834, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 25858 + }, + { + "epoch": 1.2546766917293233, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 25859 + }, + { + "epoch": 1.2547368421052632, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 25860 + }, + { + "epoch": 1.254796992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 25861 + }, + { + "epoch": 1.254857142857143, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 25862 + }, + { + "epoch": 1.2549172932330828, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 25863 + }, + { + "epoch": 1.2549774436090226, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 25864 + }, + { + "epoch": 1.2550375939849623, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 25865 + }, + { + "epoch": 1.2550977443609024, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 25866 + }, + { + "epoch": 1.255157894736842, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 25867 + }, + { + "epoch": 1.255218045112782, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 25868 + }, + { + "epoch": 1.2552781954887218, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 25869 + }, + { + "epoch": 1.2553383458646616, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 25870 + }, + { + "epoch": 1.2553984962406015, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 25871 + }, + { + "epoch": 1.2554586466165414, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 25872 + }, + { + "epoch": 1.2555187969924813, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 25873 + }, + { + "epoch": 1.2555789473684211, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 25874 + }, + { + "epoch": 1.255639097744361, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9406, + "step": 25875 + }, + { + "epoch": 1.2556992481203006, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 25876 + }, + { + "epoch": 1.2557593984962405, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 25877 + }, + { + "epoch": 1.2558195488721804, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 25878 + }, + { + "epoch": 1.2558796992481203, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 25879 + }, + { + "epoch": 1.2559398496240601, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 25880 + }, + { + "epoch": 1.256, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 25881 + }, + { + "epoch": 1.2560601503759399, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2022, + "step": 25882 + }, + { + "epoch": 1.2561203007518797, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 25883 + }, + { + "epoch": 1.2561804511278196, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0994, + "step": 25884 + }, + { + "epoch": 1.2562406015037593, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 25885 + }, + { + "epoch": 1.2563007518796994, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 25886 + }, + { + "epoch": 1.256360902255639, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 25887 + }, + { + "epoch": 1.2564210526315789, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1722, + "step": 25888 + }, + { + "epoch": 1.2564812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 25889 + }, + { + "epoch": 1.2565413533834586, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 25890 + }, + { + "epoch": 1.2566015037593985, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 25891 + }, + { + "epoch": 1.2566616541353384, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 25892 + }, + { + "epoch": 1.2567218045112782, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 25893 + }, + { + "epoch": 1.2567819548872181, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2157, + "step": 25894 + }, + { + "epoch": 1.256842105263158, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 25895 + }, + { + "epoch": 1.2569022556390976, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 25896 + }, + { + "epoch": 1.2569624060150377, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 25897 + }, + { + "epoch": 1.2570225563909774, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 25898 + }, + { + "epoch": 1.2570827067669172, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 25899 + }, + { + "epoch": 1.2571428571428571, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 25900 + }, + { + "epoch": 1.257203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 25901 + }, + { + "epoch": 1.2572631578947369, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 25902 + }, + { + "epoch": 1.2573233082706767, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 25903 + }, + { + "epoch": 1.2573834586466166, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 25904 + }, + { + "epoch": 1.2574436090225563, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 25905 + }, + { + "epoch": 1.2575037593984963, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 25906 + }, + { + "epoch": 1.257563909774436, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 25907 + }, + { + "epoch": 1.2576240601503759, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 25908 + }, + { + "epoch": 1.2576842105263157, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 25909 + }, + { + "epoch": 1.2577443609022556, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 25910 + }, + { + "epoch": 1.2578045112781955, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 25911 + }, + { + "epoch": 1.2578646616541354, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 25912 + }, + { + "epoch": 1.2579248120300752, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 25913 + }, + { + "epoch": 1.257984962406015, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 25914 + }, + { + "epoch": 1.258045112781955, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 25915 + }, + { + "epoch": 1.2581052631578946, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 25916 + }, + { + "epoch": 1.2581654135338347, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 25917 + }, + { + "epoch": 1.2582255639097744, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 25918 + }, + { + "epoch": 1.2582857142857142, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 25919 + }, + { + "epoch": 1.258345864661654, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 25920 + }, + { + "epoch": 1.258406015037594, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 25921 + }, + { + "epoch": 1.2584661654135338, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 25922 + }, + { + "epoch": 1.2585263157894737, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 25923 + }, + { + "epoch": 1.2585864661654136, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 25924 + }, + { + "epoch": 1.2586466165413535, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 25925 + }, + { + "epoch": 1.2587067669172933, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 25926 + }, + { + "epoch": 1.258766917293233, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 25927 + }, + { + "epoch": 1.2588270676691729, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 25928 + }, + { + "epoch": 1.2588872180451127, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 25929 + }, + { + "epoch": 1.2589473684210526, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 25930 + }, + { + "epoch": 1.2590075187969925, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 25931 + }, + { + "epoch": 1.2590676691729323, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 25932 + }, + { + "epoch": 1.2591278195488722, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 25933 + }, + { + "epoch": 1.259187969924812, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 25934 + }, + { + "epoch": 1.259248120300752, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 25935 + }, + { + "epoch": 1.2593082706766916, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 25936 + }, + { + "epoch": 1.2593684210526317, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 25937 + }, + { + "epoch": 1.2594285714285713, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 25938 + }, + { + "epoch": 1.2594887218045112, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 25939 + }, + { + "epoch": 1.259548872180451, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 25940 + }, + { + "epoch": 1.259609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 25941 + }, + { + "epoch": 1.2596691729323308, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 25942 + }, + { + "epoch": 1.2597293233082707, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 25943 + }, + { + "epoch": 1.2597894736842106, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 25944 + }, + { + "epoch": 1.2598496240601504, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 25945 + }, + { + "epoch": 1.2599097744360903, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 25946 + }, + { + "epoch": 1.25996992481203, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 25947 + }, + { + "epoch": 1.26003007518797, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 25948 + }, + { + "epoch": 1.2600902255639097, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 25949 + }, + { + "epoch": 1.2601503759398496, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 25950 + }, + { + "epoch": 1.2602105263157894, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 25951 + }, + { + "epoch": 1.2602706766917293, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 25952 + }, + { + "epoch": 1.2603308270676692, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 25953 + }, + { + "epoch": 1.260390977443609, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 25954 + }, + { + "epoch": 1.260451127819549, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 25955 + }, + { + "epoch": 1.2605112781954888, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 25956 + }, + { + "epoch": 1.2605714285714287, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 25957 + }, + { + "epoch": 1.2606315789473683, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1605, + "step": 25958 + }, + { + "epoch": 1.2606917293233082, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2034, + "step": 25959 + }, + { + "epoch": 1.260751879699248, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 25960 + }, + { + "epoch": 1.260812030075188, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 25961 + }, + { + "epoch": 1.2608721804511278, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 25962 + }, + { + "epoch": 1.2609323308270677, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 25963 + }, + { + "epoch": 1.2609924812030076, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 25964 + }, + { + "epoch": 1.2610526315789474, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 25965 + }, + { + "epoch": 1.2611127819548873, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 25966 + }, + { + "epoch": 1.261172932330827, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 25967 + }, + { + "epoch": 1.261233082706767, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 25968 + }, + { + "epoch": 1.2612932330827067, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 25969 + }, + { + "epoch": 1.2613533834586466, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 25970 + }, + { + "epoch": 1.2614135338345864, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 25971 + }, + { + "epoch": 1.2614736842105263, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 25972 + }, + { + "epoch": 1.2615338345864662, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 25973 + }, + { + "epoch": 1.261593984962406, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 25974 + }, + { + "epoch": 1.261654135338346, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 25975 + }, + { + "epoch": 1.2617142857142858, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 25976 + }, + { + "epoch": 1.2617744360902257, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 25977 + }, + { + "epoch": 1.2618345864661653, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 25978 + }, + { + "epoch": 1.2618947368421052, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2074, + "step": 25979 + }, + { + "epoch": 1.261954887218045, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 25980 + }, + { + "epoch": 1.262015037593985, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1658, + "step": 25981 + }, + { + "epoch": 1.2620751879699248, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 25982 + }, + { + "epoch": 1.2621353383458647, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 25983 + }, + { + "epoch": 1.2621954887218045, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 25984 + }, + { + "epoch": 1.2622556390977444, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 25985 + }, + { + "epoch": 1.2623157894736843, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 25986 + }, + { + "epoch": 1.262375939849624, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 25987 + }, + { + "epoch": 1.262436090225564, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 25988 + }, + { + "epoch": 1.2624962406015037, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 25989 + }, + { + "epoch": 1.2625563909774435, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 25990 + }, + { + "epoch": 1.2626165413533834, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 25991 + }, + { + "epoch": 1.2626766917293233, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 25992 + }, + { + "epoch": 1.2627368421052632, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 25993 + }, + { + "epoch": 1.262796992481203, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 25994 + }, + { + "epoch": 1.262857142857143, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 25995 + }, + { + "epoch": 1.2629172932330828, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 25996 + }, + { + "epoch": 1.2629774436090226, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 25997 + }, + { + "epoch": 1.2630375939849623, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 25998 + }, + { + "epoch": 1.2630977443609024, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 25999 + }, + { + "epoch": 1.263157894736842, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 26000 + }, + { + "epoch": 1.263218045112782, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 26001 + }, + { + "epoch": 1.2632781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 26002 + }, + { + "epoch": 1.2633383458646616, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 26003 + }, + { + "epoch": 1.2633984962406015, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 26004 + }, + { + "epoch": 1.2634586466165414, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 26005 + }, + { + "epoch": 1.2635187969924813, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 26006 + }, + { + "epoch": 1.2635789473684211, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1631, + "step": 26007 + }, + { + "epoch": 1.263639097744361, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 26008 + }, + { + "epoch": 1.2636992481203007, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 26009 + }, + { + "epoch": 1.2637593984962405, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 26010 + }, + { + "epoch": 1.2638195488721804, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 26011 + }, + { + "epoch": 1.2638796992481203, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 26012 + }, + { + "epoch": 1.2639398496240601, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 26013 + }, + { + "epoch": 1.264, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 26014 + }, + { + "epoch": 1.2640601503759399, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 26015 + }, + { + "epoch": 1.2641203007518798, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 26016 + }, + { + "epoch": 1.2641804511278196, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 26017 + }, + { + "epoch": 1.2642406015037593, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 26018 + }, + { + "epoch": 1.2643007518796994, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 26019 + }, + { + "epoch": 1.264360902255639, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 26020 + }, + { + "epoch": 1.264421052631579, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.2632, + "step": 26021 + }, + { + "epoch": 1.2644812030075188, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 26022 + }, + { + "epoch": 1.2645413533834586, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 26023 + }, + { + "epoch": 1.2646015037593985, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 26024 + }, + { + "epoch": 1.2646616541353384, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 26025 + }, + { + "epoch": 1.2647218045112782, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 26026 + }, + { + "epoch": 1.2647819548872181, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 26027 + }, + { + "epoch": 1.264842105263158, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 26028 + }, + { + "epoch": 1.2649022556390976, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9595, + "step": 26029 + }, + { + "epoch": 1.2649624060150377, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 26030 + }, + { + "epoch": 1.2650225563909774, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 26031 + }, + { + "epoch": 1.2650827067669173, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 26032 + }, + { + "epoch": 1.2651428571428571, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 26033 + }, + { + "epoch": 1.265203007518797, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 26034 + }, + { + "epoch": 1.2652631578947369, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 26035 + }, + { + "epoch": 1.2653233082706767, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 26036 + }, + { + "epoch": 1.2653834586466166, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 26037 + }, + { + "epoch": 1.2654436090225563, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 26038 + }, + { + "epoch": 1.2655037593984964, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 26039 + }, + { + "epoch": 1.265563909774436, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 26040 + }, + { + "epoch": 1.2656240601503759, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 26041 + }, + { + "epoch": 1.2656842105263157, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 26042 + }, + { + "epoch": 1.2657443609022556, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 26043 + }, + { + "epoch": 1.2658045112781955, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 26044 + }, + { + "epoch": 1.2658646616541354, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 26045 + }, + { + "epoch": 1.2659248120300752, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 26046 + }, + { + "epoch": 1.265984962406015, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 26047 + }, + { + "epoch": 1.266045112781955, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 26048 + }, + { + "epoch": 1.2661052631578946, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 26049 + }, + { + "epoch": 1.2661654135338347, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 26050 + }, + { + "epoch": 1.2662255639097744, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 26051 + }, + { + "epoch": 1.2662857142857142, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 26052 + }, + { + "epoch": 1.266345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 26053 + }, + { + "epoch": 1.266406015037594, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 26054 + }, + { + "epoch": 1.2664661654135339, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 26055 + }, + { + "epoch": 1.2665263157894737, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 26056 + }, + { + "epoch": 1.2665864661654136, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 26057 + }, + { + "epoch": 1.2666466165413535, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 26058 + }, + { + "epoch": 1.2667067669172933, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 26059 + }, + { + "epoch": 1.266766917293233, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 26060 + }, + { + "epoch": 1.2668270676691729, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 26061 + }, + { + "epoch": 1.2668872180451127, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 26062 + }, + { + "epoch": 1.2669473684210526, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 26063 + }, + { + "epoch": 1.2670075187969925, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 26064 + }, + { + "epoch": 1.2670676691729323, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 26065 + }, + { + "epoch": 1.2671278195488722, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 26066 + }, + { + "epoch": 1.267187969924812, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 26067 + }, + { + "epoch": 1.267248120300752, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2332, + "step": 26068 + }, + { + "epoch": 1.2673082706766916, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 26069 + }, + { + "epoch": 1.2673684210526317, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 26070 + }, + { + "epoch": 1.2674285714285713, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 26071 + }, + { + "epoch": 1.2674887218045112, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2388, + "step": 26072 + }, + { + "epoch": 1.267548872180451, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 26073 + }, + { + "epoch": 1.267609022556391, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 26074 + }, + { + "epoch": 1.2676691729323308, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 26075 + }, + { + "epoch": 1.2677293233082707, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 26076 + }, + { + "epoch": 1.2677894736842106, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 26077 + }, + { + "epoch": 1.2678496240601504, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 26078 + }, + { + "epoch": 1.2679097744360903, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 26079 + }, + { + "epoch": 1.26796992481203, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 26080 + }, + { + "epoch": 1.26803007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 26081 + }, + { + "epoch": 1.2680902255639097, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 26082 + }, + { + "epoch": 1.2681503759398496, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 26083 + }, + { + "epoch": 1.2682105263157895, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 26084 + }, + { + "epoch": 1.2682706766917293, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 26085 + }, + { + "epoch": 1.2683308270676692, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2541, + "step": 26086 + }, + { + "epoch": 1.268390977443609, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 26087 + }, + { + "epoch": 1.268451127819549, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 26088 + }, + { + "epoch": 1.2685112781954886, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 26089 + }, + { + "epoch": 1.2685714285714287, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 26090 + }, + { + "epoch": 1.2686315789473683, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 26091 + }, + { + "epoch": 1.2686917293233082, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 26092 + }, + { + "epoch": 1.268751879699248, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 26093 + }, + { + "epoch": 1.268812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 26094 + }, + { + "epoch": 1.2688721804511278, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 26095 + }, + { + "epoch": 1.2689323308270677, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 26096 + }, + { + "epoch": 1.2689924812030076, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 26097 + }, + { + "epoch": 1.2690526315789474, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 26098 + }, + { + "epoch": 1.2691127819548873, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 26099 + }, + { + "epoch": 1.269172932330827, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1927, + "step": 26100 + }, + { + "epoch": 1.269233082706767, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 26101 + }, + { + "epoch": 1.2692932330827067, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 26102 + }, + { + "epoch": 1.2693533834586466, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 26103 + }, + { + "epoch": 1.2694135338345864, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 26104 + }, + { + "epoch": 1.2694736842105263, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1656, + "step": 26105 + }, + { + "epoch": 1.2695338345864662, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 26106 + }, + { + "epoch": 1.269593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 26107 + }, + { + "epoch": 1.269654135338346, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 26108 + }, + { + "epoch": 1.2697142857142858, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 26109 + }, + { + "epoch": 1.2697744360902257, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 26110 + }, + { + "epoch": 1.2698345864661653, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 26111 + }, + { + "epoch": 1.2698947368421052, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 26112 + }, + { + "epoch": 1.269954887218045, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 26113 + }, + { + "epoch": 1.270015037593985, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 26114 + }, + { + "epoch": 1.2700751879699248, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 26115 + }, + { + "epoch": 1.2701353383458647, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 26116 + }, + { + "epoch": 1.2701954887218045, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 26117 + }, + { + "epoch": 1.2702556390977444, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 26118 + }, + { + "epoch": 1.2703157894736843, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 26119 + }, + { + "epoch": 1.270375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 26120 + }, + { + "epoch": 1.270436090225564, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 26121 + }, + { + "epoch": 1.2704962406015037, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 26122 + }, + { + "epoch": 1.2705563909774436, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1582, + "step": 26123 + }, + { + "epoch": 1.2706165413533834, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 26124 + }, + { + "epoch": 1.2706766917293233, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 26125 + }, + { + "epoch": 1.2707368421052632, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 26126 + }, + { + "epoch": 1.270796992481203, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 26127 + }, + { + "epoch": 1.270857142857143, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 26128 + }, + { + "epoch": 1.2709172932330828, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 26129 + }, + { + "epoch": 1.2709774436090227, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 26130 + }, + { + "epoch": 1.2710375939849623, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 26131 + }, + { + "epoch": 1.2710977443609024, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 26132 + }, + { + "epoch": 1.271157894736842, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 26133 + }, + { + "epoch": 1.271218045112782, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2424, + "step": 26134 + }, + { + "epoch": 1.2712781954887218, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 26135 + }, + { + "epoch": 1.2713383458646617, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 26136 + }, + { + "epoch": 1.2713984962406015, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 26137 + }, + { + "epoch": 1.2714586466165414, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 26138 + }, + { + "epoch": 1.2715187969924813, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 26139 + }, + { + "epoch": 1.271578947368421, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 26140 + }, + { + "epoch": 1.271639097744361, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 26141 + }, + { + "epoch": 1.2716992481203007, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 26142 + }, + { + "epoch": 1.2717593984962405, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 26143 + }, + { + "epoch": 1.2718195488721804, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 26144 + }, + { + "epoch": 1.2718796992481203, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 26145 + }, + { + "epoch": 1.2719398496240601, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 26146 + }, + { + "epoch": 1.272, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 26147 + }, + { + "epoch": 1.27206015037594, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 26148 + }, + { + "epoch": 1.2721203007518798, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 26149 + }, + { + "epoch": 1.2721804511278196, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 26150 + }, + { + "epoch": 1.2722406015037593, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 26151 + }, + { + "epoch": 1.2723007518796994, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 26152 + }, + { + "epoch": 1.272360902255639, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 26153 + }, + { + "epoch": 1.272421052631579, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 26154 + }, + { + "epoch": 1.2724812030075188, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 26155 + }, + { + "epoch": 1.2725413533834586, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 26156 + }, + { + "epoch": 1.2726015037593985, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 26157 + }, + { + "epoch": 1.2726616541353384, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 26158 + }, + { + "epoch": 1.2727218045112783, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 26159 + }, + { + "epoch": 1.2727819548872181, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 26160 + }, + { + "epoch": 1.272842105263158, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 26161 + }, + { + "epoch": 1.2729022556390976, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 26162 + }, + { + "epoch": 1.2729624060150375, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 26163 + }, + { + "epoch": 1.2730225563909774, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 26164 + }, + { + "epoch": 1.2730827067669173, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 26165 + }, + { + "epoch": 1.2731428571428571, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 26166 + }, + { + "epoch": 1.273203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.207, + "step": 26167 + }, + { + "epoch": 1.2732631578947369, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 26168 + }, + { + "epoch": 1.2733233082706767, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 26169 + }, + { + "epoch": 1.2733834586466166, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 26170 + }, + { + "epoch": 1.2734436090225563, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 26171 + }, + { + "epoch": 1.2735037593984964, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 26172 + }, + { + "epoch": 1.273563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 26173 + }, + { + "epoch": 1.2736240601503759, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 26174 + }, + { + "epoch": 1.2736842105263158, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 26175 + }, + { + "epoch": 1.2737443609022556, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 26176 + }, + { + "epoch": 1.2738045112781955, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 26177 + }, + { + "epoch": 1.2738646616541354, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 26178 + }, + { + "epoch": 1.2739248120300752, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 26179 + }, + { + "epoch": 1.273984962406015, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 26180 + }, + { + "epoch": 1.274045112781955, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 26181 + }, + { + "epoch": 1.2741052631578946, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 26182 + }, + { + "epoch": 1.2741654135338347, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 26183 + }, + { + "epoch": 1.2742255639097744, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 26184 + }, + { + "epoch": 1.2742857142857142, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 26185 + }, + { + "epoch": 1.2743458646616541, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 26186 + }, + { + "epoch": 1.274406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 26187 + }, + { + "epoch": 1.2744661654135339, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 26188 + }, + { + "epoch": 1.2745263157894737, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 26189 + }, + { + "epoch": 1.2745864661654136, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 26190 + }, + { + "epoch": 1.2746466165413533, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 26191 + }, + { + "epoch": 1.2747067669172933, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 26192 + }, + { + "epoch": 1.274766917293233, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 26193 + }, + { + "epoch": 1.2748270676691729, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 26194 + }, + { + "epoch": 1.2748872180451127, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2143, + "step": 26195 + }, + { + "epoch": 1.2749473684210526, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 26196 + }, + { + "epoch": 1.2750075187969925, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 26197 + }, + { + "epoch": 1.2750676691729323, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 26198 + }, + { + "epoch": 1.2751278195488722, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 26199 + }, + { + "epoch": 1.275187969924812, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 26200 + }, + { + "epoch": 1.275248120300752, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 26201 + }, + { + "epoch": 1.2753082706766916, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 26202 + }, + { + "epoch": 1.2753684210526317, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 26203 + }, + { + "epoch": 1.2754285714285714, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 26204 + }, + { + "epoch": 1.2754887218045112, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 26205 + }, + { + "epoch": 1.275548872180451, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 26206 + }, + { + "epoch": 1.275609022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 26207 + }, + { + "epoch": 1.2756691729323308, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 26208 + }, + { + "epoch": 1.2757293233082707, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 26209 + }, + { + "epoch": 1.2757894736842106, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 26210 + }, + { + "epoch": 1.2758496240601505, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 26211 + }, + { + "epoch": 1.2759097744360903, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 26212 + }, + { + "epoch": 1.27596992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 26213 + }, + { + "epoch": 1.2760300751879698, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 26214 + }, + { + "epoch": 1.2760902255639097, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 26215 + }, + { + "epoch": 1.2761503759398496, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 26216 + }, + { + "epoch": 1.2762105263157895, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 26217 + }, + { + "epoch": 1.2762706766917293, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 26218 + }, + { + "epoch": 1.2763308270676692, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 26219 + }, + { + "epoch": 1.276390977443609, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 26220 + }, + { + "epoch": 1.276451127819549, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 26221 + }, + { + "epoch": 1.2765112781954886, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 26222 + }, + { + "epoch": 1.2765714285714287, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 26223 + }, + { + "epoch": 1.2766315789473683, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 26224 + }, + { + "epoch": 1.2766917293233082, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 26225 + }, + { + "epoch": 1.276751879699248, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 26226 + }, + { + "epoch": 1.276812030075188, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 26227 + }, + { + "epoch": 1.2768721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 26228 + }, + { + "epoch": 1.2769323308270677, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 26229 + }, + { + "epoch": 1.2769924812030076, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 26230 + }, + { + "epoch": 1.2770526315789474, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 26231 + }, + { + "epoch": 1.2771127819548873, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 26232 + }, + { + "epoch": 1.277172932330827, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 26233 + }, + { + "epoch": 1.277233082706767, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 26234 + }, + { + "epoch": 1.2772932330827067, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2318, + "step": 26235 + }, + { + "epoch": 1.2773533834586466, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9416, + "step": 26236 + }, + { + "epoch": 1.2774135338345864, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 26237 + }, + { + "epoch": 1.2774736842105263, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 26238 + }, + { + "epoch": 1.2775338345864662, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 26239 + }, + { + "epoch": 1.277593984962406, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 26240 + }, + { + "epoch": 1.277654135338346, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 26241 + }, + { + "epoch": 1.2777142857142856, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 26242 + }, + { + "epoch": 1.2777744360902257, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 26243 + }, + { + "epoch": 1.2778345864661653, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 26244 + }, + { + "epoch": 1.2778947368421052, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 26245 + }, + { + "epoch": 1.277954887218045, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 26246 + }, + { + "epoch": 1.278015037593985, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1626, + "step": 26247 + }, + { + "epoch": 1.2780751879699248, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 26248 + }, + { + "epoch": 1.2781353383458647, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 26249 + }, + { + "epoch": 1.2781954887218046, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 26250 + }, + { + "epoch": 1.2782556390977444, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 26251 + }, + { + "epoch": 1.2783157894736843, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 26252 + }, + { + "epoch": 1.278375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 26253 + }, + { + "epoch": 1.278436090225564, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 26254 + }, + { + "epoch": 1.2784962406015037, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 26255 + }, + { + "epoch": 1.2785563909774436, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 26256 + }, + { + "epoch": 1.2786165413533834, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 26257 + }, + { + "epoch": 1.2786766917293233, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 26258 + }, + { + "epoch": 1.2787368421052632, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 26259 + }, + { + "epoch": 1.278796992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 26260 + }, + { + "epoch": 1.278857142857143, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 26261 + }, + { + "epoch": 1.2789172932330828, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 26262 + }, + { + "epoch": 1.2789774436090227, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 26263 + }, + { + "epoch": 1.2790375939849623, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 26264 + }, + { + "epoch": 1.2790977443609022, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 26265 + }, + { + "epoch": 1.279157894736842, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 26266 + }, + { + "epoch": 1.279218045112782, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 26267 + }, + { + "epoch": 1.2792781954887218, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 26268 + }, + { + "epoch": 1.2793383458646617, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 26269 + }, + { + "epoch": 1.2793984962406015, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 26270 + }, + { + "epoch": 1.2794586466165414, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 26271 + }, + { + "epoch": 1.2795187969924813, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 26272 + }, + { + "epoch": 1.279578947368421, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 26273 + }, + { + "epoch": 1.279639097744361, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 26274 + }, + { + "epoch": 1.2796992481203007, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 26275 + }, + { + "epoch": 1.2797593984962405, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 26276 + }, + { + "epoch": 1.2798195488721804, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 26277 + }, + { + "epoch": 1.2798796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 26278 + }, + { + "epoch": 1.2799398496240602, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1973, + "step": 26279 + }, + { + "epoch": 1.28, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 26280 + }, + { + "epoch": 1.28006015037594, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 26281 + }, + { + "epoch": 1.2801203007518798, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 26282 + }, + { + "epoch": 1.2801804511278196, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 26283 + }, + { + "epoch": 1.2802406015037593, + "grad_norm": 1.421875, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 26284 + }, + { + "epoch": 1.2803007518796994, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 26285 + }, + { + "epoch": 1.280360902255639, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 26286 + }, + { + "epoch": 1.280421052631579, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 26287 + }, + { + "epoch": 1.2804812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 26288 + }, + { + "epoch": 1.2805413533834586, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 26289 + }, + { + "epoch": 1.2806015037593985, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 26290 + }, + { + "epoch": 1.2806616541353384, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 26291 + }, + { + "epoch": 1.2807218045112783, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 26292 + }, + { + "epoch": 1.280781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 26293 + }, + { + "epoch": 1.280842105263158, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 26294 + }, + { + "epoch": 1.2809022556390977, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 26295 + }, + { + "epoch": 1.2809624060150375, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 26296 + }, + { + "epoch": 1.2810225563909774, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 26297 + }, + { + "epoch": 1.2810827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 26298 + }, + { + "epoch": 1.2811428571428571, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 26299 + }, + { + "epoch": 1.281203007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 26300 + }, + { + "epoch": 1.2812631578947369, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 26301 + }, + { + "epoch": 1.2813233082706768, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2115, + "step": 26302 + }, + { + "epoch": 1.2813834586466166, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 26303 + }, + { + "epoch": 1.2814436090225563, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 26304 + }, + { + "epoch": 1.2815037593984964, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 26305 + }, + { + "epoch": 1.281563909774436, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 26306 + }, + { + "epoch": 1.2816240601503759, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 26307 + }, + { + "epoch": 1.2816842105263158, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 26308 + }, + { + "epoch": 1.2817443609022556, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 26309 + }, + { + "epoch": 1.2818045112781955, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 26310 + }, + { + "epoch": 1.2818646616541354, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 26311 + }, + { + "epoch": 1.2819248120300752, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 26312 + }, + { + "epoch": 1.2819849624060151, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 26313 + }, + { + "epoch": 1.282045112781955, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 26314 + }, + { + "epoch": 1.2821052631578946, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 26315 + }, + { + "epoch": 1.2821654135338345, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 26316 + }, + { + "epoch": 1.2822255639097744, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 26317 + }, + { + "epoch": 1.2822857142857143, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 26318 + }, + { + "epoch": 1.2823458646616541, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 26319 + }, + { + "epoch": 1.282406015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 26320 + }, + { + "epoch": 1.2824661654135339, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 26321 + }, + { + "epoch": 1.2825263157894737, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 26322 + }, + { + "epoch": 1.2825864661654136, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 26323 + }, + { + "epoch": 1.2826466165413533, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 26324 + }, + { + "epoch": 1.2827067669172934, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 26325 + }, + { + "epoch": 1.282766917293233, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 26326 + }, + { + "epoch": 1.2828270676691729, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 26327 + }, + { + "epoch": 1.2828872180451127, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 26328 + }, + { + "epoch": 1.2829473684210526, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 26329 + }, + { + "epoch": 1.2830075187969925, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 26330 + }, + { + "epoch": 1.2830676691729324, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 26331 + }, + { + "epoch": 1.2831278195488722, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 26332 + }, + { + "epoch": 1.283187969924812, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 26333 + }, + { + "epoch": 1.283248120300752, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 26334 + }, + { + "epoch": 1.2833082706766916, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 26335 + }, + { + "epoch": 1.2833684210526317, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2529, + "step": 26336 + }, + { + "epoch": 1.2834285714285714, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 26337 + }, + { + "epoch": 1.2834887218045112, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2736, + "step": 26338 + }, + { + "epoch": 1.283548872180451, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 26339 + }, + { + "epoch": 1.283609022556391, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 26340 + }, + { + "epoch": 1.2836691729323308, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 26341 + }, + { + "epoch": 1.2837293233082707, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 26342 + }, + { + "epoch": 1.2837894736842106, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 26343 + }, + { + "epoch": 1.2838496240601502, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 26344 + }, + { + "epoch": 1.2839097744360903, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 26345 + }, + { + "epoch": 1.28396992481203, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 26346 + }, + { + "epoch": 1.2840300751879699, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 26347 + }, + { + "epoch": 1.2840902255639097, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 26348 + }, + { + "epoch": 1.2841503759398496, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 26349 + }, + { + "epoch": 1.2842105263157895, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 26350 + }, + { + "epoch": 1.2842706766917293, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 26351 + }, + { + "epoch": 1.2843308270676692, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 26352 + }, + { + "epoch": 1.284390977443609, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 26353 + }, + { + "epoch": 1.284451127819549, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.148, + "step": 26354 + }, + { + "epoch": 1.2845112781954886, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 26355 + }, + { + "epoch": 1.2845714285714287, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 26356 + }, + { + "epoch": 1.2846315789473683, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 26357 + }, + { + "epoch": 1.2846917293233082, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 26358 + }, + { + "epoch": 1.284751879699248, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 26359 + }, + { + "epoch": 1.284812030075188, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 26360 + }, + { + "epoch": 1.2848721804511278, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 26361 + }, + { + "epoch": 1.2849323308270677, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 26362 + }, + { + "epoch": 1.2849924812030076, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 26363 + }, + { + "epoch": 1.2850526315789474, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 26364 + }, + { + "epoch": 1.2851127819548873, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 26365 + }, + { + "epoch": 1.285172932330827, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 26366 + }, + { + "epoch": 1.2852330827067668, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 26367 + }, + { + "epoch": 1.2852932330827067, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 26368 + }, + { + "epoch": 1.2853533834586466, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 26369 + }, + { + "epoch": 1.2854135338345865, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 26370 + }, + { + "epoch": 1.2854736842105263, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 26371 + }, + { + "epoch": 1.2855338345864662, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 26372 + }, + { + "epoch": 1.285593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 26373 + }, + { + "epoch": 1.285654135338346, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 26374 + }, + { + "epoch": 1.2857142857142856, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 26375 + }, + { + "epoch": 1.2857744360902257, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 26376 + }, + { + "epoch": 1.2858345864661653, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 26377 + }, + { + "epoch": 1.2858947368421052, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 26378 + }, + { + "epoch": 1.285954887218045, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 26379 + }, + { + "epoch": 1.286015037593985, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 26380 + }, + { + "epoch": 1.2860751879699248, + "grad_norm": 1.5234375, + "learning_rate": 3e-05, + "loss": 1.1632, + "step": 26381 + }, + { + "epoch": 1.2861353383458647, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 26382 + }, + { + "epoch": 1.2861954887218046, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 26383 + }, + { + "epoch": 1.2862556390977444, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 26384 + }, + { + "epoch": 1.2863157894736843, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 26385 + }, + { + "epoch": 1.286375939849624, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 26386 + }, + { + "epoch": 1.286436090225564, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 26387 + }, + { + "epoch": 1.2864962406015037, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 26388 + }, + { + "epoch": 1.2865563909774436, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 26389 + }, + { + "epoch": 1.2866165413533834, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 26390 + }, + { + "epoch": 1.2866766917293233, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 26391 + }, + { + "epoch": 1.2867368421052632, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 26392 + }, + { + "epoch": 1.286796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 26393 + }, + { + "epoch": 1.286857142857143, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 26394 + }, + { + "epoch": 1.2869172932330826, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 26395 + }, + { + "epoch": 1.2869774436090227, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 26396 + }, + { + "epoch": 1.2870375939849623, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.1623, + "step": 26397 + }, + { + "epoch": 1.2870977443609022, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 26398 + }, + { + "epoch": 1.287157894736842, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 26399 + }, + { + "epoch": 1.287218045112782, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 26400 + }, + { + "epoch": 1.2872781954887218, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 26401 + }, + { + "epoch": 1.2873383458646617, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 26402 + }, + { + "epoch": 1.2873984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 26403 + }, + { + "epoch": 1.2874586466165414, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 26404 + }, + { + "epoch": 1.2875187969924813, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 26405 + }, + { + "epoch": 1.287578947368421, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 26406 + }, + { + "epoch": 1.287639097744361, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 26407 + }, + { + "epoch": 1.2876992481203007, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 26408 + }, + { + "epoch": 1.2877593984962405, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 26409 + }, + { + "epoch": 1.2878195488721804, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 26410 + }, + { + "epoch": 1.2878796992481203, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 26411 + }, + { + "epoch": 1.2879398496240602, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 26412 + }, + { + "epoch": 1.288, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 26413 + }, + { + "epoch": 1.28806015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 26414 + }, + { + "epoch": 1.2881203007518798, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 26415 + }, + { + "epoch": 1.2881804511278196, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 26416 + }, + { + "epoch": 1.2882406015037593, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 26417 + }, + { + "epoch": 1.2883007518796992, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 26418 + }, + { + "epoch": 1.288360902255639, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0994, + "step": 26419 + }, + { + "epoch": 1.288421052631579, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 26420 + }, + { + "epoch": 1.2884812030075188, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 26421 + }, + { + "epoch": 1.2885413533834587, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 26422 + }, + { + "epoch": 1.2886015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 26423 + }, + { + "epoch": 1.2886616541353384, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2152, + "step": 26424 + }, + { + "epoch": 1.2887218045112783, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 26425 + }, + { + "epoch": 1.288781954887218, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 26426 + }, + { + "epoch": 1.288842105263158, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 26427 + }, + { + "epoch": 1.2889022556390977, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 26428 + }, + { + "epoch": 1.2889624060150375, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9258, + "step": 26429 + }, + { + "epoch": 1.2890225563909774, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 26430 + }, + { + "epoch": 1.2890827067669173, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 26431 + }, + { + "epoch": 1.2891428571428571, + "grad_norm": 1.5234375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 26432 + }, + { + "epoch": 1.289203007518797, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 26433 + }, + { + "epoch": 1.2892631578947369, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 26434 + }, + { + "epoch": 1.2893233082706768, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 26435 + }, + { + "epoch": 1.2893834586466166, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 26436 + }, + { + "epoch": 1.2894436090225563, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 26437 + }, + { + "epoch": 1.2895037593984964, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 26438 + }, + { + "epoch": 1.289563909774436, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 26439 + }, + { + "epoch": 1.289624060150376, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 26440 + }, + { + "epoch": 1.2896842105263158, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 26441 + }, + { + "epoch": 1.2897443609022556, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 26442 + }, + { + "epoch": 1.2898045112781955, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 26443 + }, + { + "epoch": 1.2898646616541354, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 26444 + }, + { + "epoch": 1.2899248120300753, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 26445 + }, + { + "epoch": 1.289984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 26446 + }, + { + "epoch": 1.290045112781955, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 26447 + }, + { + "epoch": 1.2901052631578946, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 26448 + }, + { + "epoch": 1.2901654135338345, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 26449 + }, + { + "epoch": 1.2902255639097744, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 26450 + }, + { + "epoch": 1.2902857142857143, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 26451 + }, + { + "epoch": 1.2903458646616541, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 26452 + }, + { + "epoch": 1.290406015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 26453 + }, + { + "epoch": 1.2904661654135339, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 26454 + }, + { + "epoch": 1.2905263157894737, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 26455 + }, + { + "epoch": 1.2905864661654136, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2273, + "step": 26456 + }, + { + "epoch": 1.2906466165413533, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 26457 + }, + { + "epoch": 1.2907067669172934, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 26458 + }, + { + "epoch": 1.290766917293233, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 26459 + }, + { + "epoch": 1.2908270676691729, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 26460 + }, + { + "epoch": 1.2908872180451127, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 26461 + }, + { + "epoch": 1.2909473684210526, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 26462 + }, + { + "epoch": 1.2910075187969925, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 26463 + }, + { + "epoch": 1.2910676691729324, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 26464 + }, + { + "epoch": 1.2911278195488722, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 26465 + }, + { + "epoch": 1.291187969924812, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 26466 + }, + { + "epoch": 1.291248120300752, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 26467 + }, + { + "epoch": 1.2913082706766916, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 26468 + }, + { + "epoch": 1.2913684210526315, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1688, + "step": 26469 + }, + { + "epoch": 1.2914285714285714, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 26470 + }, + { + "epoch": 1.2914887218045112, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 26471 + }, + { + "epoch": 1.2915488721804511, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 26472 + }, + { + "epoch": 1.291609022556391, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 26473 + }, + { + "epoch": 1.2916691729323309, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 26474 + }, + { + "epoch": 1.2917293233082707, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 26475 + }, + { + "epoch": 1.2917894736842106, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 26476 + }, + { + "epoch": 1.2918496240601502, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 26477 + }, + { + "epoch": 1.2919097744360903, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 26478 + }, + { + "epoch": 1.29196992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 26479 + }, + { + "epoch": 1.2920300751879699, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 26480 + }, + { + "epoch": 1.2920902255639097, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 26481 + }, + { + "epoch": 1.2921503759398496, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 26482 + }, + { + "epoch": 1.2922105263157895, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 26483 + }, + { + "epoch": 1.2922706766917293, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 26484 + }, + { + "epoch": 1.2923308270676692, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 26485 + }, + { + "epoch": 1.292390977443609, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 26486 + }, + { + "epoch": 1.292451127819549, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 26487 + }, + { + "epoch": 1.2925112781954886, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 26488 + }, + { + "epoch": 1.2925714285714287, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 26489 + }, + { + "epoch": 1.2926315789473684, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 26490 + }, + { + "epoch": 1.2926917293233082, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 26491 + }, + { + "epoch": 1.292751879699248, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 26492 + }, + { + "epoch": 1.292812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 26493 + }, + { + "epoch": 1.2928721804511278, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 26494 + }, + { + "epoch": 1.2929323308270677, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 26495 + }, + { + "epoch": 1.2929924812030076, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 26496 + }, + { + "epoch": 1.2930526315789472, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 26497 + }, + { + "epoch": 1.2931127819548873, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 26498 + }, + { + "epoch": 1.293172932330827, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2074, + "step": 26499 + }, + { + "epoch": 1.2932330827067668, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 26500 + }, + { + "epoch": 1.2932932330827067, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 26501 + }, + { + "epoch": 1.2933533834586466, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 26502 + }, + { + "epoch": 1.2934135338345865, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 26503 + }, + { + "epoch": 1.2934736842105263, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 26504 + }, + { + "epoch": 1.2935338345864662, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 26505 + }, + { + "epoch": 1.293593984962406, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 26506 + }, + { + "epoch": 1.293654135338346, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 26507 + }, + { + "epoch": 1.2937142857142856, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 26508 + }, + { + "epoch": 1.2937744360902257, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 26509 + }, + { + "epoch": 1.2938345864661653, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 26510 + }, + { + "epoch": 1.2938947368421052, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 26511 + }, + { + "epoch": 1.293954887218045, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 26512 + }, + { + "epoch": 1.294015037593985, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 26513 + }, + { + "epoch": 1.2940751879699248, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 26514 + }, + { + "epoch": 1.2941353383458647, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 26515 + }, + { + "epoch": 1.2941954887218046, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 26516 + }, + { + "epoch": 1.2942556390977444, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 26517 + }, + { + "epoch": 1.2943157894736843, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 26518 + }, + { + "epoch": 1.294375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 26519 + }, + { + "epoch": 1.2944360902255638, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 26520 + }, + { + "epoch": 1.2944962406015037, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 26521 + }, + { + "epoch": 1.2945563909774436, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 26522 + }, + { + "epoch": 1.2946165413533834, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.3064, + "step": 26523 + }, + { + "epoch": 1.2946766917293233, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 26524 + }, + { + "epoch": 1.2947368421052632, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 26525 + }, + { + "epoch": 1.294796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 26526 + }, + { + "epoch": 1.294857142857143, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 26527 + }, + { + "epoch": 1.2949172932330826, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 26528 + }, + { + "epoch": 1.2949774436090227, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 26529 + }, + { + "epoch": 1.2950375939849623, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 26530 + }, + { + "epoch": 1.2950977443609022, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 26531 + }, + { + "epoch": 1.295157894736842, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 26532 + }, + { + "epoch": 1.295218045112782, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 26533 + }, + { + "epoch": 1.2952781954887218, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 26534 + }, + { + "epoch": 1.2953383458646617, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 26535 + }, + { + "epoch": 1.2953984962406015, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 26536 + }, + { + "epoch": 1.2954586466165414, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 26537 + }, + { + "epoch": 1.2955187969924813, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 26538 + }, + { + "epoch": 1.295578947368421, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 26539 + }, + { + "epoch": 1.295639097744361, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 26540 + }, + { + "epoch": 1.2956992481203007, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 26541 + }, + { + "epoch": 1.2957593984962406, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 26542 + }, + { + "epoch": 1.2958195488721804, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 26543 + }, + { + "epoch": 1.2958796992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 26544 + }, + { + "epoch": 1.2959398496240602, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 26545 + }, + { + "epoch": 1.296, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 26546 + }, + { + "epoch": 1.29606015037594, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 26547 + }, + { + "epoch": 1.2961203007518796, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 26548 + }, + { + "epoch": 1.2961804511278197, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2031, + "step": 26549 + }, + { + "epoch": 1.2962406015037593, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 26550 + }, + { + "epoch": 1.2963007518796992, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 26551 + }, + { + "epoch": 1.296360902255639, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 26552 + }, + { + "epoch": 1.296421052631579, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 26553 + }, + { + "epoch": 1.2964812030075188, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 26554 + }, + { + "epoch": 1.2965413533834587, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 26555 + }, + { + "epoch": 1.2966015037593985, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 26556 + }, + { + "epoch": 1.2966616541353384, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 26557 + }, + { + "epoch": 1.2967218045112783, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 26558 + }, + { + "epoch": 1.296781954887218, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 26559 + }, + { + "epoch": 1.296842105263158, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 26560 + }, + { + "epoch": 1.2969022556390977, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 26561 + }, + { + "epoch": 1.2969624060150375, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 26562 + }, + { + "epoch": 1.2970225563909774, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1677, + "step": 26563 + }, + { + "epoch": 1.2970827067669173, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 26564 + }, + { + "epoch": 1.2971428571428572, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 26565 + }, + { + "epoch": 1.297203007518797, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 26566 + }, + { + "epoch": 1.297263157894737, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9145, + "step": 26567 + }, + { + "epoch": 1.2973233082706768, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 26568 + }, + { + "epoch": 1.2973834586466166, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 26569 + }, + { + "epoch": 1.2974436090225563, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.147, + "step": 26570 + }, + { + "epoch": 1.2975037593984962, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 26571 + }, + { + "epoch": 1.297563909774436, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 26572 + }, + { + "epoch": 1.297624060150376, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 26573 + }, + { + "epoch": 1.2976842105263158, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 26574 + }, + { + "epoch": 1.2977443609022556, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 26575 + }, + { + "epoch": 1.2978045112781955, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 26576 + }, + { + "epoch": 1.2978646616541354, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 26577 + }, + { + "epoch": 1.2979248120300753, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 26578 + }, + { + "epoch": 1.297984962406015, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 26579 + }, + { + "epoch": 1.298045112781955, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 26580 + }, + { + "epoch": 1.2981052631578947, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 26581 + }, + { + "epoch": 1.2981654135338345, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 26582 + }, + { + "epoch": 1.2982255639097744, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 26583 + }, + { + "epoch": 1.2982857142857143, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 26584 + }, + { + "epoch": 1.2983458646616541, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 26585 + }, + { + "epoch": 1.298406015037594, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 26586 + }, + { + "epoch": 1.2984661654135339, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 26587 + }, + { + "epoch": 1.2985263157894738, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 26588 + }, + { + "epoch": 1.2985864661654136, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 26589 + }, + { + "epoch": 1.2986466165413533, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 26590 + }, + { + "epoch": 1.2987067669172934, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 26591 + }, + { + "epoch": 1.298766917293233, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 26592 + }, + { + "epoch": 1.2988270676691729, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 26593 + }, + { + "epoch": 1.2988872180451128, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 26594 + }, + { + "epoch": 1.2989473684210526, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 26595 + }, + { + "epoch": 1.2990075187969925, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 26596 + }, + { + "epoch": 1.2990676691729324, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 26597 + }, + { + "epoch": 1.2991278195488722, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 26598 + }, + { + "epoch": 1.299187969924812, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 26599 + }, + { + "epoch": 1.299248120300752, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 26600 + }, + { + "epoch": 1.2993082706766916, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 26601 + }, + { + "epoch": 1.2993684210526315, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 26602 + }, + { + "epoch": 1.2994285714285714, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 26603 + }, + { + "epoch": 1.2994887218045112, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 26604 + }, + { + "epoch": 1.2995488721804511, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 26605 + }, + { + "epoch": 1.299609022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 26606 + }, + { + "epoch": 1.2996691729323309, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 26607 + }, + { + "epoch": 1.2997293233082707, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 26608 + }, + { + "epoch": 1.2997894736842106, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 26609 + }, + { + "epoch": 1.2998496240601503, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 26610 + }, + { + "epoch": 1.2999097744360903, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 26611 + }, + { + "epoch": 1.29996992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 26612 + }, + { + "epoch": 1.3000300751879699, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 26613 + }, + { + "epoch": 1.3000902255639097, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 26614 + }, + { + "epoch": 1.3001503759398496, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 26615 + }, + { + "epoch": 1.3002105263157895, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.172, + "step": 26616 + }, + { + "epoch": 1.3002706766917294, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 26617 + }, + { + "epoch": 1.3003308270676692, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 26618 + }, + { + "epoch": 1.300390977443609, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 26619 + }, + { + "epoch": 1.300451127819549, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 26620 + }, + { + "epoch": 1.3005112781954886, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 26621 + }, + { + "epoch": 1.3005714285714285, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 26622 + }, + { + "epoch": 1.3006315789473684, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 26623 + }, + { + "epoch": 1.3006917293233082, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 26624 + }, + { + "epoch": 1.300751879699248, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 26625 + }, + { + "epoch": 1.300812030075188, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 26626 + }, + { + "epoch": 1.3008721804511278, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 26627 + }, + { + "epoch": 1.3009323308270677, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 26628 + }, + { + "epoch": 1.3009924812030076, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 26629 + }, + { + "epoch": 1.3010526315789472, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 26630 + }, + { + "epoch": 1.3011127819548873, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 26631 + }, + { + "epoch": 1.301172932330827, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 26632 + }, + { + "epoch": 1.3012330827067669, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 26633 + }, + { + "epoch": 1.3012932330827067, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 26634 + }, + { + "epoch": 1.3013533834586466, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 26635 + }, + { + "epoch": 1.3014135338345865, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 26636 + }, + { + "epoch": 1.3014736842105263, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 26637 + }, + { + "epoch": 1.3015338345864662, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 26638 + }, + { + "epoch": 1.301593984962406, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.136, + "step": 26639 + }, + { + "epoch": 1.301654135338346, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 26640 + }, + { + "epoch": 1.3017142857142856, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 26641 + }, + { + "epoch": 1.3017744360902257, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 26642 + }, + { + "epoch": 1.3018345864661653, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 26643 + }, + { + "epoch": 1.3018947368421052, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 26644 + }, + { + "epoch": 1.301954887218045, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 26645 + }, + { + "epoch": 1.302015037593985, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 26646 + }, + { + "epoch": 1.3020751879699248, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 26647 + }, + { + "epoch": 1.3021353383458647, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 26648 + }, + { + "epoch": 1.3021954887218046, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1612, + "step": 26649 + }, + { + "epoch": 1.3022556390977444, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 26650 + }, + { + "epoch": 1.3023157894736843, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 26651 + }, + { + "epoch": 1.302375939849624, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 26652 + }, + { + "epoch": 1.3024360902255638, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9501, + "step": 26653 + }, + { + "epoch": 1.3024962406015037, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 26654 + }, + { + "epoch": 1.3025563909774436, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 26655 + }, + { + "epoch": 1.3026165413533835, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 26656 + }, + { + "epoch": 1.3026766917293233, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 26657 + }, + { + "epoch": 1.3027368421052632, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 26658 + }, + { + "epoch": 1.302796992481203, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 26659 + }, + { + "epoch": 1.302857142857143, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 26660 + }, + { + "epoch": 1.3029172932330826, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 26661 + }, + { + "epoch": 1.3029774436090227, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 26662 + }, + { + "epoch": 1.3030375939849623, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 26663 + }, + { + "epoch": 1.3030977443609022, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 26664 + }, + { + "epoch": 1.303157894736842, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 26665 + }, + { + "epoch": 1.303218045112782, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 26666 + }, + { + "epoch": 1.3032781954887218, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2376, + "step": 26667 + }, + { + "epoch": 1.3033383458646617, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 26668 + }, + { + "epoch": 1.3033984962406016, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 26669 + }, + { + "epoch": 1.3034586466165414, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 26670 + }, + { + "epoch": 1.3035187969924813, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 26671 + }, + { + "epoch": 1.303578947368421, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 26672 + }, + { + "epoch": 1.3036390977443608, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 26673 + }, + { + "epoch": 1.3036992481203007, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 26674 + }, + { + "epoch": 1.3037593984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.25, + "step": 26675 + }, + { + "epoch": 1.3038195488721804, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 26676 + }, + { + "epoch": 1.3038796992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 26677 + }, + { + "epoch": 1.3039398496240602, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 26678 + }, + { + "epoch": 1.304, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 26679 + }, + { + "epoch": 1.30406015037594, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 26680 + }, + { + "epoch": 1.3041203007518796, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 26681 + }, + { + "epoch": 1.3041804511278197, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 26682 + }, + { + "epoch": 1.3042406015037593, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 26683 + }, + { + "epoch": 1.3043007518796992, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 26684 + }, + { + "epoch": 1.304360902255639, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 26685 + }, + { + "epoch": 1.304421052631579, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 26686 + }, + { + "epoch": 1.3044812030075188, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 26687 + }, + { + "epoch": 1.3045413533834587, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 26688 + }, + { + "epoch": 1.3046015037593985, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1904, + "step": 26689 + }, + { + "epoch": 1.3046616541353384, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 26690 + }, + { + "epoch": 1.3047218045112783, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 26691 + }, + { + "epoch": 1.304781954887218, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 26692 + }, + { + "epoch": 1.304842105263158, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 26693 + }, + { + "epoch": 1.3049022556390977, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 26694 + }, + { + "epoch": 1.3049624060150375, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 26695 + }, + { + "epoch": 1.3050225563909774, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 26696 + }, + { + "epoch": 1.3050827067669173, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 26697 + }, + { + "epoch": 1.3051428571428572, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 26698 + }, + { + "epoch": 1.305203007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 26699 + }, + { + "epoch": 1.305263157894737, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 26700 + }, + { + "epoch": 1.3053233082706768, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 26701 + }, + { + "epoch": 1.3053834586466166, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 26702 + }, + { + "epoch": 1.3054436090225563, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 26703 + }, + { + "epoch": 1.3055037593984962, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 26704 + }, + { + "epoch": 1.305563909774436, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 26705 + }, + { + "epoch": 1.305624060150376, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 26706 + }, + { + "epoch": 1.3056842105263158, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 26707 + }, + { + "epoch": 1.3057443609022557, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2051, + "step": 26708 + }, + { + "epoch": 1.3058045112781955, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 26709 + }, + { + "epoch": 1.3058646616541354, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 26710 + }, + { + "epoch": 1.3059248120300753, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 26711 + }, + { + "epoch": 1.305984962406015, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 26712 + }, + { + "epoch": 1.306045112781955, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 26713 + }, + { + "epoch": 1.3061052631578947, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 26714 + }, + { + "epoch": 1.3061654135338345, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 26715 + }, + { + "epoch": 1.3062255639097744, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 26716 + }, + { + "epoch": 1.3062857142857143, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 26717 + }, + { + "epoch": 1.3063458646616541, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 26718 + }, + { + "epoch": 1.306406015037594, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 26719 + }, + { + "epoch": 1.3064661654135339, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 26720 + }, + { + "epoch": 1.3065263157894738, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 26721 + }, + { + "epoch": 1.3065864661654136, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 26722 + }, + { + "epoch": 1.3066466165413533, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 26723 + }, + { + "epoch": 1.3067067669172931, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 26724 + }, + { + "epoch": 1.306766917293233, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 26725 + }, + { + "epoch": 1.306827067669173, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 26726 + }, + { + "epoch": 1.3068872180451128, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 26727 + }, + { + "epoch": 1.3069473684210526, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 26728 + }, + { + "epoch": 1.3070075187969925, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.372, + "step": 26729 + }, + { + "epoch": 1.3070676691729324, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.4232, + "step": 26730 + }, + { + "epoch": 1.3071278195488722, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 26731 + }, + { + "epoch": 1.307187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 26732 + }, + { + "epoch": 1.307248120300752, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 26733 + }, + { + "epoch": 1.3073082706766916, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 26734 + }, + { + "epoch": 1.3073684210526315, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 26735 + }, + { + "epoch": 1.3074285714285714, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 26736 + }, + { + "epoch": 1.3074887218045113, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 26737 + }, + { + "epoch": 1.3075488721804511, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 26738 + }, + { + "epoch": 1.307609022556391, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 26739 + }, + { + "epoch": 1.3076691729323309, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 26740 + }, + { + "epoch": 1.3077293233082707, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 26741 + }, + { + "epoch": 1.3077894736842106, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 26742 + }, + { + "epoch": 1.3078496240601503, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 26743 + }, + { + "epoch": 1.3079097744360904, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 26744 + }, + { + "epoch": 1.30796992481203, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 26745 + }, + { + "epoch": 1.3080300751879699, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 26746 + }, + { + "epoch": 1.3080902255639097, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 26747 + }, + { + "epoch": 1.3081503759398496, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 26748 + }, + { + "epoch": 1.3082105263157895, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 26749 + }, + { + "epoch": 1.3082706766917294, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 26750 + }, + { + "epoch": 1.3083308270676692, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 26751 + }, + { + "epoch": 1.308390977443609, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 26752 + }, + { + "epoch": 1.308451127819549, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 26753 + }, + { + "epoch": 1.3085112781954886, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 26754 + }, + { + "epoch": 1.3085714285714285, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 26755 + }, + { + "epoch": 1.3086315789473684, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 26756 + }, + { + "epoch": 1.3086917293233082, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 26757 + }, + { + "epoch": 1.308751879699248, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 26758 + }, + { + "epoch": 1.308812030075188, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 26759 + }, + { + "epoch": 1.3088721804511279, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 26760 + }, + { + "epoch": 1.3089323308270677, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 26761 + }, + { + "epoch": 1.3089924812030076, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 26762 + }, + { + "epoch": 1.3090526315789472, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 26763 + }, + { + "epoch": 1.3091127819548873, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 26764 + }, + { + "epoch": 1.309172932330827, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 26765 + }, + { + "epoch": 1.3092330827067669, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2588, + "step": 26766 + }, + { + "epoch": 1.3092932330827067, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 26767 + }, + { + "epoch": 1.3093533834586466, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 26768 + }, + { + "epoch": 1.3094135338345865, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 26769 + }, + { + "epoch": 1.3094736842105263, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1835, + "step": 26770 + }, + { + "epoch": 1.3095338345864662, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 26771 + }, + { + "epoch": 1.309593984962406, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 26772 + }, + { + "epoch": 1.309654135338346, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 26773 + }, + { + "epoch": 1.3097142857142856, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 26774 + }, + { + "epoch": 1.3097744360902257, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 26775 + }, + { + "epoch": 1.3098345864661654, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 26776 + }, + { + "epoch": 1.3098947368421052, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 26777 + }, + { + "epoch": 1.309954887218045, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 26778 + }, + { + "epoch": 1.310015037593985, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 26779 + }, + { + "epoch": 1.3100751879699248, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 26780 + }, + { + "epoch": 1.3101353383458647, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.2033, + "step": 26781 + }, + { + "epoch": 1.3101954887218046, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 26782 + }, + { + "epoch": 1.3102556390977442, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 26783 + }, + { + "epoch": 1.3103157894736843, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 26784 + }, + { + "epoch": 1.310375939849624, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 26785 + }, + { + "epoch": 1.3104360902255638, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 26786 + }, + { + "epoch": 1.3104962406015037, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 26787 + }, + { + "epoch": 1.3105563909774436, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 26788 + }, + { + "epoch": 1.3106165413533835, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 26789 + }, + { + "epoch": 1.3106766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 26790 + }, + { + "epoch": 1.3107368421052632, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 26791 + }, + { + "epoch": 1.310796992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1763, + "step": 26792 + }, + { + "epoch": 1.310857142857143, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 26793 + }, + { + "epoch": 1.3109172932330826, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9254, + "step": 26794 + }, + { + "epoch": 1.3109774436090227, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 26795 + }, + { + "epoch": 1.3110375939849623, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 26796 + }, + { + "epoch": 1.3110977443609022, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 26797 + }, + { + "epoch": 1.311157894736842, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 26798 + }, + { + "epoch": 1.311218045112782, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 26799 + }, + { + "epoch": 1.3112781954887218, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 26800 + }, + { + "epoch": 1.3113383458646617, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 26801 + }, + { + "epoch": 1.3113984962406016, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 26802 + }, + { + "epoch": 1.3114586466165414, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 26803 + }, + { + "epoch": 1.3115187969924813, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 26804 + }, + { + "epoch": 1.311578947368421, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 26805 + }, + { + "epoch": 1.3116390977443608, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 26806 + }, + { + "epoch": 1.3116992481203007, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 26807 + }, + { + "epoch": 1.3117593984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 26808 + }, + { + "epoch": 1.3118195488721804, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 26809 + }, + { + "epoch": 1.3118796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 26810 + }, + { + "epoch": 1.3119398496240602, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 26811 + }, + { + "epoch": 1.312, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 26812 + }, + { + "epoch": 1.31206015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 26813 + }, + { + "epoch": 1.3121203007518796, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 26814 + }, + { + "epoch": 1.3121804511278197, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 26815 + }, + { + "epoch": 1.3122406015037593, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 26816 + }, + { + "epoch": 1.3123007518796992, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 26817 + }, + { + "epoch": 1.312360902255639, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 26818 + }, + { + "epoch": 1.312421052631579, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 26819 + }, + { + "epoch": 1.3124812030075188, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 26820 + }, + { + "epoch": 1.3125413533834587, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 26821 + }, + { + "epoch": 1.3126015037593985, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 26822 + }, + { + "epoch": 1.3126616541353384, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 26823 + }, + { + "epoch": 1.3127218045112783, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 26824 + }, + { + "epoch": 1.312781954887218, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 26825 + }, + { + "epoch": 1.312842105263158, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 26826 + }, + { + "epoch": 1.3129022556390977, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 26827 + }, + { + "epoch": 1.3129624060150376, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 26828 + }, + { + "epoch": 1.3130225563909774, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 26829 + }, + { + "epoch": 1.3130827067669173, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 26830 + }, + { + "epoch": 1.3131428571428572, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 26831 + }, + { + "epoch": 1.313203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 26832 + }, + { + "epoch": 1.313263157894737, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 26833 + }, + { + "epoch": 1.3133233082706766, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 26834 + }, + { + "epoch": 1.3133834586466167, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.188, + "step": 26835 + }, + { + "epoch": 1.3134436090225563, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 26836 + }, + { + "epoch": 1.3135037593984962, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 26837 + }, + { + "epoch": 1.313563909774436, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 26838 + }, + { + "epoch": 1.313624060150376, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.2315, + "step": 26839 + }, + { + "epoch": 1.3136842105263158, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 26840 + }, + { + "epoch": 1.3137443609022557, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 26841 + }, + { + "epoch": 1.3138045112781955, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 26842 + }, + { + "epoch": 1.3138646616541354, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 26843 + }, + { + "epoch": 1.3139248120300753, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 26844 + }, + { + "epoch": 1.313984962406015, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 26845 + }, + { + "epoch": 1.314045112781955, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 26846 + }, + { + "epoch": 1.3141052631578947, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 26847 + }, + { + "epoch": 1.3141654135338345, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1526, + "step": 26848 + }, + { + "epoch": 1.3142255639097744, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 26849 + }, + { + "epoch": 1.3142857142857143, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 26850 + }, + { + "epoch": 1.3143458646616542, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 26851 + }, + { + "epoch": 1.314406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 26852 + }, + { + "epoch": 1.314466165413534, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 26853 + }, + { + "epoch": 1.3145263157894738, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 26854 + }, + { + "epoch": 1.3145864661654136, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 26855 + }, + { + "epoch": 1.3146466165413533, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 26856 + }, + { + "epoch": 1.3147067669172932, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 26857 + }, + { + "epoch": 1.314766917293233, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 26858 + }, + { + "epoch": 1.314827067669173, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 26859 + }, + { + "epoch": 1.3148872180451128, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 26860 + }, + { + "epoch": 1.3149473684210526, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 26861 + }, + { + "epoch": 1.3150075187969925, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 26862 + }, + { + "epoch": 1.3150676691729324, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 26863 + }, + { + "epoch": 1.3151278195488723, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 26864 + }, + { + "epoch": 1.315187969924812, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 26865 + }, + { + "epoch": 1.315248120300752, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 26866 + }, + { + "epoch": 1.3153082706766916, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 26867 + }, + { + "epoch": 1.3153684210526315, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 26868 + }, + { + "epoch": 1.3154285714285714, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 26869 + }, + { + "epoch": 1.3154887218045113, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 26870 + }, + { + "epoch": 1.3155488721804511, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 26871 + }, + { + "epoch": 1.315609022556391, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 26872 + }, + { + "epoch": 1.3156691729323309, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 26873 + }, + { + "epoch": 1.3157293233082707, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 26874 + }, + { + "epoch": 1.3157894736842106, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 26875 + }, + { + "epoch": 1.3158496240601503, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 26876 + }, + { + "epoch": 1.3159097744360904, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 26877 + }, + { + "epoch": 1.31596992481203, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 26878 + }, + { + "epoch": 1.3160300751879699, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 26879 + }, + { + "epoch": 1.3160902255639098, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 26880 + }, + { + "epoch": 1.3161503759398496, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 26881 + }, + { + "epoch": 1.3162105263157895, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 26882 + }, + { + "epoch": 1.3162706766917294, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 26883 + }, + { + "epoch": 1.3163308270676692, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 26884 + }, + { + "epoch": 1.316390977443609, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 26885 + }, + { + "epoch": 1.316451127819549, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 26886 + }, + { + "epoch": 1.3165112781954886, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 26887 + }, + { + "epoch": 1.3165714285714285, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 26888 + }, + { + "epoch": 1.3166315789473684, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 26889 + }, + { + "epoch": 1.3166917293233082, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 26890 + }, + { + "epoch": 1.3167518796992481, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 26891 + }, + { + "epoch": 1.316812030075188, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 26892 + }, + { + "epoch": 1.3168721804511279, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 26893 + }, + { + "epoch": 1.3169323308270677, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 26894 + }, + { + "epoch": 1.3169924812030076, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 26895 + }, + { + "epoch": 1.3170526315789473, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 26896 + }, + { + "epoch": 1.3171127819548873, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 26897 + }, + { + "epoch": 1.317172932330827, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 26898 + }, + { + "epoch": 1.3172330827067669, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 26899 + }, + { + "epoch": 1.3172932330827067, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 26900 + }, + { + "epoch": 1.3173533834586466, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 26901 + }, + { + "epoch": 1.3174135338345865, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 26902 + }, + { + "epoch": 1.3174736842105264, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 26903 + }, + { + "epoch": 1.3175338345864662, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 26904 + }, + { + "epoch": 1.317593984962406, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 26905 + }, + { + "epoch": 1.317654135338346, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 26906 + }, + { + "epoch": 1.3177142857142856, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 0.934, + "step": 26907 + }, + { + "epoch": 1.3177744360902255, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 26908 + }, + { + "epoch": 1.3178345864661654, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 26909 + }, + { + "epoch": 1.3178947368421052, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 26910 + }, + { + "epoch": 1.317954887218045, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 26911 + }, + { + "epoch": 1.318015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 26912 + }, + { + "epoch": 1.3180751879699248, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 26913 + }, + { + "epoch": 1.3181353383458647, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 26914 + }, + { + "epoch": 1.3181954887218046, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 26915 + }, + { + "epoch": 1.3182556390977442, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 26916 + }, + { + "epoch": 1.3183157894736843, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 26917 + }, + { + "epoch": 1.318375939849624, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 26918 + }, + { + "epoch": 1.3184360902255639, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 26919 + }, + { + "epoch": 1.3184962406015037, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.1903, + "step": 26920 + }, + { + "epoch": 1.3185563909774436, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 26921 + }, + { + "epoch": 1.3186165413533835, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 26922 + }, + { + "epoch": 1.3186766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 26923 + }, + { + "epoch": 1.3187368421052632, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 26924 + }, + { + "epoch": 1.318796992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 26925 + }, + { + "epoch": 1.318857142857143, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 26926 + }, + { + "epoch": 1.3189172932330826, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 26927 + }, + { + "epoch": 1.3189774436090227, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 26928 + }, + { + "epoch": 1.3190375939849623, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 26929 + }, + { + "epoch": 1.3190977443609022, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 26930 + }, + { + "epoch": 1.319157894736842, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 26931 + }, + { + "epoch": 1.319218045112782, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 26932 + }, + { + "epoch": 1.3192781954887218, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 26933 + }, + { + "epoch": 1.3193383458646617, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 26934 + }, + { + "epoch": 1.3193984962406016, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 26935 + }, + { + "epoch": 1.3194586466165412, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 26936 + }, + { + "epoch": 1.3195187969924813, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 26937 + }, + { + "epoch": 1.319578947368421, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 26938 + }, + { + "epoch": 1.3196390977443608, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 26939 + }, + { + "epoch": 1.3196992481203007, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 26940 + }, + { + "epoch": 1.3197593984962406, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 26941 + }, + { + "epoch": 1.3198195488721804, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 26942 + }, + { + "epoch": 1.3198796992481203, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 26943 + }, + { + "epoch": 1.3199398496240602, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 26944 + }, + { + "epoch": 1.32, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 26945 + }, + { + "epoch": 1.32006015037594, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 26946 + }, + { + "epoch": 1.3201203007518796, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 26947 + }, + { + "epoch": 1.3201804511278197, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 26948 + }, + { + "epoch": 1.3202406015037593, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 26949 + }, + { + "epoch": 1.3203007518796992, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 26950 + }, + { + "epoch": 1.320360902255639, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 26951 + }, + { + "epoch": 1.320421052631579, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 26952 + }, + { + "epoch": 1.3204812030075188, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 26953 + }, + { + "epoch": 1.3205413533834587, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 26954 + }, + { + "epoch": 1.3206015037593986, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 26955 + }, + { + "epoch": 1.3206616541353384, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 26956 + }, + { + "epoch": 1.3207218045112783, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 26957 + }, + { + "epoch": 1.320781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 26958 + }, + { + "epoch": 1.3208421052631578, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 26959 + }, + { + "epoch": 1.3209022556390977, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 26960 + }, + { + "epoch": 1.3209624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 26961 + }, + { + "epoch": 1.3210225563909774, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 26962 + }, + { + "epoch": 1.3210827067669173, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 26963 + }, + { + "epoch": 1.3211428571428572, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 26964 + }, + { + "epoch": 1.321203007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 26965 + }, + { + "epoch": 1.321263157894737, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 26966 + }, + { + "epoch": 1.3213233082706766, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 26967 + }, + { + "epoch": 1.3213834586466167, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 26968 + }, + { + "epoch": 1.3214436090225563, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 26969 + }, + { + "epoch": 1.3215037593984962, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 26970 + }, + { + "epoch": 1.321563909774436, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 26971 + }, + { + "epoch": 1.321624060150376, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 26972 + }, + { + "epoch": 1.3216842105263158, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 26973 + }, + { + "epoch": 1.3217443609022557, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 26974 + }, + { + "epoch": 1.3218045112781955, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 26975 + }, + { + "epoch": 1.3218646616541354, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 26976 + }, + { + "epoch": 1.3219248120300753, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 26977 + }, + { + "epoch": 1.321984962406015, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 26978 + }, + { + "epoch": 1.322045112781955, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 26979 + }, + { + "epoch": 1.3221052631578947, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 26980 + }, + { + "epoch": 1.3221654135338345, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 26981 + }, + { + "epoch": 1.3222255639097744, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 26982 + }, + { + "epoch": 1.3222857142857143, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 26983 + }, + { + "epoch": 1.3223458646616542, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 26984 + }, + { + "epoch": 1.322406015037594, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 26985 + }, + { + "epoch": 1.322466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 26986 + }, + { + "epoch": 1.3225263157894735, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 26987 + }, + { + "epoch": 1.3225864661654136, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 26988 + }, + { + "epoch": 1.3226466165413533, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 26989 + }, + { + "epoch": 1.3227067669172932, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 26990 + }, + { + "epoch": 1.322766917293233, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 26991 + }, + { + "epoch": 1.322827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2239, + "step": 26992 + }, + { + "epoch": 1.3228872180451128, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 26993 + }, + { + "epoch": 1.3229473684210526, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 26994 + }, + { + "epoch": 1.3230075187969925, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 26995 + }, + { + "epoch": 1.3230676691729324, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 26996 + }, + { + "epoch": 1.3231278195488723, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 26997 + }, + { + "epoch": 1.323187969924812, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 26998 + }, + { + "epoch": 1.323248120300752, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 26999 + }, + { + "epoch": 1.3233082706766917, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 27000 + }, + { + "epoch": 1.3233684210526315, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2382, + "step": 27001 + }, + { + "epoch": 1.3234285714285714, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 27002 + }, + { + "epoch": 1.3234887218045113, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 27003 + }, + { + "epoch": 1.3235488721804511, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 27004 + }, + { + "epoch": 1.323609022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 27005 + }, + { + "epoch": 1.3236691729323309, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 27006 + }, + { + "epoch": 1.3237293233082708, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 27007 + }, + { + "epoch": 1.3237894736842106, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 27008 + }, + { + "epoch": 1.3238496240601503, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 27009 + }, + { + "epoch": 1.3239097744360901, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 27010 + }, + { + "epoch": 1.32396992481203, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 27011 + }, + { + "epoch": 1.32403007518797, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 27012 + }, + { + "epoch": 1.3240902255639098, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 27013 + }, + { + "epoch": 1.3241503759398496, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 27014 + }, + { + "epoch": 1.3242105263157895, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 27015 + }, + { + "epoch": 1.3242706766917294, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 27016 + }, + { + "epoch": 1.3243308270676692, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 27017 + }, + { + "epoch": 1.324390977443609, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 27018 + }, + { + "epoch": 1.324451127819549, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 27019 + }, + { + "epoch": 1.3245112781954886, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 27020 + }, + { + "epoch": 1.3245714285714285, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 27021 + }, + { + "epoch": 1.3246315789473684, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 27022 + }, + { + "epoch": 1.3246917293233083, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 27023 + }, + { + "epoch": 1.3247518796992481, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 27024 + }, + { + "epoch": 1.324812030075188, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 27025 + }, + { + "epoch": 1.3248721804511279, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 27026 + }, + { + "epoch": 1.3249323308270677, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 27027 + }, + { + "epoch": 1.3249924812030076, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 27028 + }, + { + "epoch": 1.3250526315789473, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 27029 + }, + { + "epoch": 1.3251127819548874, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 27030 + }, + { + "epoch": 1.325172932330827, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 27031 + }, + { + "epoch": 1.3252330827067669, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1612, + "step": 27032 + }, + { + "epoch": 1.3252932330827067, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 27033 + }, + { + "epoch": 1.3253533834586466, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 27034 + }, + { + "epoch": 1.3254135338345865, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 27035 + }, + { + "epoch": 1.3254736842105264, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 27036 + }, + { + "epoch": 1.3255338345864662, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 27037 + }, + { + "epoch": 1.3255939849624059, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 27038 + }, + { + "epoch": 1.325654135338346, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 27039 + }, + { + "epoch": 1.3257142857142856, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 27040 + }, + { + "epoch": 1.3257744360902255, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 27041 + }, + { + "epoch": 1.3258345864661654, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1996, + "step": 27042 + }, + { + "epoch": 1.3258947368421052, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 27043 + }, + { + "epoch": 1.325954887218045, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 27044 + }, + { + "epoch": 1.326015037593985, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 27045 + }, + { + "epoch": 1.3260751879699249, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 27046 + }, + { + "epoch": 1.3261353383458647, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 27047 + }, + { + "epoch": 1.3261954887218046, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 27048 + }, + { + "epoch": 1.3262556390977442, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 27049 + }, + { + "epoch": 1.3263157894736843, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 27050 + }, + { + "epoch": 1.326375939849624, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 27051 + }, + { + "epoch": 1.3264360902255639, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 27052 + }, + { + "epoch": 1.3264962406015037, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.2144, + "step": 27053 + }, + { + "epoch": 1.3265563909774436, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 27054 + }, + { + "epoch": 1.3266165413533835, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 27055 + }, + { + "epoch": 1.3266766917293233, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 27056 + }, + { + "epoch": 1.3267368421052632, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 27057 + }, + { + "epoch": 1.326796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 27058 + }, + { + "epoch": 1.326857142857143, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 27059 + }, + { + "epoch": 1.3269172932330826, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 27060 + }, + { + "epoch": 1.3269774436090225, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1264, + "step": 27061 + }, + { + "epoch": 1.3270375939849623, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 27062 + }, + { + "epoch": 1.3270977443609022, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 27063 + }, + { + "epoch": 1.327157894736842, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 27064 + }, + { + "epoch": 1.327218045112782, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 27065 + }, + { + "epoch": 1.3272781954887218, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 27066 + }, + { + "epoch": 1.3273383458646617, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 27067 + }, + { + "epoch": 1.3273984962406016, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 27068 + }, + { + "epoch": 1.3274586466165412, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 27069 + }, + { + "epoch": 1.3275187969924813, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 27070 + }, + { + "epoch": 1.327578947368421, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 27071 + }, + { + "epoch": 1.3276390977443608, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 27072 + }, + { + "epoch": 1.3276992481203007, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 27073 + }, + { + "epoch": 1.3277593984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 27074 + }, + { + "epoch": 1.3278195488721805, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 27075 + }, + { + "epoch": 1.3278796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 27076 + }, + { + "epoch": 1.3279398496240602, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 27077 + }, + { + "epoch": 1.328, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 27078 + }, + { + "epoch": 1.32806015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 27079 + }, + { + "epoch": 1.3281203007518796, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 27080 + }, + { + "epoch": 1.3281804511278197, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 27081 + }, + { + "epoch": 1.3282406015037593, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 27082 + }, + { + "epoch": 1.3283007518796992, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 27083 + }, + { + "epoch": 1.328360902255639, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 27084 + }, + { + "epoch": 1.328421052631579, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 27085 + }, + { + "epoch": 1.3284812030075188, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 27086 + }, + { + "epoch": 1.3285413533834587, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 27087 + }, + { + "epoch": 1.3286015037593986, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 27088 + }, + { + "epoch": 1.3286616541353382, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 27089 + }, + { + "epoch": 1.3287218045112783, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 27090 + }, + { + "epoch": 1.328781954887218, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 27091 + }, + { + "epoch": 1.3288421052631578, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 27092 + }, + { + "epoch": 1.3289022556390977, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 27093 + }, + { + "epoch": 1.3289624060150376, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 27094 + }, + { + "epoch": 1.3290225563909774, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 27095 + }, + { + "epoch": 1.3290827067669173, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 27096 + }, + { + "epoch": 1.3291428571428572, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 27097 + }, + { + "epoch": 1.329203007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 27098 + }, + { + "epoch": 1.329263157894737, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 27099 + }, + { + "epoch": 1.3293233082706766, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 27100 + }, + { + "epoch": 1.3293834586466167, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 27101 + }, + { + "epoch": 1.3294436090225563, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 27102 + }, + { + "epoch": 1.3295037593984962, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 27103 + }, + { + "epoch": 1.329563909774436, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 27104 + }, + { + "epoch": 1.329624060150376, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 27105 + }, + { + "epoch": 1.3296842105263158, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 27106 + }, + { + "epoch": 1.3297443609022557, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 27107 + }, + { + "epoch": 1.3298045112781955, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 27108 + }, + { + "epoch": 1.3298646616541354, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 27109 + }, + { + "epoch": 1.3299248120300753, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 27110 + }, + { + "epoch": 1.329984962406015, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 27111 + }, + { + "epoch": 1.3300451127819548, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 27112 + }, + { + "epoch": 1.3301052631578947, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 27113 + }, + { + "epoch": 1.3301654135338346, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 27114 + }, + { + "epoch": 1.3302255639097744, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 27115 + }, + { + "epoch": 1.3302857142857143, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 27116 + }, + { + "epoch": 1.3303458646616542, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 27117 + }, + { + "epoch": 1.330406015037594, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 27118 + }, + { + "epoch": 1.330466165413534, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 27119 + }, + { + "epoch": 1.3305263157894736, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 27120 + }, + { + "epoch": 1.3305864661654137, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2496, + "step": 27121 + }, + { + "epoch": 1.3306466165413533, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 27122 + }, + { + "epoch": 1.3307067669172932, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 27123 + }, + { + "epoch": 1.330766917293233, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 27124 + }, + { + "epoch": 1.330827067669173, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 27125 + }, + { + "epoch": 1.3308872180451128, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 27126 + }, + { + "epoch": 1.3309473684210527, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 27127 + }, + { + "epoch": 1.3310075187969925, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 27128 + }, + { + "epoch": 1.3310676691729324, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 27129 + }, + { + "epoch": 1.3311278195488723, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 27130 + }, + { + "epoch": 1.331187969924812, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 27131 + }, + { + "epoch": 1.331248120300752, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 27132 + }, + { + "epoch": 1.3313082706766917, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 27133 + }, + { + "epoch": 1.3313684210526315, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 27134 + }, + { + "epoch": 1.3314285714285714, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1646, + "step": 27135 + }, + { + "epoch": 1.3314887218045113, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 27136 + }, + { + "epoch": 1.3315488721804511, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 27137 + }, + { + "epoch": 1.331609022556391, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 27138 + }, + { + "epoch": 1.331669172932331, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 27139 + }, + { + "epoch": 1.3317293233082705, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 27140 + }, + { + "epoch": 1.3317894736842106, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 27141 + }, + { + "epoch": 1.3318496240601503, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9487, + "step": 27142 + }, + { + "epoch": 1.3319097744360902, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 27143 + }, + { + "epoch": 1.33196992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 27144 + }, + { + "epoch": 1.33203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 27145 + }, + { + "epoch": 1.3320902255639098, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 27146 + }, + { + "epoch": 1.3321503759398496, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 27147 + }, + { + "epoch": 1.3322105263157895, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 27148 + }, + { + "epoch": 1.3322706766917294, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 27149 + }, + { + "epoch": 1.3323308270676693, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 27150 + }, + { + "epoch": 1.332390977443609, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 27151 + }, + { + "epoch": 1.332451127819549, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 27152 + }, + { + "epoch": 1.3325112781954886, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 27153 + }, + { + "epoch": 1.3325714285714285, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 27154 + }, + { + "epoch": 1.3326315789473684, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 27155 + }, + { + "epoch": 1.3326917293233083, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 27156 + }, + { + "epoch": 1.3327518796992481, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 27157 + }, + { + "epoch": 1.332812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 27158 + }, + { + "epoch": 1.3328721804511279, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 27159 + }, + { + "epoch": 1.3329323308270677, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 27160 + }, + { + "epoch": 1.3329924812030076, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 27161 + }, + { + "epoch": 1.3330526315789473, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 27162 + }, + { + "epoch": 1.3331127819548871, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 27163 + }, + { + "epoch": 1.333172932330827, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 27164 + }, + { + "epoch": 1.3332330827067669, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 27165 + }, + { + "epoch": 1.3332932330827068, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 27166 + }, + { + "epoch": 1.3333533834586466, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 27167 + }, + { + "epoch": 1.3334135338345865, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 27168 + }, + { + "epoch": 1.3334736842105264, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 27169 + }, + { + "epoch": 1.3335338345864662, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 27170 + }, + { + "epoch": 1.3335939849624059, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 27171 + }, + { + "epoch": 1.333654135338346, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1564, + "step": 27172 + }, + { + "epoch": 1.3337142857142856, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 27173 + }, + { + "epoch": 1.3337744360902255, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 27174 + }, + { + "epoch": 1.3338345864661654, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 27175 + }, + { + "epoch": 1.3338947368421052, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 27176 + }, + { + "epoch": 1.3339548872180451, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 27177 + }, + { + "epoch": 1.334015037593985, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1998, + "step": 27178 + }, + { + "epoch": 1.3340751879699249, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 27179 + }, + { + "epoch": 1.3341353383458647, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 27180 + }, + { + "epoch": 1.3341954887218046, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 27181 + }, + { + "epoch": 1.3342556390977443, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 27182 + }, + { + "epoch": 1.3343157894736843, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1529, + "step": 27183 + }, + { + "epoch": 1.334375939849624, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 27184 + }, + { + "epoch": 1.3344360902255639, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 27185 + }, + { + "epoch": 1.3344962406015037, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 27186 + }, + { + "epoch": 1.3345563909774436, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 27187 + }, + { + "epoch": 1.3346165413533835, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1271, + "step": 27188 + }, + { + "epoch": 1.3346766917293234, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 27189 + }, + { + "epoch": 1.3347368421052632, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 27190 + }, + { + "epoch": 1.3347969924812029, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 27191 + }, + { + "epoch": 1.334857142857143, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 27192 + }, + { + "epoch": 1.3349172932330826, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 27193 + }, + { + "epoch": 1.3349774436090225, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1436, + "step": 27194 + }, + { + "epoch": 1.3350375939849624, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 27195 + }, + { + "epoch": 1.3350977443609022, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1853, + "step": 27196 + }, + { + "epoch": 1.335157894736842, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 27197 + }, + { + "epoch": 1.335218045112782, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 27198 + }, + { + "epoch": 1.3352781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 27199 + }, + { + "epoch": 1.3353383458646617, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 27200 + }, + { + "epoch": 1.3353984962406016, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 27201 + }, + { + "epoch": 1.3354586466165412, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 27202 + }, + { + "epoch": 1.3355187969924813, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 27203 + }, + { + "epoch": 1.335578947368421, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 27204 + }, + { + "epoch": 1.3356390977443608, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 27205 + }, + { + "epoch": 1.3356992481203007, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 27206 + }, + { + "epoch": 1.3357593984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 27207 + }, + { + "epoch": 1.3358195488721805, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 27208 + }, + { + "epoch": 1.3358796992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 27209 + }, + { + "epoch": 1.3359398496240602, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 27210 + }, + { + "epoch": 1.336, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 27211 + }, + { + "epoch": 1.33606015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 27212 + }, + { + "epoch": 1.3361203007518796, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 27213 + }, + { + "epoch": 1.3361804511278195, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 27214 + }, + { + "epoch": 1.3362406015037593, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 27215 + }, + { + "epoch": 1.3363007518796992, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 27216 + }, + { + "epoch": 1.336360902255639, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 27217 + }, + { + "epoch": 1.336421052631579, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 27218 + }, + { + "epoch": 1.3364812030075188, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 27219 + }, + { + "epoch": 1.3365413533834587, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 27220 + }, + { + "epoch": 1.3366015037593986, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 27221 + }, + { + "epoch": 1.3366616541353382, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 27222 + }, + { + "epoch": 1.3367218045112783, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 27223 + }, + { + "epoch": 1.336781954887218, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 27224 + }, + { + "epoch": 1.3368421052631578, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 27225 + }, + { + "epoch": 1.3369022556390977, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 27226 + }, + { + "epoch": 1.3369624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 27227 + }, + { + "epoch": 1.3370225563909774, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 27228 + }, + { + "epoch": 1.3370827067669173, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 27229 + }, + { + "epoch": 1.3371428571428572, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 27230 + }, + { + "epoch": 1.337203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 27231 + }, + { + "epoch": 1.337263157894737, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 27232 + }, + { + "epoch": 1.3373233082706766, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 27233 + }, + { + "epoch": 1.3373834586466167, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 27234 + }, + { + "epoch": 1.3374436090225563, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 27235 + }, + { + "epoch": 1.3375037593984962, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 27236 + }, + { + "epoch": 1.337563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 27237 + }, + { + "epoch": 1.337624060150376, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 27238 + }, + { + "epoch": 1.3376842105263158, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 27239 + }, + { + "epoch": 1.3377443609022557, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 27240 + }, + { + "epoch": 1.3378045112781956, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 27241 + }, + { + "epoch": 1.3378646616541352, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 27242 + }, + { + "epoch": 1.3379248120300753, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 27243 + }, + { + "epoch": 1.337984962406015, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 27244 + }, + { + "epoch": 1.3380451127819548, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 27245 + }, + { + "epoch": 1.3381052631578947, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 27246 + }, + { + "epoch": 1.3381654135338346, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 27247 + }, + { + "epoch": 1.3382255639097744, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 27248 + }, + { + "epoch": 1.3382857142857143, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 27249 + }, + { + "epoch": 1.3383458646616542, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 27250 + }, + { + "epoch": 1.338406015037594, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 27251 + }, + { + "epoch": 1.338466165413534, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 27252 + }, + { + "epoch": 1.3385263157894736, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 27253 + }, + { + "epoch": 1.3385864661654137, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 27254 + }, + { + "epoch": 1.3386466165413533, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 27255 + }, + { + "epoch": 1.3387067669172932, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 27256 + }, + { + "epoch": 1.338766917293233, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 27257 + }, + { + "epoch": 1.338827067669173, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9357, + "step": 27258 + }, + { + "epoch": 1.3388872180451128, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1964, + "step": 27259 + }, + { + "epoch": 1.3389473684210527, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 27260 + }, + { + "epoch": 1.3390075187969925, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 27261 + }, + { + "epoch": 1.3390676691729324, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 27262 + }, + { + "epoch": 1.3391278195488723, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 27263 + }, + { + "epoch": 1.339187969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 27264 + }, + { + "epoch": 1.3392481203007518, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 27265 + }, + { + "epoch": 1.3393082706766917, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 27266 + }, + { + "epoch": 1.3393684210526315, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 27267 + }, + { + "epoch": 1.3394285714285714, + "grad_norm": 1.390625, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 27268 + }, + { + "epoch": 1.3394887218045113, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 27269 + }, + { + "epoch": 1.3395488721804512, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 27270 + }, + { + "epoch": 1.339609022556391, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2279, + "step": 27271 + }, + { + "epoch": 1.339669172932331, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 27272 + }, + { + "epoch": 1.3397293233082705, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 27273 + }, + { + "epoch": 1.3397894736842106, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 27274 + }, + { + "epoch": 1.3398496240601503, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 27275 + }, + { + "epoch": 1.3399097744360902, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 27276 + }, + { + "epoch": 1.33996992481203, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 27277 + }, + { + "epoch": 1.34003007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 27278 + }, + { + "epoch": 1.3400902255639098, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 27279 + }, + { + "epoch": 1.3401503759398496, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 27280 + }, + { + "epoch": 1.3402105263157895, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 27281 + }, + { + "epoch": 1.3402706766917294, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 27282 + }, + { + "epoch": 1.3403308270676693, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 27283 + }, + { + "epoch": 1.340390977443609, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 27284 + }, + { + "epoch": 1.340451127819549, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 27285 + }, + { + "epoch": 1.3405112781954887, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 27286 + }, + { + "epoch": 1.3405714285714285, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 27287 + }, + { + "epoch": 1.3406315789473684, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 27288 + }, + { + "epoch": 1.3406917293233083, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 27289 + }, + { + "epoch": 1.3407518796992481, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 27290 + }, + { + "epoch": 1.340812030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1553, + "step": 27291 + }, + { + "epoch": 1.3408721804511279, + "grad_norm": 1.4765625, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 27292 + }, + { + "epoch": 1.3409323308270675, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 27293 + }, + { + "epoch": 1.3409924812030076, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 27294 + }, + { + "epoch": 1.3410526315789473, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 27295 + }, + { + "epoch": 1.3411127819548871, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 27296 + }, + { + "epoch": 1.341172932330827, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 27297 + }, + { + "epoch": 1.3412330827067669, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 27298 + }, + { + "epoch": 1.3412932330827068, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 27299 + }, + { + "epoch": 1.3413533834586466, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 27300 + }, + { + "epoch": 1.3414135338345865, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1421, + "step": 27301 + }, + { + "epoch": 1.3414736842105264, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 27302 + }, + { + "epoch": 1.3415338345864662, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 27303 + }, + { + "epoch": 1.341593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 27304 + }, + { + "epoch": 1.341654135338346, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 27305 + }, + { + "epoch": 1.3417142857142856, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 27306 + }, + { + "epoch": 1.3417744360902255, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 27307 + }, + { + "epoch": 1.3418345864661654, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 27308 + }, + { + "epoch": 1.3418947368421053, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1704, + "step": 27309 + }, + { + "epoch": 1.3419548872180451, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 27310 + }, + { + "epoch": 1.342015037593985, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 27311 + }, + { + "epoch": 1.3420751879699249, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 27312 + }, + { + "epoch": 1.3421353383458647, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 27313 + }, + { + "epoch": 1.3421954887218046, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 27314 + }, + { + "epoch": 1.3422556390977443, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 27315 + }, + { + "epoch": 1.3423157894736841, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 27316 + }, + { + "epoch": 1.342375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 27317 + }, + { + "epoch": 1.3424360902255639, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 27318 + }, + { + "epoch": 1.3424962406015037, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 27319 + }, + { + "epoch": 1.3425563909774436, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 27320 + }, + { + "epoch": 1.3426165413533835, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 27321 + }, + { + "epoch": 1.3426766917293234, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 27322 + }, + { + "epoch": 1.3427368421052632, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 27323 + }, + { + "epoch": 1.3427969924812029, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 27324 + }, + { + "epoch": 1.342857142857143, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 27325 + }, + { + "epoch": 1.3429172932330826, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 27326 + }, + { + "epoch": 1.3429774436090225, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 27327 + }, + { + "epoch": 1.3430375939849624, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 27328 + }, + { + "epoch": 1.3430977443609022, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1748, + "step": 27329 + }, + { + "epoch": 1.343157894736842, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 27330 + }, + { + "epoch": 1.343218045112782, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 27331 + }, + { + "epoch": 1.3432781954887218, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 27332 + }, + { + "epoch": 1.3433383458646617, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 27333 + }, + { + "epoch": 1.3433984962406016, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 27334 + }, + { + "epoch": 1.3434586466165412, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 27335 + }, + { + "epoch": 1.3435187969924813, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 27336 + }, + { + "epoch": 1.343578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 27337 + }, + { + "epoch": 1.3436390977443609, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 27338 + }, + { + "epoch": 1.3436992481203007, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 27339 + }, + { + "epoch": 1.3437593984962406, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 27340 + }, + { + "epoch": 1.3438195488721805, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 27341 + }, + { + "epoch": 1.3438796992481203, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 27342 + }, + { + "epoch": 1.3439398496240602, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1923, + "step": 27343 + }, + { + "epoch": 1.3439999999999999, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 27344 + }, + { + "epoch": 1.34406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 27345 + }, + { + "epoch": 1.3441203007518796, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 27346 + }, + { + "epoch": 1.3441804511278195, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 27347 + }, + { + "epoch": 1.3442406015037593, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 27348 + }, + { + "epoch": 1.3443007518796992, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 27349 + }, + { + "epoch": 1.344360902255639, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 27350 + }, + { + "epoch": 1.344421052631579, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 27351 + }, + { + "epoch": 1.3444812030075188, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 27352 + }, + { + "epoch": 1.3445413533834587, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 27353 + }, + { + "epoch": 1.3446015037593986, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 27354 + }, + { + "epoch": 1.3446616541353382, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 27355 + }, + { + "epoch": 1.3447218045112783, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 27356 + }, + { + "epoch": 1.344781954887218, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 27357 + }, + { + "epoch": 1.3448421052631578, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 27358 + }, + { + "epoch": 1.3449022556390977, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 27359 + }, + { + "epoch": 1.3449624060150376, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 27360 + }, + { + "epoch": 1.3450225563909775, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 27361 + }, + { + "epoch": 1.3450827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 27362 + }, + { + "epoch": 1.3451428571428572, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 27363 + }, + { + "epoch": 1.345203007518797, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 27364 + }, + { + "epoch": 1.345263157894737, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 27365 + }, + { + "epoch": 1.3453233082706766, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 27366 + }, + { + "epoch": 1.3453834586466165, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 27367 + }, + { + "epoch": 1.3454436090225563, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 27368 + }, + { + "epoch": 1.3455037593984962, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 27369 + }, + { + "epoch": 1.345563909774436, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 27370 + }, + { + "epoch": 1.345624060150376, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 27371 + }, + { + "epoch": 1.3456842105263158, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 27372 + }, + { + "epoch": 1.3457443609022557, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 27373 + }, + { + "epoch": 1.3458045112781956, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 27374 + }, + { + "epoch": 1.3458646616541352, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 27375 + }, + { + "epoch": 1.3459248120300753, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 27376 + }, + { + "epoch": 1.345984962406015, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 27377 + }, + { + "epoch": 1.3460451127819548, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 27378 + }, + { + "epoch": 1.3461052631578947, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 27379 + }, + { + "epoch": 1.3461654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 27380 + }, + { + "epoch": 1.3462255639097744, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 27381 + }, + { + "epoch": 1.3462857142857143, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 27382 + }, + { + "epoch": 1.3463458646616542, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 27383 + }, + { + "epoch": 1.346406015037594, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 27384 + }, + { + "epoch": 1.346466165413534, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 27385 + }, + { + "epoch": 1.3465263157894736, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1731, + "step": 27386 + }, + { + "epoch": 1.3465864661654137, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 27387 + }, + { + "epoch": 1.3466466165413533, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 27388 + }, + { + "epoch": 1.3467067669172932, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 27389 + }, + { + "epoch": 1.346766917293233, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 27390 + }, + { + "epoch": 1.346827067669173, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9327, + "step": 27391 + }, + { + "epoch": 1.3468872180451128, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 27392 + }, + { + "epoch": 1.3469473684210527, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 27393 + }, + { + "epoch": 1.3470075187969925, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9667, + "step": 27394 + }, + { + "epoch": 1.3470676691729324, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 27395 + }, + { + "epoch": 1.3471278195488723, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 27396 + }, + { + "epoch": 1.347187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 27397 + }, + { + "epoch": 1.3472481203007518, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.2527, + "step": 27398 + }, + { + "epoch": 1.3473082706766917, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 27399 + }, + { + "epoch": 1.3473684210526315, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 27400 + }, + { + "epoch": 1.3474285714285714, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 27401 + }, + { + "epoch": 1.3474887218045113, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 27402 + }, + { + "epoch": 1.3475488721804512, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 27403 + }, + { + "epoch": 1.347609022556391, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 27404 + }, + { + "epoch": 1.347669172932331, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 27405 + }, + { + "epoch": 1.3477293233082706, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 27406 + }, + { + "epoch": 1.3477894736842106, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 27407 + }, + { + "epoch": 1.3478496240601503, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 27408 + }, + { + "epoch": 1.3479097744360902, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 27409 + }, + { + "epoch": 1.34796992481203, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 27410 + }, + { + "epoch": 1.34803007518797, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 27411 + }, + { + "epoch": 1.3480902255639098, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 27412 + }, + { + "epoch": 1.3481503759398497, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 27413 + }, + { + "epoch": 1.3482105263157895, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 27414 + }, + { + "epoch": 1.3482706766917294, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 27415 + }, + { + "epoch": 1.3483308270676693, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1464, + "step": 27416 + }, + { + "epoch": 1.348390977443609, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 27417 + }, + { + "epoch": 1.3484511278195488, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 27418 + }, + { + "epoch": 1.3485112781954887, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 27419 + }, + { + "epoch": 1.3485714285714285, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 27420 + }, + { + "epoch": 1.3486315789473684, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 27421 + }, + { + "epoch": 1.3486917293233083, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 27422 + }, + { + "epoch": 1.3487518796992481, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 27423 + }, + { + "epoch": 1.348812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 27424 + }, + { + "epoch": 1.348872180451128, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 27425 + }, + { + "epoch": 1.3489323308270675, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 27426 + }, + { + "epoch": 1.3489924812030076, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 27427 + }, + { + "epoch": 1.3490526315789473, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 27428 + }, + { + "epoch": 1.3491127819548872, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 27429 + }, + { + "epoch": 1.349172932330827, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 27430 + }, + { + "epoch": 1.349233082706767, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 27431 + }, + { + "epoch": 1.3492932330827068, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 27432 + }, + { + "epoch": 1.3493533834586466, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1689, + "step": 27433 + }, + { + "epoch": 1.3494135338345865, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 27434 + }, + { + "epoch": 1.3494736842105264, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 27435 + }, + { + "epoch": 1.3495338345864663, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 27436 + }, + { + "epoch": 1.349593984962406, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 27437 + }, + { + "epoch": 1.349654135338346, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 27438 + }, + { + "epoch": 1.3497142857142856, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 27439 + }, + { + "epoch": 1.3497744360902255, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 27440 + }, + { + "epoch": 1.3498345864661654, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 27441 + }, + { + "epoch": 1.3498947368421053, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 27442 + }, + { + "epoch": 1.3499548872180451, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 27443 + }, + { + "epoch": 1.350015037593985, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 27444 + }, + { + "epoch": 1.3500751879699249, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 27445 + }, + { + "epoch": 1.3501353383458647, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 27446 + }, + { + "epoch": 1.3501954887218046, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1529, + "step": 27447 + }, + { + "epoch": 1.3502556390977443, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 27448 + }, + { + "epoch": 1.3503157894736841, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 27449 + }, + { + "epoch": 1.350375939849624, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 27450 + }, + { + "epoch": 1.3504360902255639, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 27451 + }, + { + "epoch": 1.3504962406015038, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 27452 + }, + { + "epoch": 1.3505563909774436, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 27453 + }, + { + "epoch": 1.3506165413533835, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 27454 + }, + { + "epoch": 1.3506766917293234, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 27455 + }, + { + "epoch": 1.3507368421052632, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 27456 + }, + { + "epoch": 1.3507969924812029, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 27457 + }, + { + "epoch": 1.350857142857143, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 27458 + }, + { + "epoch": 1.3509172932330826, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.155, + "step": 27459 + }, + { + "epoch": 1.3509774436090225, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 27460 + }, + { + "epoch": 1.3510375939849624, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 27461 + }, + { + "epoch": 1.3510977443609022, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 27462 + }, + { + "epoch": 1.3511578947368421, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 27463 + }, + { + "epoch": 1.351218045112782, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 27464 + }, + { + "epoch": 1.3512781954887219, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 27465 + }, + { + "epoch": 1.3513383458646617, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 27466 + }, + { + "epoch": 1.3513984962406016, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 27467 + }, + { + "epoch": 1.3514586466165412, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 27468 + }, + { + "epoch": 1.3515187969924813, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 27469 + }, + { + "epoch": 1.351578947368421, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 27470 + }, + { + "epoch": 1.3516390977443609, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 27471 + }, + { + "epoch": 1.3516992481203007, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 27472 + }, + { + "epoch": 1.3517593984962406, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 27473 + }, + { + "epoch": 1.3518195488721805, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.3969, + "step": 27474 + }, + { + "epoch": 1.3518796992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 27475 + }, + { + "epoch": 1.3519398496240602, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 27476 + }, + { + "epoch": 1.3519999999999999, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2499, + "step": 27477 + }, + { + "epoch": 1.35206015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 27478 + }, + { + "epoch": 1.3521203007518796, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 27479 + }, + { + "epoch": 1.3521804511278195, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 27480 + }, + { + "epoch": 1.3522406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 27481 + }, + { + "epoch": 1.3523007518796992, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 27482 + }, + { + "epoch": 1.352360902255639, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 27483 + }, + { + "epoch": 1.352421052631579, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 27484 + }, + { + "epoch": 1.3524812030075188, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 27485 + }, + { + "epoch": 1.3525413533834587, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 27486 + }, + { + "epoch": 1.3526015037593986, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 27487 + }, + { + "epoch": 1.3526616541353382, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 27488 + }, + { + "epoch": 1.3527218045112783, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 27489 + }, + { + "epoch": 1.352781954887218, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 27490 + }, + { + "epoch": 1.3528421052631578, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 27491 + }, + { + "epoch": 1.3529022556390977, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 27492 + }, + { + "epoch": 1.3529624060150376, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 27493 + }, + { + "epoch": 1.3530225563909775, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 27494 + }, + { + "epoch": 1.3530827067669173, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 27495 + }, + { + "epoch": 1.3531428571428572, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 27496 + }, + { + "epoch": 1.353203007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 27497 + }, + { + "epoch": 1.353263157894737, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 27498 + }, + { + "epoch": 1.3533233082706766, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 27499 + }, + { + "epoch": 1.3533834586466165, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 27500 + }, + { + "epoch": 1.3534436090225563, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 27501 + }, + { + "epoch": 1.3535037593984962, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 27502 + }, + { + "epoch": 1.353563909774436, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 27503 + }, + { + "epoch": 1.353624060150376, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 27504 + }, + { + "epoch": 1.3536842105263158, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 27505 + }, + { + "epoch": 1.3537443609022557, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 27506 + }, + { + "epoch": 1.3538045112781956, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 27507 + }, + { + "epoch": 1.3538646616541352, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 27508 + }, + { + "epoch": 1.3539248120300753, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 27509 + }, + { + "epoch": 1.353984962406015, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 27510 + }, + { + "epoch": 1.3540451127819548, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 27511 + }, + { + "epoch": 1.3541052631578947, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 27512 + }, + { + "epoch": 1.3541654135338346, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 27513 + }, + { + "epoch": 1.3542255639097744, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 27514 + }, + { + "epoch": 1.3542857142857143, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 27515 + }, + { + "epoch": 1.3543458646616542, + "grad_norm": 1.5234375, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 27516 + }, + { + "epoch": 1.354406015037594, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1944, + "step": 27517 + }, + { + "epoch": 1.354466165413534, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 27518 + }, + { + "epoch": 1.3545263157894736, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 27519 + }, + { + "epoch": 1.3545864661654137, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 27520 + }, + { + "epoch": 1.3546466165413533, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 27521 + }, + { + "epoch": 1.3547067669172932, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 27522 + }, + { + "epoch": 1.354766917293233, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 27523 + }, + { + "epoch": 1.354827067669173, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 27524 + }, + { + "epoch": 1.3548872180451128, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 27525 + }, + { + "epoch": 1.3549473684210527, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 27526 + }, + { + "epoch": 1.3550075187969925, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 27527 + }, + { + "epoch": 1.3550676691729322, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 27528 + }, + { + "epoch": 1.3551278195488723, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 27529 + }, + { + "epoch": 1.355187969924812, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 27530 + }, + { + "epoch": 1.3552481203007518, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 27531 + }, + { + "epoch": 1.3553082706766917, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 27532 + }, + { + "epoch": 1.3553684210526316, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 27533 + }, + { + "epoch": 1.3554285714285714, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 27534 + }, + { + "epoch": 1.3554887218045113, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2137, + "step": 27535 + }, + { + "epoch": 1.3555488721804512, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 27536 + }, + { + "epoch": 1.355609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 27537 + }, + { + "epoch": 1.355669172932331, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 27538 + }, + { + "epoch": 1.3557293233082706, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 27539 + }, + { + "epoch": 1.3557894736842107, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 27540 + }, + { + "epoch": 1.3558496240601503, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 27541 + }, + { + "epoch": 1.3559097744360902, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 27542 + }, + { + "epoch": 1.35596992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 27543 + }, + { + "epoch": 1.35603007518797, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 27544 + }, + { + "epoch": 1.3560902255639098, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 27545 + }, + { + "epoch": 1.3561503759398497, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 27546 + }, + { + "epoch": 1.3562105263157895, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1773, + "step": 27547 + }, + { + "epoch": 1.3562706766917294, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 27548 + }, + { + "epoch": 1.3563308270676693, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 27549 + }, + { + "epoch": 1.356390977443609, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 27550 + }, + { + "epoch": 1.3564511278195488, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 27551 + }, + { + "epoch": 1.3565112781954887, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 27552 + }, + { + "epoch": 1.3565714285714285, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 27553 + }, + { + "epoch": 1.3566315789473684, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 27554 + }, + { + "epoch": 1.3566917293233083, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 27555 + }, + { + "epoch": 1.3567518796992482, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 27556 + }, + { + "epoch": 1.356812030075188, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 27557 + }, + { + "epoch": 1.356872180451128, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 27558 + }, + { + "epoch": 1.3569323308270675, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 27559 + }, + { + "epoch": 1.3569924812030076, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 27560 + }, + { + "epoch": 1.3570526315789473, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 27561 + }, + { + "epoch": 1.3571127819548872, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 27562 + }, + { + "epoch": 1.357172932330827, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 27563 + }, + { + "epoch": 1.357233082706767, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 27564 + }, + { + "epoch": 1.3572932330827068, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 27565 + }, + { + "epoch": 1.3573533834586466, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 27566 + }, + { + "epoch": 1.3574135338345865, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 27567 + }, + { + "epoch": 1.3574736842105264, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 27568 + }, + { + "epoch": 1.3575338345864663, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.162, + "step": 27569 + }, + { + "epoch": 1.357593984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 27570 + }, + { + "epoch": 1.357654135338346, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1784, + "step": 27571 + }, + { + "epoch": 1.3577142857142857, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 27572 + }, + { + "epoch": 1.3577744360902255, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 27573 + }, + { + "epoch": 1.3578345864661654, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 27574 + }, + { + "epoch": 1.3578947368421053, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 27575 + }, + { + "epoch": 1.3579548872180451, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 27576 + }, + { + "epoch": 1.358015037593985, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2132, + "step": 27577 + }, + { + "epoch": 1.3580751879699249, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 27578 + }, + { + "epoch": 1.3581353383458645, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 27579 + }, + { + "epoch": 1.3581954887218046, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 27580 + }, + { + "epoch": 1.3582556390977443, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 27581 + }, + { + "epoch": 1.3583157894736841, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 27582 + }, + { + "epoch": 1.358375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 27583 + }, + { + "epoch": 1.3584360902255639, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 27584 + }, + { + "epoch": 1.3584962406015038, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 27585 + }, + { + "epoch": 1.3585563909774436, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1721, + "step": 27586 + }, + { + "epoch": 1.3586165413533835, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 27587 + }, + { + "epoch": 1.3586766917293234, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 27588 + }, + { + "epoch": 1.3587368421052632, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 27589 + }, + { + "epoch": 1.358796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 27590 + }, + { + "epoch": 1.358857142857143, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 27591 + }, + { + "epoch": 1.3589172932330826, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 27592 + }, + { + "epoch": 1.3589774436090225, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1924, + "step": 27593 + }, + { + "epoch": 1.3590375939849624, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 27594 + }, + { + "epoch": 1.3590977443609022, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 27595 + }, + { + "epoch": 1.3591578947368421, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 27596 + }, + { + "epoch": 1.359218045112782, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 27597 + }, + { + "epoch": 1.3592781954887219, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1638, + "step": 27598 + }, + { + "epoch": 1.3593383458646617, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2173, + "step": 27599 + }, + { + "epoch": 1.3593984962406016, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 27600 + }, + { + "epoch": 1.3594586466165413, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 27601 + }, + { + "epoch": 1.3595187969924811, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 27602 + }, + { + "epoch": 1.359578947368421, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 27603 + }, + { + "epoch": 1.3596390977443609, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 27604 + }, + { + "epoch": 1.3596992481203007, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 27605 + }, + { + "epoch": 1.3597593984962406, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 27606 + }, + { + "epoch": 1.3598195488721805, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 27607 + }, + { + "epoch": 1.3598796992481204, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 27608 + }, + { + "epoch": 1.3599398496240602, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 27609 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 27610 + }, + { + "epoch": 1.36006015037594, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 27611 + }, + { + "epoch": 1.3601203007518796, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 27612 + }, + { + "epoch": 1.3601804511278195, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2428, + "step": 27613 + }, + { + "epoch": 1.3602406015037594, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 27614 + }, + { + "epoch": 1.3603007518796992, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 27615 + }, + { + "epoch": 1.360360902255639, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 27616 + }, + { + "epoch": 1.360421052631579, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 27617 + }, + { + "epoch": 1.3604812030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 27618 + }, + { + "epoch": 1.3605413533834587, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 27619 + }, + { + "epoch": 1.3606015037593986, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 27620 + }, + { + "epoch": 1.3606616541353382, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 27621 + }, + { + "epoch": 1.3607218045112783, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 27622 + }, + { + "epoch": 1.360781954887218, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 27623 + }, + { + "epoch": 1.3608421052631579, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 27624 + }, + { + "epoch": 1.3609022556390977, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1735, + "step": 27625 + }, + { + "epoch": 1.3609624060150376, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 27626 + }, + { + "epoch": 1.3610225563909775, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2511, + "step": 27627 + }, + { + "epoch": 1.3610827067669173, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 27628 + }, + { + "epoch": 1.3611428571428572, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 27629 + }, + { + "epoch": 1.3612030075187969, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 27630 + }, + { + "epoch": 1.361263157894737, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 27631 + }, + { + "epoch": 1.3613233082706766, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 27632 + }, + { + "epoch": 1.3613834586466165, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 27633 + }, + { + "epoch": 1.3614436090225563, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 27634 + }, + { + "epoch": 1.3615037593984962, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 27635 + }, + { + "epoch": 1.361563909774436, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 27636 + }, + { + "epoch": 1.361624060150376, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 27637 + }, + { + "epoch": 1.3616842105263158, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 27638 + }, + { + "epoch": 1.3617443609022557, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 27639 + }, + { + "epoch": 1.3618045112781956, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 27640 + }, + { + "epoch": 1.3618646616541352, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 27641 + }, + { + "epoch": 1.3619248120300753, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 27642 + }, + { + "epoch": 1.361984962406015, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 27643 + }, + { + "epoch": 1.3620451127819548, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 27644 + }, + { + "epoch": 1.3621052631578947, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 27645 + }, + { + "epoch": 1.3621654135338346, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 27646 + }, + { + "epoch": 1.3622255639097745, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 27647 + }, + { + "epoch": 1.3622857142857143, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 27648 + }, + { + "epoch": 1.3623458646616542, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 27649 + }, + { + "epoch": 1.362406015037594, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.2539, + "step": 27650 + }, + { + "epoch": 1.362466165413534, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 27651 + }, + { + "epoch": 1.3625263157894736, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 27652 + }, + { + "epoch": 1.3625864661654135, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 27653 + }, + { + "epoch": 1.3626466165413533, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 27654 + }, + { + "epoch": 1.3627067669172932, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 27655 + }, + { + "epoch": 1.362766917293233, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 27656 + }, + { + "epoch": 1.362827067669173, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 27657 + }, + { + "epoch": 1.3628872180451128, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 27658 + }, + { + "epoch": 1.3629473684210527, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 27659 + }, + { + "epoch": 1.3630075187969926, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 27660 + }, + { + "epoch": 1.3630676691729322, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 27661 + }, + { + "epoch": 1.3631278195488723, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 27662 + }, + { + "epoch": 1.363187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 27663 + }, + { + "epoch": 1.3632481203007518, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 27664 + }, + { + "epoch": 1.3633082706766917, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 27665 + }, + { + "epoch": 1.3633684210526316, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 27666 + }, + { + "epoch": 1.3634285714285714, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 27667 + }, + { + "epoch": 1.3634887218045113, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 27668 + }, + { + "epoch": 1.3635488721804512, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 27669 + }, + { + "epoch": 1.363609022556391, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 27670 + }, + { + "epoch": 1.363669172932331, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 27671 + }, + { + "epoch": 1.3637293233082706, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 27672 + }, + { + "epoch": 1.3637894736842107, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 27673 + }, + { + "epoch": 1.3638496240601503, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 27674 + }, + { + "epoch": 1.3639097744360902, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 27675 + }, + { + "epoch": 1.36396992481203, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2098, + "step": 27676 + }, + { + "epoch": 1.36403007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 27677 + }, + { + "epoch": 1.3640902255639098, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 27678 + }, + { + "epoch": 1.3641503759398497, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 27679 + }, + { + "epoch": 1.3642105263157895, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 27680 + }, + { + "epoch": 1.3642706766917292, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 27681 + }, + { + "epoch": 1.3643308270676693, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 27682 + }, + { + "epoch": 1.364390977443609, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 27683 + }, + { + "epoch": 1.3644511278195488, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 27684 + }, + { + "epoch": 1.3645112781954887, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 27685 + }, + { + "epoch": 1.3645714285714285, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1744, + "step": 27686 + }, + { + "epoch": 1.3646315789473684, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 27687 + }, + { + "epoch": 1.3646917293233083, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 27688 + }, + { + "epoch": 1.3647518796992482, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 27689 + }, + { + "epoch": 1.364812030075188, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 27690 + }, + { + "epoch": 1.364872180451128, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 27691 + }, + { + "epoch": 1.3649323308270676, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 27692 + }, + { + "epoch": 1.3649924812030076, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 27693 + }, + { + "epoch": 1.3650526315789473, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 27694 + }, + { + "epoch": 1.3651127819548872, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 27695 + }, + { + "epoch": 1.365172932330827, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 27696 + }, + { + "epoch": 1.365233082706767, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9086, + "step": 27697 + }, + { + "epoch": 1.3652932330827068, + "grad_norm": 1.8359375, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 27698 + }, + { + "epoch": 1.3653533834586467, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 27699 + }, + { + "epoch": 1.3654135338345865, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 27700 + }, + { + "epoch": 1.3654736842105264, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 27701 + }, + { + "epoch": 1.3655338345864663, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 27702 + }, + { + "epoch": 1.365593984962406, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 27703 + }, + { + "epoch": 1.3656541353383458, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 27704 + }, + { + "epoch": 1.3657142857142857, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 27705 + }, + { + "epoch": 1.3657744360902255, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 27706 + }, + { + "epoch": 1.3658345864661654, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 27707 + }, + { + "epoch": 1.3658947368421053, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 27708 + }, + { + "epoch": 1.3659548872180451, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 27709 + }, + { + "epoch": 1.366015037593985, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 27710 + }, + { + "epoch": 1.3660751879699249, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 27711 + }, + { + "epoch": 1.3661353383458645, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 27712 + }, + { + "epoch": 1.3661954887218046, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 27713 + }, + { + "epoch": 1.3662556390977443, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 27714 + }, + { + "epoch": 1.3663157894736842, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 27715 + }, + { + "epoch": 1.366375939849624, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 27716 + }, + { + "epoch": 1.366436090225564, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 27717 + }, + { + "epoch": 1.3664962406015038, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 27718 + }, + { + "epoch": 1.3665563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 27719 + }, + { + "epoch": 1.3666165413533835, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 27720 + }, + { + "epoch": 1.3666766917293234, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 27721 + }, + { + "epoch": 1.3667368421052632, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 27722 + }, + { + "epoch": 1.366796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 27723 + }, + { + "epoch": 1.366857142857143, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 27724 + }, + { + "epoch": 1.3669172932330826, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 27725 + }, + { + "epoch": 1.3669774436090225, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 27726 + }, + { + "epoch": 1.3670375939849624, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 27727 + }, + { + "epoch": 1.3670977443609023, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 27728 + }, + { + "epoch": 1.3671578947368421, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 27729 + }, + { + "epoch": 1.367218045112782, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 27730 + }, + { + "epoch": 1.3672781954887219, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 27731 + }, + { + "epoch": 1.3673383458646615, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 27732 + }, + { + "epoch": 1.3673984962406016, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 27733 + }, + { + "epoch": 1.3674586466165413, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 27734 + }, + { + "epoch": 1.3675187969924811, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 27735 + }, + { + "epoch": 1.367578947368421, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 27736 + }, + { + "epoch": 1.3676390977443609, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 27737 + }, + { + "epoch": 1.3676992481203007, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 27738 + }, + { + "epoch": 1.3677593984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 27739 + }, + { + "epoch": 1.3678195488721805, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 27740 + }, + { + "epoch": 1.3678796992481204, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 27741 + }, + { + "epoch": 1.3679398496240602, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 27742 + }, + { + "epoch": 1.3679999999999999, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 27743 + }, + { + "epoch": 1.36806015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 27744 + }, + { + "epoch": 1.3681203007518796, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 27745 + }, + { + "epoch": 1.3681804511278195, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 27746 + }, + { + "epoch": 1.3682406015037594, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 27747 + }, + { + "epoch": 1.3683007518796992, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 27748 + }, + { + "epoch": 1.3683609022556391, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 27749 + }, + { + "epoch": 1.368421052631579, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 27750 + }, + { + "epoch": 1.3684812030075189, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 27751 + }, + { + "epoch": 1.3685413533834587, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 27752 + }, + { + "epoch": 1.3686015037593986, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 27753 + }, + { + "epoch": 1.3686616541353382, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 27754 + }, + { + "epoch": 1.3687218045112781, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 27755 + }, + { + "epoch": 1.368781954887218, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 27756 + }, + { + "epoch": 1.3688421052631579, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 27757 + }, + { + "epoch": 1.3689022556390977, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 27758 + }, + { + "epoch": 1.3689624060150376, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 27759 + }, + { + "epoch": 1.3690225563909775, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 27760 + }, + { + "epoch": 1.3690827067669173, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 27761 + }, + { + "epoch": 1.3691428571428572, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2071, + "step": 27762 + }, + { + "epoch": 1.3692030075187969, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 27763 + }, + { + "epoch": 1.369263157894737, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 27764 + }, + { + "epoch": 1.3693233082706766, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 27765 + }, + { + "epoch": 1.3693834586466165, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 27766 + }, + { + "epoch": 1.3694436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 27767 + }, + { + "epoch": 1.3695037593984962, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 27768 + }, + { + "epoch": 1.369563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 27769 + }, + { + "epoch": 1.369624060150376, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 27770 + }, + { + "epoch": 1.3696842105263158, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 27771 + }, + { + "epoch": 1.3697443609022557, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 27772 + }, + { + "epoch": 1.3698045112781956, + "grad_norm": 1.8984375, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 27773 + }, + { + "epoch": 1.3698646616541352, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 27774 + }, + { + "epoch": 1.3699248120300753, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 27775 + }, + { + "epoch": 1.369984962406015, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 27776 + }, + { + "epoch": 1.3700451127819548, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 27777 + }, + { + "epoch": 1.3701052631578947, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 27778 + }, + { + "epoch": 1.3701654135338346, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 27779 + }, + { + "epoch": 1.3702255639097745, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1865, + "step": 27780 + }, + { + "epoch": 1.3702857142857143, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 27781 + }, + { + "epoch": 1.3703458646616542, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 27782 + }, + { + "epoch": 1.3704060150375938, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 27783 + }, + { + "epoch": 1.370466165413534, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 27784 + }, + { + "epoch": 1.3705263157894736, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 27785 + }, + { + "epoch": 1.3705864661654135, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 27786 + }, + { + "epoch": 1.3706466165413533, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 27787 + }, + { + "epoch": 1.3707067669172932, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 27788 + }, + { + "epoch": 1.370766917293233, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.932, + "step": 27789 + }, + { + "epoch": 1.370827067669173, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 27790 + }, + { + "epoch": 1.3708872180451128, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 27791 + }, + { + "epoch": 1.3709473684210527, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 27792 + }, + { + "epoch": 1.3710075187969926, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 27793 + }, + { + "epoch": 1.3710676691729322, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 27794 + }, + { + "epoch": 1.3711278195488723, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 27795 + }, + { + "epoch": 1.371187969924812, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 27796 + }, + { + "epoch": 1.3712481203007518, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 27797 + }, + { + "epoch": 1.3713082706766917, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 27798 + }, + { + "epoch": 1.3713684210526316, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 27799 + }, + { + "epoch": 1.3714285714285714, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 27800 + }, + { + "epoch": 1.3714887218045113, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 27801 + }, + { + "epoch": 1.3715488721804512, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 27802 + }, + { + "epoch": 1.371609022556391, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 27803 + }, + { + "epoch": 1.371669172932331, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 27804 + }, + { + "epoch": 1.3717293233082706, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 27805 + }, + { + "epoch": 1.3717894736842104, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 27806 + }, + { + "epoch": 1.3718496240601503, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 27807 + }, + { + "epoch": 1.3719097744360902, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 27808 + }, + { + "epoch": 1.37196992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1891, + "step": 27809 + }, + { + "epoch": 1.37203007518797, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 27810 + }, + { + "epoch": 1.3720902255639098, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 27811 + }, + { + "epoch": 1.3721503759398497, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 27812 + }, + { + "epoch": 1.3722105263157895, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 27813 + }, + { + "epoch": 1.3722706766917292, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 27814 + }, + { + "epoch": 1.3723308270676693, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 27815 + }, + { + "epoch": 1.372390977443609, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 27816 + }, + { + "epoch": 1.3724511278195488, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 27817 + }, + { + "epoch": 1.3725112781954887, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 27818 + }, + { + "epoch": 1.3725714285714286, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 27819 + }, + { + "epoch": 1.3726315789473684, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 27820 + }, + { + "epoch": 1.3726917293233083, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 27821 + }, + { + "epoch": 1.3727518796992482, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 27822 + }, + { + "epoch": 1.372812030075188, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1982, + "step": 27823 + }, + { + "epoch": 1.372872180451128, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 27824 + }, + { + "epoch": 1.3729323308270676, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 27825 + }, + { + "epoch": 1.3729924812030077, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 27826 + }, + { + "epoch": 1.3730526315789473, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 27827 + }, + { + "epoch": 1.3731127819548872, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 27828 + }, + { + "epoch": 1.373172932330827, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 27829 + }, + { + "epoch": 1.373233082706767, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 27830 + }, + { + "epoch": 1.3732932330827068, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 27831 + }, + { + "epoch": 1.3733533834586467, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 27832 + }, + { + "epoch": 1.3734135338345865, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 27833 + }, + { + "epoch": 1.3734736842105262, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 27834 + }, + { + "epoch": 1.3735338345864663, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 27835 + }, + { + "epoch": 1.373593984962406, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 27836 + }, + { + "epoch": 1.3736541353383458, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 27837 + }, + { + "epoch": 1.3737142857142857, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 27838 + }, + { + "epoch": 1.3737744360902255, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 27839 + }, + { + "epoch": 1.3738345864661654, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 27840 + }, + { + "epoch": 1.3738947368421053, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 27841 + }, + { + "epoch": 1.3739548872180452, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2164, + "step": 27842 + }, + { + "epoch": 1.374015037593985, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 27843 + }, + { + "epoch": 1.374075187969925, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 27844 + }, + { + "epoch": 1.3741353383458645, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 27845 + }, + { + "epoch": 1.3741954887218046, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 27846 + }, + { + "epoch": 1.3742556390977443, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 27847 + }, + { + "epoch": 1.3743157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 27848 + }, + { + "epoch": 1.374375939849624, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 27849 + }, + { + "epoch": 1.374436090225564, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 27850 + }, + { + "epoch": 1.3744962406015038, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 27851 + }, + { + "epoch": 1.3745563909774436, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 27852 + }, + { + "epoch": 1.3746165413533835, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 27853 + }, + { + "epoch": 1.3746766917293234, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 27854 + }, + { + "epoch": 1.3747368421052633, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 27855 + }, + { + "epoch": 1.374796992481203, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 27856 + }, + { + "epoch": 1.3748571428571428, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 27857 + }, + { + "epoch": 1.3749172932330826, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 27858 + }, + { + "epoch": 1.3749774436090225, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 27859 + }, + { + "epoch": 1.3750375939849624, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 27860 + }, + { + "epoch": 1.3750977443609023, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 27861 + }, + { + "epoch": 1.3751578947368421, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 27862 + }, + { + "epoch": 1.375218045112782, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 27863 + }, + { + "epoch": 1.3752781954887219, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 27864 + }, + { + "epoch": 1.3753383458646615, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 27865 + }, + { + "epoch": 1.3753984962406016, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 27866 + }, + { + "epoch": 1.3754586466165413, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 27867 + }, + { + "epoch": 1.3755187969924811, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 27868 + }, + { + "epoch": 1.375578947368421, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 27869 + }, + { + "epoch": 1.3756390977443609, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 27870 + }, + { + "epoch": 1.3756992481203008, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 27871 + }, + { + "epoch": 1.3757593984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 27872 + }, + { + "epoch": 1.3758195488721805, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 27873 + }, + { + "epoch": 1.3758796992481204, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 27874 + }, + { + "epoch": 1.3759398496240602, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 27875 + }, + { + "epoch": 1.376, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 27876 + }, + { + "epoch": 1.37606015037594, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 27877 + }, + { + "epoch": 1.3761203007518796, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 27878 + }, + { + "epoch": 1.3761804511278195, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 27879 + }, + { + "epoch": 1.3762406015037594, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 27880 + }, + { + "epoch": 1.3763007518796992, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 27881 + }, + { + "epoch": 1.3763609022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 27882 + }, + { + "epoch": 1.376421052631579, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 27883 + }, + { + "epoch": 1.3764812030075189, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.2499, + "step": 27884 + }, + { + "epoch": 1.3765413533834585, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 27885 + }, + { + "epoch": 1.3766015037593986, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 27886 + }, + { + "epoch": 1.3766616541353383, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 27887 + }, + { + "epoch": 1.3767218045112781, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 27888 + }, + { + "epoch": 1.376781954887218, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 27889 + }, + { + "epoch": 1.3768421052631579, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 27890 + }, + { + "epoch": 1.3769022556390977, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 27891 + }, + { + "epoch": 1.3769624060150376, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 27892 + }, + { + "epoch": 1.3770225563909775, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 27893 + }, + { + "epoch": 1.3770827067669174, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 27894 + }, + { + "epoch": 1.3771428571428572, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 27895 + }, + { + "epoch": 1.3772030075187969, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 27896 + }, + { + "epoch": 1.377263157894737, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 27897 + }, + { + "epoch": 1.3773233082706766, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2482, + "step": 27898 + }, + { + "epoch": 1.3773834586466165, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 27899 + }, + { + "epoch": 1.3774436090225564, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 27900 + }, + { + "epoch": 1.3775037593984962, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 27901 + }, + { + "epoch": 1.377563909774436, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 27902 + }, + { + "epoch": 1.377624060150376, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 27903 + }, + { + "epoch": 1.3776842105263158, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 27904 + }, + { + "epoch": 1.3777443609022557, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 27905 + }, + { + "epoch": 1.3778045112781956, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 27906 + }, + { + "epoch": 1.3778646616541352, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 27907 + }, + { + "epoch": 1.377924812030075, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 27908 + }, + { + "epoch": 1.377984962406015, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 27909 + }, + { + "epoch": 1.3780451127819549, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 27910 + }, + { + "epoch": 1.3781052631578947, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 27911 + }, + { + "epoch": 1.3781654135338346, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 27912 + }, + { + "epoch": 1.3782255639097745, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 27913 + }, + { + "epoch": 1.3782857142857143, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 27914 + }, + { + "epoch": 1.3783458646616542, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 27915 + }, + { + "epoch": 1.3784060150375939, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 27916 + }, + { + "epoch": 1.378466165413534, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 27917 + }, + { + "epoch": 1.3785263157894736, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 27918 + }, + { + "epoch": 1.3785864661654135, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 27919 + }, + { + "epoch": 1.3786466165413533, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 27920 + }, + { + "epoch": 1.3787067669172932, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 27921 + }, + { + "epoch": 1.378766917293233, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 27922 + }, + { + "epoch": 1.378827067669173, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 27923 + }, + { + "epoch": 1.3788872180451128, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 27924 + }, + { + "epoch": 1.3789473684210527, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 27925 + }, + { + "epoch": 1.3790075187969926, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 27926 + }, + { + "epoch": 1.3790676691729322, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 27927 + }, + { + "epoch": 1.3791278195488723, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 27928 + }, + { + "epoch": 1.379187969924812, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 27929 + }, + { + "epoch": 1.3792481203007518, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 27930 + }, + { + "epoch": 1.3793082706766917, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 27931 + }, + { + "epoch": 1.3793684210526316, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 27932 + }, + { + "epoch": 1.3794285714285714, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 27933 + }, + { + "epoch": 1.3794887218045113, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 27934 + }, + { + "epoch": 1.3795488721804512, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 27935 + }, + { + "epoch": 1.3796090225563908, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 27936 + }, + { + "epoch": 1.379669172932331, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 27937 + }, + { + "epoch": 1.3797293233082706, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 27938 + }, + { + "epoch": 1.3797894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 27939 + }, + { + "epoch": 1.3798496240601503, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 27940 + }, + { + "epoch": 1.3799097744360902, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 27941 + }, + { + "epoch": 1.37996992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 27942 + }, + { + "epoch": 1.38003007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 27943 + }, + { + "epoch": 1.3800902255639098, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 27944 + }, + { + "epoch": 1.3801503759398497, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 27945 + }, + { + "epoch": 1.3802105263157896, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 27946 + }, + { + "epoch": 1.3802706766917292, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9387, + "step": 27947 + }, + { + "epoch": 1.3803308270676693, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 27948 + }, + { + "epoch": 1.380390977443609, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 27949 + }, + { + "epoch": 1.3804511278195488, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 27950 + }, + { + "epoch": 1.3805112781954887, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 27951 + }, + { + "epoch": 1.3805714285714286, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 27952 + }, + { + "epoch": 1.3806315789473684, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 27953 + }, + { + "epoch": 1.3806917293233083, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 27954 + }, + { + "epoch": 1.3807518796992482, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 27955 + }, + { + "epoch": 1.380812030075188, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 27956 + }, + { + "epoch": 1.380872180451128, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 27957 + }, + { + "epoch": 1.3809323308270676, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 27958 + }, + { + "epoch": 1.3809924812030074, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1646, + "step": 27959 + }, + { + "epoch": 1.3810526315789473, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 27960 + }, + { + "epoch": 1.3811127819548872, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 27961 + }, + { + "epoch": 1.381172932330827, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 27962 + }, + { + "epoch": 1.381233082706767, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 27963 + }, + { + "epoch": 1.3812932330827068, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 27964 + }, + { + "epoch": 1.3813533834586467, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 27965 + }, + { + "epoch": 1.3814135338345865, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 27966 + }, + { + "epoch": 1.3814736842105262, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 27967 + }, + { + "epoch": 1.3815338345864663, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 27968 + }, + { + "epoch": 1.381593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 27969 + }, + { + "epoch": 1.3816541353383458, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 27970 + }, + { + "epoch": 1.3817142857142857, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 27971 + }, + { + "epoch": 1.3817744360902255, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 27972 + }, + { + "epoch": 1.3818345864661654, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 27973 + }, + { + "epoch": 1.3818947368421053, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 27974 + }, + { + "epoch": 1.3819548872180452, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 27975 + }, + { + "epoch": 1.382015037593985, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 27976 + }, + { + "epoch": 1.382075187969925, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 27977 + }, + { + "epoch": 1.3821353383458646, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 27978 + }, + { + "epoch": 1.3821954887218046, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 27979 + }, + { + "epoch": 1.3822556390977443, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 27980 + }, + { + "epoch": 1.3823157894736842, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 27981 + }, + { + "epoch": 1.382375939849624, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 27982 + }, + { + "epoch": 1.382436090225564, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 27983 + }, + { + "epoch": 1.3824962406015038, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 27984 + }, + { + "epoch": 1.3825563909774436, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 27985 + }, + { + "epoch": 1.3826165413533835, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 27986 + }, + { + "epoch": 1.3826766917293232, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 27987 + }, + { + "epoch": 1.3827368421052633, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 27988 + }, + { + "epoch": 1.382796992481203, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.2153, + "step": 27989 + }, + { + "epoch": 1.3828571428571428, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 27990 + }, + { + "epoch": 1.3829172932330827, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 27991 + }, + { + "epoch": 1.3829774436090225, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 27992 + }, + { + "epoch": 1.3830375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 27993 + }, + { + "epoch": 1.3830977443609023, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 27994 + }, + { + "epoch": 1.3831578947368421, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 27995 + }, + { + "epoch": 1.383218045112782, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 27996 + }, + { + "epoch": 1.3832781954887219, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 27997 + }, + { + "epoch": 1.3833383458646615, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 27998 + }, + { + "epoch": 1.3833984962406016, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 27999 + }, + { + "epoch": 1.3834586466165413, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 28000 + }, + { + "epoch": 1.3835187969924811, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 28001 + }, + { + "epoch": 1.383578947368421, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 28002 + }, + { + "epoch": 1.383639097744361, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 28003 + }, + { + "epoch": 1.3836992481203008, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 28004 + }, + { + "epoch": 1.3837593984962406, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 28005 + }, + { + "epoch": 1.3838195488721805, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 28006 + }, + { + "epoch": 1.3838796992481204, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 28007 + }, + { + "epoch": 1.3839398496240602, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 28008 + }, + { + "epoch": 1.384, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 28009 + }, + { + "epoch": 1.3840601503759398, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 28010 + }, + { + "epoch": 1.3841203007518796, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 28011 + }, + { + "epoch": 1.3841804511278195, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 28012 + }, + { + "epoch": 1.3842406015037594, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 28013 + }, + { + "epoch": 1.3843007518796993, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 28014 + }, + { + "epoch": 1.3843609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 28015 + }, + { + "epoch": 1.384421052631579, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 28016 + }, + { + "epoch": 1.3844812030075189, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 28017 + }, + { + "epoch": 1.3845413533834585, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 28018 + }, + { + "epoch": 1.3846015037593986, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 28019 + }, + { + "epoch": 1.3846616541353383, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 28020 + }, + { + "epoch": 1.3847218045112781, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 28021 + }, + { + "epoch": 1.384781954887218, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 28022 + }, + { + "epoch": 1.3848421052631579, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 28023 + }, + { + "epoch": 1.3849022556390977, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 28024 + }, + { + "epoch": 1.3849624060150376, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 28025 + }, + { + "epoch": 1.3850225563909775, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 28026 + }, + { + "epoch": 1.3850827067669174, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 28027 + }, + { + "epoch": 1.3851428571428572, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 28028 + }, + { + "epoch": 1.3852030075187969, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 28029 + }, + { + "epoch": 1.385263157894737, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 28030 + }, + { + "epoch": 1.3853233082706766, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 28031 + }, + { + "epoch": 1.3853834586466165, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 28032 + }, + { + "epoch": 1.3854436090225564, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 28033 + }, + { + "epoch": 1.3855037593984962, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 28034 + }, + { + "epoch": 1.385563909774436, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 28035 + }, + { + "epoch": 1.385624060150376, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 28036 + }, + { + "epoch": 1.3856842105263159, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 28037 + }, + { + "epoch": 1.3857443609022555, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 28038 + }, + { + "epoch": 1.3858045112781956, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 28039 + }, + { + "epoch": 1.3858646616541352, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 28040 + }, + { + "epoch": 1.3859248120300751, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 28041 + }, + { + "epoch": 1.385984962406015, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1494, + "step": 28042 + }, + { + "epoch": 1.3860451127819549, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1685, + "step": 28043 + }, + { + "epoch": 1.3861052631578947, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 28044 + }, + { + "epoch": 1.3861654135338346, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 28045 + }, + { + "epoch": 1.3862255639097745, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 28046 + }, + { + "epoch": 1.3862857142857143, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 28047 + }, + { + "epoch": 1.3863458646616542, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 28048 + }, + { + "epoch": 1.3864060150375939, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 28049 + }, + { + "epoch": 1.386466165413534, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 28050 + }, + { + "epoch": 1.3865263157894736, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 28051 + }, + { + "epoch": 1.3865864661654135, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1974, + "step": 28052 + }, + { + "epoch": 1.3866466165413533, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 28053 + }, + { + "epoch": 1.3867067669172932, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 28054 + }, + { + "epoch": 1.386766917293233, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 28055 + }, + { + "epoch": 1.386827067669173, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 28056 + }, + { + "epoch": 1.3868872180451128, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 28057 + }, + { + "epoch": 1.3869473684210527, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 28058 + }, + { + "epoch": 1.3870075187969926, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 28059 + }, + { + "epoch": 1.3870676691729322, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 28060 + }, + { + "epoch": 1.387127819548872, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 28061 + }, + { + "epoch": 1.387187969924812, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 28062 + }, + { + "epoch": 1.3872481203007518, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 28063 + }, + { + "epoch": 1.3873082706766917, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 28064 + }, + { + "epoch": 1.3873684210526316, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.3851, + "step": 28065 + }, + { + "epoch": 1.3874285714285715, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 28066 + }, + { + "epoch": 1.3874887218045113, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 28067 + }, + { + "epoch": 1.3875488721804512, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 28068 + }, + { + "epoch": 1.3876090225563908, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 28069 + }, + { + "epoch": 1.387669172932331, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 28070 + }, + { + "epoch": 1.3877293233082706, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 28071 + }, + { + "epoch": 1.3877894736842105, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 28072 + }, + { + "epoch": 1.3878496240601503, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 28073 + }, + { + "epoch": 1.3879097744360902, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 28074 + }, + { + "epoch": 1.38796992481203, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1982, + "step": 28075 + }, + { + "epoch": 1.38803007518797, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 28076 + }, + { + "epoch": 1.3880902255639098, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 28077 + }, + { + "epoch": 1.3881503759398497, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 28078 + }, + { + "epoch": 1.3882105263157896, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 28079 + }, + { + "epoch": 1.3882706766917292, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 28080 + }, + { + "epoch": 1.3883308270676693, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 28081 + }, + { + "epoch": 1.388390977443609, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 28082 + }, + { + "epoch": 1.3884511278195488, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 28083 + }, + { + "epoch": 1.3885112781954887, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 28084 + }, + { + "epoch": 1.3885714285714286, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 28085 + }, + { + "epoch": 1.3886315789473684, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 28086 + }, + { + "epoch": 1.3886917293233083, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 28087 + }, + { + "epoch": 1.3887518796992482, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 28088 + }, + { + "epoch": 1.388812030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 28089 + }, + { + "epoch": 1.388872180451128, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 28090 + }, + { + "epoch": 1.3889323308270676, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 28091 + }, + { + "epoch": 1.3889924812030074, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 28092 + }, + { + "epoch": 1.3890526315789473, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 28093 + }, + { + "epoch": 1.3891127819548872, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 28094 + }, + { + "epoch": 1.389172932330827, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 28095 + }, + { + "epoch": 1.389233082706767, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 28096 + }, + { + "epoch": 1.3892932330827068, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 28097 + }, + { + "epoch": 1.3893533834586467, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 28098 + }, + { + "epoch": 1.3894135338345865, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 28099 + }, + { + "epoch": 1.3894736842105262, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 28100 + }, + { + "epoch": 1.3895338345864663, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 28101 + }, + { + "epoch": 1.389593984962406, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 28102 + }, + { + "epoch": 1.3896541353383458, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 28103 + }, + { + "epoch": 1.3897142857142857, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 28104 + }, + { + "epoch": 1.3897744360902256, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 28105 + }, + { + "epoch": 1.3898345864661654, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 28106 + }, + { + "epoch": 1.3898947368421053, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 28107 + }, + { + "epoch": 1.3899548872180452, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 28108 + }, + { + "epoch": 1.390015037593985, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 28109 + }, + { + "epoch": 1.390075187969925, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 28110 + }, + { + "epoch": 1.3901353383458646, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 28111 + }, + { + "epoch": 1.3901954887218044, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 28112 + }, + { + "epoch": 1.3902556390977443, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 28113 + }, + { + "epoch": 1.3903157894736842, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 28114 + }, + { + "epoch": 1.390375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 28115 + }, + { + "epoch": 1.390436090225564, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 28116 + }, + { + "epoch": 1.3904962406015038, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 28117 + }, + { + "epoch": 1.3905563909774437, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 28118 + }, + { + "epoch": 1.3906165413533835, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 28119 + }, + { + "epoch": 1.3906766917293232, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 28120 + }, + { + "epoch": 1.3907368421052633, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 28121 + }, + { + "epoch": 1.390796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 28122 + }, + { + "epoch": 1.3908571428571428, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.2421, + "step": 28123 + }, + { + "epoch": 1.3909172932330827, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 28124 + }, + { + "epoch": 1.3909774436090225, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 28125 + }, + { + "epoch": 1.3910375939849624, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 28126 + }, + { + "epoch": 1.3910977443609023, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 28127 + }, + { + "epoch": 1.3911578947368421, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 28128 + }, + { + "epoch": 1.391218045112782, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 28129 + }, + { + "epoch": 1.391278195488722, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 28130 + }, + { + "epoch": 1.3913383458646615, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 28131 + }, + { + "epoch": 1.3913984962406016, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 28132 + }, + { + "epoch": 1.3914586466165413, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 28133 + }, + { + "epoch": 1.3915187969924812, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 28134 + }, + { + "epoch": 1.391578947368421, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 28135 + }, + { + "epoch": 1.391639097744361, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 28136 + }, + { + "epoch": 1.3916992481203008, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 28137 + }, + { + "epoch": 1.3917593984962406, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 28138 + }, + { + "epoch": 1.3918195488721805, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 28139 + }, + { + "epoch": 1.3918796992481204, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 28140 + }, + { + "epoch": 1.3919398496240603, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 28141 + }, + { + "epoch": 1.392, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 28142 + }, + { + "epoch": 1.3920601503759398, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 28143 + }, + { + "epoch": 1.3921203007518796, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 28144 + }, + { + "epoch": 1.3921804511278195, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 28145 + }, + { + "epoch": 1.3922406015037594, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 28146 + }, + { + "epoch": 1.3923007518796993, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 28147 + }, + { + "epoch": 1.3923609022556391, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 28148 + }, + { + "epoch": 1.392421052631579, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 28149 + }, + { + "epoch": 1.3924812030075189, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 28150 + }, + { + "epoch": 1.3925413533834585, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 28151 + }, + { + "epoch": 1.3926015037593986, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 28152 + }, + { + "epoch": 1.3926616541353383, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 28153 + }, + { + "epoch": 1.3927218045112781, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 28154 + }, + { + "epoch": 1.392781954887218, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 28155 + }, + { + "epoch": 1.3928421052631579, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 28156 + }, + { + "epoch": 1.3929022556390978, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 28157 + }, + { + "epoch": 1.3929624060150376, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 28158 + }, + { + "epoch": 1.3930225563909775, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 28159 + }, + { + "epoch": 1.3930827067669174, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 28160 + }, + { + "epoch": 1.3931428571428572, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 28161 + }, + { + "epoch": 1.3932030075187969, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 28162 + }, + { + "epoch": 1.3932631578947368, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 28163 + }, + { + "epoch": 1.3933233082706766, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 28164 + }, + { + "epoch": 1.3933834586466165, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 28165 + }, + { + "epoch": 1.3934436090225564, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 28166 + }, + { + "epoch": 1.3935037593984962, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 28167 + }, + { + "epoch": 1.3935639097744361, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 28168 + }, + { + "epoch": 1.393624060150376, + "grad_norm": 2.578125, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 28169 + }, + { + "epoch": 1.3936842105263159, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 28170 + }, + { + "epoch": 1.3937443609022555, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 28171 + }, + { + "epoch": 1.3938045112781956, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 28172 + }, + { + "epoch": 1.3938646616541353, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 28173 + }, + { + "epoch": 1.3939248120300751, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.188, + "step": 28174 + }, + { + "epoch": 1.393984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 28175 + }, + { + "epoch": 1.3940451127819549, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 28176 + }, + { + "epoch": 1.3941052631578947, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 28177 + }, + { + "epoch": 1.3941654135338346, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 28178 + }, + { + "epoch": 1.3942255639097745, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 28179 + }, + { + "epoch": 1.3942857142857144, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 28180 + }, + { + "epoch": 1.3943458646616542, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 28181 + }, + { + "epoch": 1.3944060150375939, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 28182 + }, + { + "epoch": 1.394466165413534, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 28183 + }, + { + "epoch": 1.3945263157894736, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 28184 + }, + { + "epoch": 1.3945864661654135, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 28185 + }, + { + "epoch": 1.3946466165413534, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 28186 + }, + { + "epoch": 1.3947067669172932, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 28187 + }, + { + "epoch": 1.394766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 28188 + }, + { + "epoch": 1.394827067669173, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 28189 + }, + { + "epoch": 1.3948872180451128, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 28190 + }, + { + "epoch": 1.3949473684210527, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2608, + "step": 28191 + }, + { + "epoch": 1.3950075187969926, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 28192 + }, + { + "epoch": 1.3950676691729322, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 28193 + }, + { + "epoch": 1.395127819548872, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 28194 + }, + { + "epoch": 1.395187969924812, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 28195 + }, + { + "epoch": 1.3952481203007518, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 28196 + }, + { + "epoch": 1.3953082706766917, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 28197 + }, + { + "epoch": 1.3953684210526316, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 28198 + }, + { + "epoch": 1.3954285714285715, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 28199 + }, + { + "epoch": 1.3954887218045113, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 28200 + }, + { + "epoch": 1.3955488721804512, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 28201 + }, + { + "epoch": 1.3956090225563909, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 28202 + }, + { + "epoch": 1.395669172932331, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 28203 + }, + { + "epoch": 1.3957293233082706, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 28204 + }, + { + "epoch": 1.3957894736842105, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 28205 + }, + { + "epoch": 1.3958496240601503, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 28206 + }, + { + "epoch": 1.3959097744360902, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 28207 + }, + { + "epoch": 1.39596992481203, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 28208 + }, + { + "epoch": 1.39603007518797, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 28209 + }, + { + "epoch": 1.3960902255639098, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 28210 + }, + { + "epoch": 1.3961503759398497, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 28211 + }, + { + "epoch": 1.3962105263157896, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 28212 + }, + { + "epoch": 1.3962706766917292, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 28213 + }, + { + "epoch": 1.3963308270676693, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9497, + "step": 28214 + }, + { + "epoch": 1.396390977443609, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 28215 + }, + { + "epoch": 1.3964511278195488, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 28216 + }, + { + "epoch": 1.3965112781954887, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 28217 + }, + { + "epoch": 1.3965714285714286, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 28218 + }, + { + "epoch": 1.3966315789473684, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 28219 + }, + { + "epoch": 1.3966917293233083, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 28220 + }, + { + "epoch": 1.3967518796992482, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 28221 + }, + { + "epoch": 1.3968120300751878, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 28222 + }, + { + "epoch": 1.396872180451128, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 28223 + }, + { + "epoch": 1.3969323308270676, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 28224 + }, + { + "epoch": 1.3969924812030075, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 28225 + }, + { + "epoch": 1.3970526315789473, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 28226 + }, + { + "epoch": 1.3971127819548872, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 28227 + }, + { + "epoch": 1.397172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 28228 + }, + { + "epoch": 1.397233082706767, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 28229 + }, + { + "epoch": 1.3972932330827068, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 28230 + }, + { + "epoch": 1.3973533834586467, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.3626, + "step": 28231 + }, + { + "epoch": 1.3974135338345866, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 28232 + }, + { + "epoch": 1.3974736842105262, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 28233 + }, + { + "epoch": 1.3975338345864663, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 28234 + }, + { + "epoch": 1.397593984962406, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 28235 + }, + { + "epoch": 1.3976541353383458, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 28236 + }, + { + "epoch": 1.3977142857142857, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 28237 + }, + { + "epoch": 1.3977744360902256, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 28238 + }, + { + "epoch": 1.3978345864661654, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 28239 + }, + { + "epoch": 1.3978947368421053, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 28240 + }, + { + "epoch": 1.3979548872180452, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1878, + "step": 28241 + }, + { + "epoch": 1.398015037593985, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 28242 + }, + { + "epoch": 1.398075187969925, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 28243 + }, + { + "epoch": 1.3981353383458646, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 28244 + }, + { + "epoch": 1.3981954887218044, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 28245 + }, + { + "epoch": 1.3982556390977443, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 28246 + }, + { + "epoch": 1.3983157894736842, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 28247 + }, + { + "epoch": 1.398375939849624, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 28248 + }, + { + "epoch": 1.398436090225564, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 28249 + }, + { + "epoch": 1.3984962406015038, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 28250 + }, + { + "epoch": 1.3985563909774437, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 28251 + }, + { + "epoch": 1.3986165413533835, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 28252 + }, + { + "epoch": 1.3986766917293232, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 28253 + }, + { + "epoch": 1.3987368421052633, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 28254 + }, + { + "epoch": 1.398796992481203, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 28255 + }, + { + "epoch": 1.3988571428571428, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 28256 + }, + { + "epoch": 1.3989172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 28257 + }, + { + "epoch": 1.3989774436090225, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 28258 + }, + { + "epoch": 1.3990375939849624, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 28259 + }, + { + "epoch": 1.3990977443609023, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 28260 + }, + { + "epoch": 1.3991578947368422, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 28261 + }, + { + "epoch": 1.399218045112782, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 28262 + }, + { + "epoch": 1.399278195488722, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 28263 + }, + { + "epoch": 1.3993383458646615, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 28264 + }, + { + "epoch": 1.3993984962406016, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 28265 + }, + { + "epoch": 1.3994586466165413, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 28266 + }, + { + "epoch": 1.3995187969924812, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 28267 + }, + { + "epoch": 1.399578947368421, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1124, + "step": 28268 + }, + { + "epoch": 1.399639097744361, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 28269 + }, + { + "epoch": 1.3996992481203008, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 28270 + }, + { + "epoch": 1.3997593984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 28271 + }, + { + "epoch": 1.3998195488721805, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 28272 + }, + { + "epoch": 1.3998796992481202, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 28273 + }, + { + "epoch": 1.3999398496240603, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 28274 + }, + { + "epoch": 1.4, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 28275 + }, + { + "epoch": 1.4000601503759398, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 28276 + }, + { + "epoch": 1.4001203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 28277 + }, + { + "epoch": 1.4001804511278195, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 28278 + }, + { + "epoch": 1.4002406015037594, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 28279 + }, + { + "epoch": 1.4003007518796993, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 28280 + }, + { + "epoch": 1.4003609022556391, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 28281 + }, + { + "epoch": 1.400421052631579, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 28282 + }, + { + "epoch": 1.4004812030075189, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 28283 + }, + { + "epoch": 1.4005413533834585, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 28284 + }, + { + "epoch": 1.4006015037593986, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 28285 + }, + { + "epoch": 1.4006616541353383, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 28286 + }, + { + "epoch": 1.4007218045112781, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 28287 + }, + { + "epoch": 1.400781954887218, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 28288 + }, + { + "epoch": 1.4008421052631579, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 28289 + }, + { + "epoch": 1.4009022556390978, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 28290 + }, + { + "epoch": 1.4009624060150376, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 28291 + }, + { + "epoch": 1.4010225563909775, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2857, + "step": 28292 + }, + { + "epoch": 1.4010827067669174, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 28293 + }, + { + "epoch": 1.4011428571428572, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2015, + "step": 28294 + }, + { + "epoch": 1.401203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 28295 + }, + { + "epoch": 1.4012631578947368, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 28296 + }, + { + "epoch": 1.4013233082706766, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 28297 + }, + { + "epoch": 1.4013834586466165, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 28298 + }, + { + "epoch": 1.4014436090225564, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 28299 + }, + { + "epoch": 1.4015037593984963, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 28300 + }, + { + "epoch": 1.4015639097744361, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 28301 + }, + { + "epoch": 1.401624060150376, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 28302 + }, + { + "epoch": 1.4016842105263159, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 28303 + }, + { + "epoch": 1.4017443609022555, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 28304 + }, + { + "epoch": 1.4018045112781956, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 28305 + }, + { + "epoch": 1.4018646616541353, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 28306 + }, + { + "epoch": 1.4019248120300751, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 28307 + }, + { + "epoch": 1.401984962406015, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 28308 + }, + { + "epoch": 1.4020451127819549, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 28309 + }, + { + "epoch": 1.4021052631578947, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 28310 + }, + { + "epoch": 1.4021654135338346, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 28311 + }, + { + "epoch": 1.4022255639097745, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 28312 + }, + { + "epoch": 1.4022857142857144, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 28313 + }, + { + "epoch": 1.4023458646616542, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 28314 + }, + { + "epoch": 1.4024060150375939, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 28315 + }, + { + "epoch": 1.402466165413534, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 28316 + }, + { + "epoch": 1.4025263157894736, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 28317 + }, + { + "epoch": 1.4025864661654135, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 28318 + }, + { + "epoch": 1.4026466165413534, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 28319 + }, + { + "epoch": 1.4027067669172932, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 28320 + }, + { + "epoch": 1.402766917293233, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 28321 + }, + { + "epoch": 1.402827067669173, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 28322 + }, + { + "epoch": 1.4028872180451128, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 28323 + }, + { + "epoch": 1.4029473684210525, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 28324 + }, + { + "epoch": 1.4030075187969926, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 28325 + }, + { + "epoch": 1.4030676691729322, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 28326 + }, + { + "epoch": 1.4031278195488721, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 28327 + }, + { + "epoch": 1.403187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 28328 + }, + { + "epoch": 1.4032481203007519, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 28329 + }, + { + "epoch": 1.4033082706766917, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 28330 + }, + { + "epoch": 1.4033684210526316, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 28331 + }, + { + "epoch": 1.4034285714285715, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 28332 + }, + { + "epoch": 1.4034887218045113, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 28333 + }, + { + "epoch": 1.4035488721804512, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 28334 + }, + { + "epoch": 1.4036090225563909, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 28335 + }, + { + "epoch": 1.403669172932331, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 28336 + }, + { + "epoch": 1.4037293233082706, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 28337 + }, + { + "epoch": 1.4037894736842105, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 28338 + }, + { + "epoch": 1.4038496240601503, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 28339 + }, + { + "epoch": 1.4039097744360902, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 28340 + }, + { + "epoch": 1.40396992481203, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 28341 + }, + { + "epoch": 1.40403007518797, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 28342 + }, + { + "epoch": 1.4040902255639098, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 28343 + }, + { + "epoch": 1.4041503759398497, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 28344 + }, + { + "epoch": 1.4042105263157896, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 28345 + }, + { + "epoch": 1.4042706766917292, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 28346 + }, + { + "epoch": 1.404330827067669, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 28347 + }, + { + "epoch": 1.404390977443609, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 28348 + }, + { + "epoch": 1.4044511278195488, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 28349 + }, + { + "epoch": 1.4045112781954887, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 28350 + }, + { + "epoch": 1.4045714285714286, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 28351 + }, + { + "epoch": 1.4046315789473685, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1943, + "step": 28352 + }, + { + "epoch": 1.4046917293233083, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 28353 + }, + { + "epoch": 1.4047518796992482, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 28354 + }, + { + "epoch": 1.4048120300751878, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 28355 + }, + { + "epoch": 1.404872180451128, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 28356 + }, + { + "epoch": 1.4049323308270676, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 28357 + }, + { + "epoch": 1.4049924812030075, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 28358 + }, + { + "epoch": 1.4050526315789473, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 28359 + }, + { + "epoch": 1.4051127819548872, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 28360 + }, + { + "epoch": 1.405172932330827, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 28361 + }, + { + "epoch": 1.405233082706767, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 28362 + }, + { + "epoch": 1.4052932330827068, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 28363 + }, + { + "epoch": 1.4053533834586467, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 28364 + }, + { + "epoch": 1.4054135338345866, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 28365 + }, + { + "epoch": 1.4054736842105262, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 28366 + }, + { + "epoch": 1.4055338345864663, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 28367 + }, + { + "epoch": 1.405593984962406, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 28368 + }, + { + "epoch": 1.4056541353383458, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 28369 + }, + { + "epoch": 1.4057142857142857, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 28370 + }, + { + "epoch": 1.4057744360902256, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 28371 + }, + { + "epoch": 1.4058345864661654, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 28372 + }, + { + "epoch": 1.4058947368421053, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 28373 + }, + { + "epoch": 1.4059548872180452, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 28374 + }, + { + "epoch": 1.4060150375939848, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 28375 + }, + { + "epoch": 1.406075187969925, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1636, + "step": 28376 + }, + { + "epoch": 1.4061353383458646, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.8769, + "step": 28377 + }, + { + "epoch": 1.4061954887218044, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 28378 + }, + { + "epoch": 1.4062556390977443, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 28379 + }, + { + "epoch": 1.4063157894736842, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 28380 + }, + { + "epoch": 1.406375939849624, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 28381 + }, + { + "epoch": 1.406436090225564, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 28382 + }, + { + "epoch": 1.4064962406015038, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 28383 + }, + { + "epoch": 1.4065563909774437, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 28384 + }, + { + "epoch": 1.4066165413533835, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.283, + "step": 28385 + }, + { + "epoch": 1.4066766917293232, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 28386 + }, + { + "epoch": 1.4067368421052633, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 28387 + }, + { + "epoch": 1.406796992481203, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 28388 + }, + { + "epoch": 1.4068571428571428, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 28389 + }, + { + "epoch": 1.4069172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 28390 + }, + { + "epoch": 1.4069774436090225, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 28391 + }, + { + "epoch": 1.4070375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 28392 + }, + { + "epoch": 1.4070977443609023, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 28393 + }, + { + "epoch": 1.4071578947368422, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 28394 + }, + { + "epoch": 1.407218045112782, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 28395 + }, + { + "epoch": 1.407278195488722, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 28396 + }, + { + "epoch": 1.4073383458646616, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 28397 + }, + { + "epoch": 1.4073984962406014, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 28398 + }, + { + "epoch": 1.4074586466165413, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 28399 + }, + { + "epoch": 1.4075187969924812, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 28400 + }, + { + "epoch": 1.407578947368421, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 28401 + }, + { + "epoch": 1.407639097744361, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 28402 + }, + { + "epoch": 1.4076992481203008, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 28403 + }, + { + "epoch": 1.4077593984962407, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 28404 + }, + { + "epoch": 1.4078195488721805, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 28405 + }, + { + "epoch": 1.4078796992481202, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 28406 + }, + { + "epoch": 1.4079398496240603, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 28407 + }, + { + "epoch": 1.408, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 28408 + }, + { + "epoch": 1.4080601503759398, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 28409 + }, + { + "epoch": 1.4081203007518797, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 28410 + }, + { + "epoch": 1.4081804511278195, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 28411 + }, + { + "epoch": 1.4082406015037594, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 28412 + }, + { + "epoch": 1.4083007518796993, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 28413 + }, + { + "epoch": 1.4083609022556391, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 28414 + }, + { + "epoch": 1.408421052631579, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 28415 + }, + { + "epoch": 1.408481203007519, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 28416 + }, + { + "epoch": 1.4085413533834585, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 28417 + }, + { + "epoch": 1.4086015037593986, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 28418 + }, + { + "epoch": 1.4086616541353383, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 28419 + }, + { + "epoch": 1.4087218045112782, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 28420 + }, + { + "epoch": 1.408781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 28421 + }, + { + "epoch": 1.408842105263158, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 28422 + }, + { + "epoch": 1.4089022556390978, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 28423 + }, + { + "epoch": 1.4089624060150376, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1782, + "step": 28424 + }, + { + "epoch": 1.4090225563909775, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 28425 + }, + { + "epoch": 1.4090827067669172, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 28426 + }, + { + "epoch": 1.4091428571428573, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 28427 + }, + { + "epoch": 1.409203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 28428 + }, + { + "epoch": 1.4092631578947368, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 28429 + }, + { + "epoch": 1.4093233082706766, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 28430 + }, + { + "epoch": 1.4093834586466165, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 28431 + }, + { + "epoch": 1.4094436090225564, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 28432 + }, + { + "epoch": 1.4095037593984963, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 28433 + }, + { + "epoch": 1.4095639097744361, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 28434 + }, + { + "epoch": 1.409624060150376, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 28435 + }, + { + "epoch": 1.4096842105263159, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 28436 + }, + { + "epoch": 1.4097443609022555, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 28437 + }, + { + "epoch": 1.4098045112781956, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 28438 + }, + { + "epoch": 1.4098646616541353, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 28439 + }, + { + "epoch": 1.4099248120300751, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 28440 + }, + { + "epoch": 1.409984962406015, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 28441 + }, + { + "epoch": 1.4100451127819549, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 28442 + }, + { + "epoch": 1.4101052631578948, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 28443 + }, + { + "epoch": 1.4101654135338346, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 28444 + }, + { + "epoch": 1.4102255639097745, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 28445 + }, + { + "epoch": 1.4102857142857144, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 28446 + }, + { + "epoch": 1.4103458646616542, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 28447 + }, + { + "epoch": 1.4104060150375939, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 28448 + }, + { + "epoch": 1.4104661654135338, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 28449 + }, + { + "epoch": 1.4105263157894736, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 28450 + }, + { + "epoch": 1.4105864661654135, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 28451 + }, + { + "epoch": 1.4106466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 28452 + }, + { + "epoch": 1.4107067669172932, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1918, + "step": 28453 + }, + { + "epoch": 1.4107669172932331, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 28454 + }, + { + "epoch": 1.410827067669173, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1681, + "step": 28455 + }, + { + "epoch": 1.4108872180451129, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 28456 + }, + { + "epoch": 1.4109473684210525, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 28457 + }, + { + "epoch": 1.4110075187969926, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 28458 + }, + { + "epoch": 1.4110676691729322, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 28459 + }, + { + "epoch": 1.4111278195488721, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 28460 + }, + { + "epoch": 1.411187969924812, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 28461 + }, + { + "epoch": 1.4112481203007519, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 28462 + }, + { + "epoch": 1.4113082706766917, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 28463 + }, + { + "epoch": 1.4113684210526316, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 28464 + }, + { + "epoch": 1.4114285714285715, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 28465 + }, + { + "epoch": 1.4114887218045113, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 28466 + }, + { + "epoch": 1.4115488721804512, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 28467 + }, + { + "epoch": 1.4116090225563909, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 28468 + }, + { + "epoch": 1.411669172932331, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 28469 + }, + { + "epoch": 1.4117293233082706, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 28470 + }, + { + "epoch": 1.4117894736842105, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 28471 + }, + { + "epoch": 1.4118496240601504, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 28472 + }, + { + "epoch": 1.4119097744360902, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 28473 + }, + { + "epoch": 1.41196992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 28474 + }, + { + "epoch": 1.41203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 28475 + }, + { + "epoch": 1.4120902255639098, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 28476 + }, + { + "epoch": 1.4121503759398495, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 28477 + }, + { + "epoch": 1.4122105263157896, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 28478 + }, + { + "epoch": 1.4122706766917292, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 28479 + }, + { + "epoch": 1.412330827067669, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 28480 + }, + { + "epoch": 1.412390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 28481 + }, + { + "epoch": 1.4124511278195488, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 28482 + }, + { + "epoch": 1.4125112781954887, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 28483 + }, + { + "epoch": 1.4125714285714286, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 28484 + }, + { + "epoch": 1.4126315789473685, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 28485 + }, + { + "epoch": 1.4126917293233083, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 28486 + }, + { + "epoch": 1.4127518796992482, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 28487 + }, + { + "epoch": 1.4128120300751879, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 28488 + }, + { + "epoch": 1.412872180451128, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 28489 + }, + { + "epoch": 1.4129323308270676, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 28490 + }, + { + "epoch": 1.4129924812030075, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 28491 + }, + { + "epoch": 1.4130526315789473, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 28492 + }, + { + "epoch": 1.4131127819548872, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 28493 + }, + { + "epoch": 1.413172932330827, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 28494 + }, + { + "epoch": 1.413233082706767, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 28495 + }, + { + "epoch": 1.4132932330827068, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 28496 + }, + { + "epoch": 1.4133533834586467, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 28497 + }, + { + "epoch": 1.4134135338345866, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 28498 + }, + { + "epoch": 1.4134736842105262, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 28499 + }, + { + "epoch": 1.413533834586466, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 28500 + }, + { + "epoch": 1.413593984962406, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 28501 + }, + { + "epoch": 1.4136541353383458, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 28502 + }, + { + "epoch": 1.4137142857142857, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 28503 + }, + { + "epoch": 1.4137744360902256, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 28504 + }, + { + "epoch": 1.4138345864661654, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 28505 + }, + { + "epoch": 1.4138947368421053, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 28506 + }, + { + "epoch": 1.4139548872180452, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 28507 + }, + { + "epoch": 1.4140150375939848, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 28508 + }, + { + "epoch": 1.414075187969925, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 28509 + }, + { + "epoch": 1.4141353383458646, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 28510 + }, + { + "epoch": 1.4141954887218045, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 28511 + }, + { + "epoch": 1.4142556390977443, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 28512 + }, + { + "epoch": 1.4143157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 28513 + }, + { + "epoch": 1.414375939849624, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 28514 + }, + { + "epoch": 1.414436090225564, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 28515 + }, + { + "epoch": 1.4144962406015038, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2086, + "step": 28516 + }, + { + "epoch": 1.4145563909774437, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 28517 + }, + { + "epoch": 1.4146165413533835, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 28518 + }, + { + "epoch": 1.4146766917293232, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 28519 + }, + { + "epoch": 1.4147368421052633, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 28520 + }, + { + "epoch": 1.414796992481203, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 28521 + }, + { + "epoch": 1.4148571428571428, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.2407, + "step": 28522 + }, + { + "epoch": 1.4149172932330827, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 28523 + }, + { + "epoch": 1.4149774436090226, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 28524 + }, + { + "epoch": 1.4150375939849624, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 28525 + }, + { + "epoch": 1.4150977443609023, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 28526 + }, + { + "epoch": 1.4151578947368422, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 28527 + }, + { + "epoch": 1.4152180451127818, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 28528 + }, + { + "epoch": 1.415278195488722, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 28529 + }, + { + "epoch": 1.4153383458646616, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 28530 + }, + { + "epoch": 1.4153984962406014, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 28531 + }, + { + "epoch": 1.4154586466165413, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 28532 + }, + { + "epoch": 1.4155187969924812, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 28533 + }, + { + "epoch": 1.415578947368421, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 28534 + }, + { + "epoch": 1.415639097744361, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2543, + "step": 28535 + }, + { + "epoch": 1.4156992481203008, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 28536 + }, + { + "epoch": 1.4157593984962407, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 28537 + }, + { + "epoch": 1.4158195488721805, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 28538 + }, + { + "epoch": 1.4158796992481202, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 28539 + }, + { + "epoch": 1.4159398496240603, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 28540 + }, + { + "epoch": 1.416, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 28541 + }, + { + "epoch": 1.4160601503759398, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 28542 + }, + { + "epoch": 1.4161203007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 28543 + }, + { + "epoch": 1.4161804511278195, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 28544 + }, + { + "epoch": 1.4162406015037594, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 28545 + }, + { + "epoch": 1.4163007518796993, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 28546 + }, + { + "epoch": 1.4163609022556392, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 28547 + }, + { + "epoch": 1.416421052631579, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1483, + "step": 28548 + }, + { + "epoch": 1.416481203007519, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 28549 + }, + { + "epoch": 1.4165413533834585, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 28550 + }, + { + "epoch": 1.4166015037593984, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 28551 + }, + { + "epoch": 1.4166616541353383, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 28552 + }, + { + "epoch": 1.4167218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 28553 + }, + { + "epoch": 1.416781954887218, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 28554 + }, + { + "epoch": 1.416842105263158, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 28555 + }, + { + "epoch": 1.4169022556390978, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 28556 + }, + { + "epoch": 1.4169624060150376, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 28557 + }, + { + "epoch": 1.4170225563909775, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 28558 + }, + { + "epoch": 1.4170827067669172, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 28559 + }, + { + "epoch": 1.4171428571428573, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 28560 + }, + { + "epoch": 1.417203007518797, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 28561 + }, + { + "epoch": 1.4172631578947368, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 28562 + }, + { + "epoch": 1.4173233082706767, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 28563 + }, + { + "epoch": 1.4173834586466165, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 28564 + }, + { + "epoch": 1.4174436090225564, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 28565 + }, + { + "epoch": 1.4175037593984963, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 28566 + }, + { + "epoch": 1.4175639097744361, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 28567 + }, + { + "epoch": 1.417624060150376, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 28568 + }, + { + "epoch": 1.4176842105263159, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 28569 + }, + { + "epoch": 1.4177443609022555, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 28570 + }, + { + "epoch": 1.4178045112781956, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 28571 + }, + { + "epoch": 1.4178646616541353, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 28572 + }, + { + "epoch": 1.4179248120300751, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 28573 + }, + { + "epoch": 1.417984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 28574 + }, + { + "epoch": 1.4180451127819549, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 28575 + }, + { + "epoch": 1.4181052631578948, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 28576 + }, + { + "epoch": 1.4181654135338346, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 28577 + }, + { + "epoch": 1.4182255639097745, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 28578 + }, + { + "epoch": 1.4182857142857141, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2432, + "step": 28579 + }, + { + "epoch": 1.4183458646616542, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 28580 + }, + { + "epoch": 1.418406015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 28581 + }, + { + "epoch": 1.4184661654135338, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 28582 + }, + { + "epoch": 1.4185263157894736, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 28583 + }, + { + "epoch": 1.4185864661654135, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 28584 + }, + { + "epoch": 1.4186466165413534, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 28585 + }, + { + "epoch": 1.4187067669172932, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 28586 + }, + { + "epoch": 1.4187669172932331, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 28587 + }, + { + "epoch": 1.418827067669173, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 28588 + }, + { + "epoch": 1.4188872180451129, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 28589 + }, + { + "epoch": 1.4189473684210525, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 28590 + }, + { + "epoch": 1.4190075187969926, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 28591 + }, + { + "epoch": 1.4190676691729323, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 28592 + }, + { + "epoch": 1.4191278195488721, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 28593 + }, + { + "epoch": 1.419187969924812, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 28594 + }, + { + "epoch": 1.4192481203007519, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 28595 + }, + { + "epoch": 1.4193082706766917, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 28596 + }, + { + "epoch": 1.4193684210526316, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 28597 + }, + { + "epoch": 1.4194285714285715, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 28598 + }, + { + "epoch": 1.4194887218045114, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 28599 + }, + { + "epoch": 1.4195488721804512, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 28600 + }, + { + "epoch": 1.4196090225563909, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 28601 + }, + { + "epoch": 1.4196691729323307, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 28602 + }, + { + "epoch": 1.4197293233082706, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 28603 + }, + { + "epoch": 1.4197894736842105, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 28604 + }, + { + "epoch": 1.4198496240601504, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 28605 + }, + { + "epoch": 1.4199097744360902, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 28606 + }, + { + "epoch": 1.41996992481203, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 28607 + }, + { + "epoch": 1.42003007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 28608 + }, + { + "epoch": 1.4200902255639098, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 28609 + }, + { + "epoch": 1.4201503759398495, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1678, + "step": 28610 + }, + { + "epoch": 1.4202105263157896, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 28611 + }, + { + "epoch": 1.4202706766917292, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 28612 + }, + { + "epoch": 1.420330827067669, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 28613 + }, + { + "epoch": 1.420390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.2328, + "step": 28614 + }, + { + "epoch": 1.4204511278195489, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 28615 + }, + { + "epoch": 1.4205112781954887, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 28616 + }, + { + "epoch": 1.4205714285714286, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 28617 + }, + { + "epoch": 1.4206315789473685, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 28618 + }, + { + "epoch": 1.4206917293233083, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 28619 + }, + { + "epoch": 1.4207518796992482, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 28620 + }, + { + "epoch": 1.4208120300751879, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 28621 + }, + { + "epoch": 1.420872180451128, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 28622 + }, + { + "epoch": 1.4209323308270676, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 28623 + }, + { + "epoch": 1.4209924812030075, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 28624 + }, + { + "epoch": 1.4210526315789473, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 28625 + }, + { + "epoch": 1.4211127819548872, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 28626 + }, + { + "epoch": 1.421172932330827, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 28627 + }, + { + "epoch": 1.421233082706767, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 28628 + }, + { + "epoch": 1.4212932330827068, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 28629 + }, + { + "epoch": 1.4213533834586465, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 28630 + }, + { + "epoch": 1.4214135338345866, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 28631 + }, + { + "epoch": 1.4214736842105262, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 28632 + }, + { + "epoch": 1.421533834586466, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 28633 + }, + { + "epoch": 1.421593984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 28634 + }, + { + "epoch": 1.4216541353383458, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 28635 + }, + { + "epoch": 1.4217142857142857, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 28636 + }, + { + "epoch": 1.4217744360902256, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 28637 + }, + { + "epoch": 1.4218345864661655, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 28638 + }, + { + "epoch": 1.4218947368421053, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 28639 + }, + { + "epoch": 1.4219548872180452, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 28640 + }, + { + "epoch": 1.4220150375939848, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 28641 + }, + { + "epoch": 1.422075187969925, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1972, + "step": 28642 + }, + { + "epoch": 1.4221353383458646, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 28643 + }, + { + "epoch": 1.4221954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 28644 + }, + { + "epoch": 1.4222556390977443, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2251, + "step": 28645 + }, + { + "epoch": 1.4223157894736842, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 28646 + }, + { + "epoch": 1.422375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 28647 + }, + { + "epoch": 1.422436090225564, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 28648 + }, + { + "epoch": 1.4224962406015038, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 28649 + }, + { + "epoch": 1.4225563909774437, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 28650 + }, + { + "epoch": 1.4226165413533836, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 28651 + }, + { + "epoch": 1.4226766917293232, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 28652 + }, + { + "epoch": 1.422736842105263, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 28653 + }, + { + "epoch": 1.422796992481203, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 28654 + }, + { + "epoch": 1.4228571428571428, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 28655 + }, + { + "epoch": 1.4229172932330827, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 28656 + }, + { + "epoch": 1.4229774436090226, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.224, + "step": 28657 + }, + { + "epoch": 1.4230375939849624, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 28658 + }, + { + "epoch": 1.4230977443609023, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 28659 + }, + { + "epoch": 1.4231578947368422, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 28660 + }, + { + "epoch": 1.4232180451127818, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 28661 + }, + { + "epoch": 1.423278195488722, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 28662 + }, + { + "epoch": 1.4233383458646616, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 28663 + }, + { + "epoch": 1.4233984962406014, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 28664 + }, + { + "epoch": 1.4234586466165413, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 28665 + }, + { + "epoch": 1.4235187969924812, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 28666 + }, + { + "epoch": 1.423578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 28667 + }, + { + "epoch": 1.423639097744361, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 28668 + }, + { + "epoch": 1.4236992481203008, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 28669 + }, + { + "epoch": 1.4237593984962407, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 28670 + }, + { + "epoch": 1.4238195488721805, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2489, + "step": 28671 + }, + { + "epoch": 1.4238796992481202, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 28672 + }, + { + "epoch": 1.4239398496240603, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 28673 + }, + { + "epoch": 1.424, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 28674 + }, + { + "epoch": 1.4240601503759398, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 28675 + }, + { + "epoch": 1.4241203007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 28676 + }, + { + "epoch": 1.4241804511278195, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 28677 + }, + { + "epoch": 1.4242406015037594, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 28678 + }, + { + "epoch": 1.4243007518796993, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 28679 + }, + { + "epoch": 1.4243609022556392, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 28680 + }, + { + "epoch": 1.4244210526315788, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 28681 + }, + { + "epoch": 1.424481203007519, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 28682 + }, + { + "epoch": 1.4245413533834586, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 28683 + }, + { + "epoch": 1.4246015037593984, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 28684 + }, + { + "epoch": 1.4246616541353383, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 28685 + }, + { + "epoch": 1.4247218045112782, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 28686 + }, + { + "epoch": 1.424781954887218, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 28687 + }, + { + "epoch": 1.424842105263158, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 28688 + }, + { + "epoch": 1.4249022556390978, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 28689 + }, + { + "epoch": 1.4249624060150377, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 28690 + }, + { + "epoch": 1.4250225563909775, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 28691 + }, + { + "epoch": 1.4250827067669172, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 28692 + }, + { + "epoch": 1.4251428571428573, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 28693 + }, + { + "epoch": 1.425203007518797, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 28694 + }, + { + "epoch": 1.4252631578947368, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 28695 + }, + { + "epoch": 1.4253233082706767, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 28696 + }, + { + "epoch": 1.4253834586466165, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 28697 + }, + { + "epoch": 1.4254436090225564, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 28698 + }, + { + "epoch": 1.4255037593984963, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 28699 + }, + { + "epoch": 1.4255639097744361, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 28700 + }, + { + "epoch": 1.425624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 28701 + }, + { + "epoch": 1.4256842105263159, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 28702 + }, + { + "epoch": 1.4257443609022555, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1271, + "step": 28703 + }, + { + "epoch": 1.4258045112781954, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 28704 + }, + { + "epoch": 1.4258646616541353, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 28705 + }, + { + "epoch": 1.4259248120300752, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 28706 + }, + { + "epoch": 1.425984962406015, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 28707 + }, + { + "epoch": 1.426045112781955, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 28708 + }, + { + "epoch": 1.4261052631578948, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 28709 + }, + { + "epoch": 1.4261654135338346, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 28710 + }, + { + "epoch": 1.4262255639097745, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 28711 + }, + { + "epoch": 1.4262857142857142, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 28712 + }, + { + "epoch": 1.4263458646616543, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 28713 + }, + { + "epoch": 1.426406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 28714 + }, + { + "epoch": 1.4264661654135338, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 28715 + }, + { + "epoch": 1.4265263157894736, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 28716 + }, + { + "epoch": 1.4265864661654135, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 28717 + }, + { + "epoch": 1.4266466165413534, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 28718 + }, + { + "epoch": 1.4267067669172933, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 28719 + }, + { + "epoch": 1.4267669172932331, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 28720 + }, + { + "epoch": 1.426827067669173, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 28721 + }, + { + "epoch": 1.4268872180451129, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 28722 + }, + { + "epoch": 1.4269473684210525, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 28723 + }, + { + "epoch": 1.4270075187969926, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 28724 + }, + { + "epoch": 1.4270676691729323, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 28725 + }, + { + "epoch": 1.4271278195488721, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2172, + "step": 28726 + }, + { + "epoch": 1.427187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 28727 + }, + { + "epoch": 1.4272481203007519, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 28728 + }, + { + "epoch": 1.4273082706766917, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 28729 + }, + { + "epoch": 1.4273684210526316, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 28730 + }, + { + "epoch": 1.4274285714285715, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 28731 + }, + { + "epoch": 1.4274887218045111, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 28732 + }, + { + "epoch": 1.4275488721804512, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 28733 + }, + { + "epoch": 1.4276090225563909, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1357, + "step": 28734 + }, + { + "epoch": 1.4276691729323308, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 28735 + }, + { + "epoch": 1.4277293233082706, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 28736 + }, + { + "epoch": 1.4277894736842105, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 28737 + }, + { + "epoch": 1.4278496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 28738 + }, + { + "epoch": 1.4279097744360902, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 28739 + }, + { + "epoch": 1.4279699248120301, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 28740 + }, + { + "epoch": 1.42803007518797, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 28741 + }, + { + "epoch": 1.4280902255639099, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 28742 + }, + { + "epoch": 1.4281503759398495, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 28743 + }, + { + "epoch": 1.4282105263157896, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 28744 + }, + { + "epoch": 1.4282706766917292, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 28745 + }, + { + "epoch": 1.4283308270676691, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 28746 + }, + { + "epoch": 1.428390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 28747 + }, + { + "epoch": 1.4284511278195489, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1704, + "step": 28748 + }, + { + "epoch": 1.4285112781954887, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 28749 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.148, + "step": 28750 + }, + { + "epoch": 1.4286315789473685, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1633, + "step": 28751 + }, + { + "epoch": 1.4286917293233083, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 28752 + }, + { + "epoch": 1.4287518796992482, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 28753 + }, + { + "epoch": 1.4288120300751879, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 28754 + }, + { + "epoch": 1.4288721804511277, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 28755 + }, + { + "epoch": 1.4289323308270676, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 28756 + }, + { + "epoch": 1.4289924812030075, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 28757 + }, + { + "epoch": 1.4290526315789474, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 28758 + }, + { + "epoch": 1.4291127819548872, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 28759 + }, + { + "epoch": 1.429172932330827, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 28760 + }, + { + "epoch": 1.429233082706767, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 28761 + }, + { + "epoch": 1.4292932330827068, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 28762 + }, + { + "epoch": 1.4293533834586465, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 28763 + }, + { + "epoch": 1.4294135338345866, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 28764 + }, + { + "epoch": 1.4294736842105262, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 28765 + }, + { + "epoch": 1.429533834586466, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 28766 + }, + { + "epoch": 1.429593984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 28767 + }, + { + "epoch": 1.4296541353383458, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 28768 + }, + { + "epoch": 1.4297142857142857, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 28769 + }, + { + "epoch": 1.4297744360902256, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1841, + "step": 28770 + }, + { + "epoch": 1.4298345864661655, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 28771 + }, + { + "epoch": 1.4298947368421053, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 28772 + }, + { + "epoch": 1.4299548872180452, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 28773 + }, + { + "epoch": 1.4300150375939849, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 28774 + }, + { + "epoch": 1.430075187969925, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 28775 + }, + { + "epoch": 1.4301353383458646, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 28776 + }, + { + "epoch": 1.4301954887218045, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 28777 + }, + { + "epoch": 1.4302556390977443, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 28778 + }, + { + "epoch": 1.4303157894736842, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2745, + "step": 28779 + }, + { + "epoch": 1.430375939849624, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 28780 + }, + { + "epoch": 1.430436090225564, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 28781 + }, + { + "epoch": 1.4304962406015038, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 28782 + }, + { + "epoch": 1.4305563909774435, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 28783 + }, + { + "epoch": 1.4306165413533836, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 28784 + }, + { + "epoch": 1.4306766917293232, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 28785 + }, + { + "epoch": 1.430736842105263, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1817, + "step": 28786 + }, + { + "epoch": 1.430796992481203, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 28787 + }, + { + "epoch": 1.4308571428571428, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 28788 + }, + { + "epoch": 1.4309172932330827, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 28789 + }, + { + "epoch": 1.4309774436090226, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 28790 + }, + { + "epoch": 1.4310375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 28791 + }, + { + "epoch": 1.4310977443609023, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 28792 + }, + { + "epoch": 1.4311578947368422, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 28793 + }, + { + "epoch": 1.4312180451127818, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 28794 + }, + { + "epoch": 1.431278195488722, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 28795 + }, + { + "epoch": 1.4313383458646616, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 28796 + }, + { + "epoch": 1.4313984962406014, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 28797 + }, + { + "epoch": 1.4314586466165413, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 28798 + }, + { + "epoch": 1.4315187969924812, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 28799 + }, + { + "epoch": 1.431578947368421, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 28800 + }, + { + "epoch": 1.431639097744361, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 28801 + }, + { + "epoch": 1.4316992481203008, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 28802 + }, + { + "epoch": 1.4317593984962407, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 28803 + }, + { + "epoch": 1.4318195488721805, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 28804 + }, + { + "epoch": 1.4318796992481202, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 28805 + }, + { + "epoch": 1.43193984962406, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 28806 + }, + { + "epoch": 1.432, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 28807 + }, + { + "epoch": 1.4320601503759398, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 28808 + }, + { + "epoch": 1.4321203007518797, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 28809 + }, + { + "epoch": 1.4321804511278196, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 28810 + }, + { + "epoch": 1.4322406015037594, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 28811 + }, + { + "epoch": 1.4323007518796993, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 28812 + }, + { + "epoch": 1.4323609022556392, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 28813 + }, + { + "epoch": 1.4324210526315788, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 28814 + }, + { + "epoch": 1.432481203007519, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 28815 + }, + { + "epoch": 1.4325413533834586, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 28816 + }, + { + "epoch": 1.4326015037593984, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 28817 + }, + { + "epoch": 1.4326616541353383, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 28818 + }, + { + "epoch": 1.4327218045112782, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.2097, + "step": 28819 + }, + { + "epoch": 1.432781954887218, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 28820 + }, + { + "epoch": 1.432842105263158, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 28821 + }, + { + "epoch": 1.4329022556390978, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 28822 + }, + { + "epoch": 1.4329624060150377, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 28823 + }, + { + "epoch": 1.4330225563909775, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 28824 + }, + { + "epoch": 1.4330827067669172, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 28825 + }, + { + "epoch": 1.4331428571428573, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 28826 + }, + { + "epoch": 1.433203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1609, + "step": 28827 + }, + { + "epoch": 1.4332631578947368, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 28828 + }, + { + "epoch": 1.4333233082706767, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2001, + "step": 28829 + }, + { + "epoch": 1.4333834586466165, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 28830 + }, + { + "epoch": 1.4334436090225564, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 28831 + }, + { + "epoch": 1.4335037593984963, + "grad_norm": 1.546875, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 28832 + }, + { + "epoch": 1.4335639097744362, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 28833 + }, + { + "epoch": 1.433624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 28834 + }, + { + "epoch": 1.433684210526316, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 28835 + }, + { + "epoch": 1.4337443609022555, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 28836 + }, + { + "epoch": 1.4338045112781954, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 28837 + }, + { + "epoch": 1.4338646616541353, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 28838 + }, + { + "epoch": 1.4339248120300752, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2598, + "step": 28839 + }, + { + "epoch": 1.433984962406015, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2336, + "step": 28840 + }, + { + "epoch": 1.434045112781955, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 28841 + }, + { + "epoch": 1.4341052631578948, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 28842 + }, + { + "epoch": 1.4341654135338346, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 28843 + }, + { + "epoch": 1.4342255639097745, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 28844 + }, + { + "epoch": 1.4342857142857142, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 28845 + }, + { + "epoch": 1.4343458646616543, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 28846 + }, + { + "epoch": 1.434406015037594, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 28847 + }, + { + "epoch": 1.4344661654135338, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 28848 + }, + { + "epoch": 1.4345263157894736, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 28849 + }, + { + "epoch": 1.4345864661654135, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 28850 + }, + { + "epoch": 1.4346466165413534, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 28851 + }, + { + "epoch": 1.4347067669172933, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 28852 + }, + { + "epoch": 1.4347669172932331, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 28853 + }, + { + "epoch": 1.434827067669173, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 28854 + }, + { + "epoch": 1.4348872180451129, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 28855 + }, + { + "epoch": 1.4349473684210525, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 28856 + }, + { + "epoch": 1.4350075187969924, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 28857 + }, + { + "epoch": 1.4350676691729323, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 28858 + }, + { + "epoch": 1.4351278195488721, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 28859 + }, + { + "epoch": 1.435187969924812, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 28860 + }, + { + "epoch": 1.4352481203007519, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 28861 + }, + { + "epoch": 1.4353082706766918, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 28862 + }, + { + "epoch": 1.4353684210526316, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 28863 + }, + { + "epoch": 1.4354285714285715, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 28864 + }, + { + "epoch": 1.4354887218045111, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 28865 + }, + { + "epoch": 1.4355488721804512, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 28866 + }, + { + "epoch": 1.435609022556391, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 28867 + }, + { + "epoch": 1.4356691729323308, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 28868 + }, + { + "epoch": 1.4357293233082706, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 28869 + }, + { + "epoch": 1.4357894736842105, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 28870 + }, + { + "epoch": 1.4358496240601504, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 28871 + }, + { + "epoch": 1.4359097744360902, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 28872 + }, + { + "epoch": 1.4359699248120301, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 28873 + }, + { + "epoch": 1.43603007518797, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 28874 + }, + { + "epoch": 1.4360902255639099, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1988, + "step": 28875 + }, + { + "epoch": 1.4361503759398495, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 28876 + }, + { + "epoch": 1.4362105263157896, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 28877 + }, + { + "epoch": 1.4362706766917293, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 28878 + }, + { + "epoch": 1.4363308270676691, + "grad_norm": 1.3671875, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 28879 + }, + { + "epoch": 1.436390977443609, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 28880 + }, + { + "epoch": 1.4364511278195489, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 28881 + }, + { + "epoch": 1.4365112781954887, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 28882 + }, + { + "epoch": 1.4365714285714286, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 28883 + }, + { + "epoch": 1.4366315789473685, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 28884 + }, + { + "epoch": 1.4366917293233084, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 28885 + }, + { + "epoch": 1.4367518796992482, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 28886 + }, + { + "epoch": 1.4368120300751879, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 28887 + }, + { + "epoch": 1.4368721804511277, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 28888 + }, + { + "epoch": 1.4369323308270676, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 28889 + }, + { + "epoch": 1.4369924812030075, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 28890 + }, + { + "epoch": 1.4370526315789474, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 28891 + }, + { + "epoch": 1.4371127819548872, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 28892 + }, + { + "epoch": 1.437172932330827, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 28893 + }, + { + "epoch": 1.437233082706767, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 28894 + }, + { + "epoch": 1.4372932330827068, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 28895 + }, + { + "epoch": 1.4373533834586465, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 28896 + }, + { + "epoch": 1.4374135338345866, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 28897 + }, + { + "epoch": 1.4374736842105262, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 28898 + }, + { + "epoch": 1.437533834586466, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 28899 + }, + { + "epoch": 1.437593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 28900 + }, + { + "epoch": 1.4376541353383459, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 28901 + }, + { + "epoch": 1.4377142857142857, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 28902 + }, + { + "epoch": 1.4377744360902256, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 28903 + }, + { + "epoch": 1.4378345864661655, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 28904 + }, + { + "epoch": 1.4378947368421053, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 28905 + }, + { + "epoch": 1.4379548872180452, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 28906 + }, + { + "epoch": 1.4380150375939849, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 28907 + }, + { + "epoch": 1.438075187969925, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 28908 + }, + { + "epoch": 1.4381353383458646, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 28909 + }, + { + "epoch": 1.4381954887218045, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 28910 + }, + { + "epoch": 1.4382556390977443, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 28911 + }, + { + "epoch": 1.4383157894736842, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 28912 + }, + { + "epoch": 1.438375939849624, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 28913 + }, + { + "epoch": 1.438436090225564, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 28914 + }, + { + "epoch": 1.4384962406015038, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 28915 + }, + { + "epoch": 1.4385563909774435, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 28916 + }, + { + "epoch": 1.4386165413533836, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 28917 + }, + { + "epoch": 1.4386766917293232, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 28918 + }, + { + "epoch": 1.438736842105263, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 28919 + }, + { + "epoch": 1.438796992481203, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 28920 + }, + { + "epoch": 1.4388571428571428, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 28921 + }, + { + "epoch": 1.4389172932330827, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 28922 + }, + { + "epoch": 1.4389774436090226, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 28923 + }, + { + "epoch": 1.4390375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 28924 + }, + { + "epoch": 1.4390977443609023, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 28925 + }, + { + "epoch": 1.4391578947368422, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 28926 + }, + { + "epoch": 1.4392180451127818, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 28927 + }, + { + "epoch": 1.439278195488722, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 28928 + }, + { + "epoch": 1.4393383458646616, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 28929 + }, + { + "epoch": 1.4393984962406015, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 28930 + }, + { + "epoch": 1.4394586466165413, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 28931 + }, + { + "epoch": 1.4395187969924812, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 28932 + }, + { + "epoch": 1.439578947368421, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 28933 + }, + { + "epoch": 1.439639097744361, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 28934 + }, + { + "epoch": 1.4396992481203008, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 28935 + }, + { + "epoch": 1.4397593984962407, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 28936 + }, + { + "epoch": 1.4398195488721806, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 28937 + }, + { + "epoch": 1.4398796992481202, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 28938 + }, + { + "epoch": 1.43993984962406, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 28939 + }, + { + "epoch": 1.44, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 28940 + }, + { + "epoch": 1.4400601503759398, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 28941 + }, + { + "epoch": 1.4401203007518797, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 28942 + }, + { + "epoch": 1.4401804511278196, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2072, + "step": 28943 + }, + { + "epoch": 1.4402406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 28944 + }, + { + "epoch": 1.4403007518796993, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 28945 + }, + { + "epoch": 1.4403609022556392, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 28946 + }, + { + "epoch": 1.4404210526315788, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 28947 + }, + { + "epoch": 1.440481203007519, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 28948 + }, + { + "epoch": 1.4405413533834586, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 28949 + }, + { + "epoch": 1.4406015037593984, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 28950 + }, + { + "epoch": 1.4406616541353383, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 28951 + }, + { + "epoch": 1.4407218045112782, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 28952 + }, + { + "epoch": 1.440781954887218, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 28953 + }, + { + "epoch": 1.440842105263158, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 28954 + }, + { + "epoch": 1.4409022556390978, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 28955 + }, + { + "epoch": 1.4409624060150377, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 28956 + }, + { + "epoch": 1.4410225563909775, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 28957 + }, + { + "epoch": 1.4410827067669172, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 28958 + }, + { + "epoch": 1.4411428571428573, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 28959 + }, + { + "epoch": 1.441203007518797, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 28960 + }, + { + "epoch": 1.4412631578947368, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.3595, + "step": 28961 + }, + { + "epoch": 1.4413233082706767, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 28962 + }, + { + "epoch": 1.4413834586466165, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 28963 + }, + { + "epoch": 1.4414436090225564, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 28964 + }, + { + "epoch": 1.4415037593984963, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 28965 + }, + { + "epoch": 1.4415639097744362, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 28966 + }, + { + "epoch": 1.4416240601503758, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 28967 + }, + { + "epoch": 1.441684210526316, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 28968 + }, + { + "epoch": 1.4417443609022556, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 28969 + }, + { + "epoch": 1.4418045112781954, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 28970 + }, + { + "epoch": 1.4418646616541353, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 28971 + }, + { + "epoch": 1.4419248120300752, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 28972 + }, + { + "epoch": 1.441984962406015, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 28973 + }, + { + "epoch": 1.442045112781955, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 28974 + }, + { + "epoch": 1.4421052631578948, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9189, + "step": 28975 + }, + { + "epoch": 1.4421654135338347, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 28976 + }, + { + "epoch": 1.4422255639097745, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 28977 + }, + { + "epoch": 1.4422857142857142, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 28978 + }, + { + "epoch": 1.4423458646616543, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 28979 + }, + { + "epoch": 1.442406015037594, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 28980 + }, + { + "epoch": 1.4424661654135338, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 28981 + }, + { + "epoch": 1.4425263157894737, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2221, + "step": 28982 + }, + { + "epoch": 1.4425864661654135, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 28983 + }, + { + "epoch": 1.4426466165413534, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 28984 + }, + { + "epoch": 1.4427067669172933, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.917, + "step": 28985 + }, + { + "epoch": 1.4427669172932331, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 28986 + }, + { + "epoch": 1.442827067669173, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 28987 + }, + { + "epoch": 1.4428872180451129, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 28988 + }, + { + "epoch": 1.4429473684210525, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2061, + "step": 28989 + }, + { + "epoch": 1.4430075187969924, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 28990 + }, + { + "epoch": 1.4430676691729323, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 28991 + }, + { + "epoch": 1.4431278195488721, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 28992 + }, + { + "epoch": 1.443187969924812, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 28993 + }, + { + "epoch": 1.443248120300752, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 28994 + }, + { + "epoch": 1.4433082706766918, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 28995 + }, + { + "epoch": 1.4433684210526316, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 28996 + }, + { + "epoch": 1.4434285714285715, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 28997 + }, + { + "epoch": 1.4434887218045112, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 28998 + }, + { + "epoch": 1.4435488721804512, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 28999 + }, + { + "epoch": 1.443609022556391, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 29000 + }, + { + "epoch": 1.4436691729323308, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 29001 + }, + { + "epoch": 1.4437293233082706, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 29002 + }, + { + "epoch": 1.4437894736842105, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 29003 + }, + { + "epoch": 1.4438496240601504, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 29004 + }, + { + "epoch": 1.4439097744360903, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 29005 + }, + { + "epoch": 1.4439699248120301, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 29006 + }, + { + "epoch": 1.44403007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 29007 + }, + { + "epoch": 1.4440902255639099, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 29008 + }, + { + "epoch": 1.4441503759398495, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 29009 + }, + { + "epoch": 1.4442105263157896, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 29010 + }, + { + "epoch": 1.4442706766917293, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 29011 + }, + { + "epoch": 1.4443308270676691, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 29012 + }, + { + "epoch": 1.444390977443609, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 29013 + }, + { + "epoch": 1.4444511278195489, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 29014 + }, + { + "epoch": 1.4445112781954887, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 29015 + }, + { + "epoch": 1.4445714285714286, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 29016 + }, + { + "epoch": 1.4446315789473685, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 29017 + }, + { + "epoch": 1.4446917293233081, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 29018 + }, + { + "epoch": 1.4447518796992482, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 29019 + }, + { + "epoch": 1.4448120300751879, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 29020 + }, + { + "epoch": 1.4448721804511278, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 29021 + }, + { + "epoch": 1.4449323308270676, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 29022 + }, + { + "epoch": 1.4449924812030075, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 29023 + }, + { + "epoch": 1.4450526315789474, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 29024 + }, + { + "epoch": 1.4451127819548872, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 29025 + }, + { + "epoch": 1.445172932330827, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 29026 + }, + { + "epoch": 1.445233082706767, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 29027 + }, + { + "epoch": 1.4452932330827069, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 29028 + }, + { + "epoch": 1.4453533834586465, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 29029 + }, + { + "epoch": 1.4454135338345866, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 29030 + }, + { + "epoch": 1.4454736842105262, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 29031 + }, + { + "epoch": 1.4455338345864661, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 29032 + }, + { + "epoch": 1.445593984962406, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 29033 + }, + { + "epoch": 1.4456541353383459, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 29034 + }, + { + "epoch": 1.4457142857142857, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 29035 + }, + { + "epoch": 1.4457744360902256, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 29036 + }, + { + "epoch": 1.4458345864661655, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 29037 + }, + { + "epoch": 1.4458947368421053, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 29038 + }, + { + "epoch": 1.4459548872180452, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1512, + "step": 29039 + }, + { + "epoch": 1.4460150375939849, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 29040 + }, + { + "epoch": 1.4460751879699247, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 29041 + }, + { + "epoch": 1.4461353383458646, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 29042 + }, + { + "epoch": 1.4461954887218045, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 29043 + }, + { + "epoch": 1.4462556390977443, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 29044 + }, + { + "epoch": 1.4463157894736842, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 29045 + }, + { + "epoch": 1.446375939849624, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 29046 + }, + { + "epoch": 1.446436090225564, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 29047 + }, + { + "epoch": 1.4464962406015038, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 29048 + }, + { + "epoch": 1.4465563909774435, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 29049 + }, + { + "epoch": 1.4466165413533836, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 29050 + }, + { + "epoch": 1.4466766917293232, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 29051 + }, + { + "epoch": 1.446736842105263, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 29052 + }, + { + "epoch": 1.446796992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 29053 + }, + { + "epoch": 1.4468571428571428, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 29054 + }, + { + "epoch": 1.4469172932330827, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 29055 + }, + { + "epoch": 1.4469774436090226, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 29056 + }, + { + "epoch": 1.4470375939849625, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.926, + "step": 29057 + }, + { + "epoch": 1.4470977443609023, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 29058 + }, + { + "epoch": 1.4471578947368422, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 29059 + }, + { + "epoch": 1.4472180451127818, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 29060 + }, + { + "epoch": 1.447278195488722, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 29061 + }, + { + "epoch": 1.4473383458646616, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1927, + "step": 29062 + }, + { + "epoch": 1.4473984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 29063 + }, + { + "epoch": 1.4474586466165413, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 29064 + }, + { + "epoch": 1.4475187969924812, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 29065 + }, + { + "epoch": 1.447578947368421, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 29066 + }, + { + "epoch": 1.447639097744361, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 29067 + }, + { + "epoch": 1.4476992481203008, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 29068 + }, + { + "epoch": 1.4477593984962405, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 29069 + }, + { + "epoch": 1.4478195488721806, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 29070 + }, + { + "epoch": 1.4478796992481202, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 29071 + }, + { + "epoch": 1.44793984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 29072 + }, + { + "epoch": 1.448, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 29073 + }, + { + "epoch": 1.4480601503759398, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 29074 + }, + { + "epoch": 1.4481203007518797, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 29075 + }, + { + "epoch": 1.4481804511278196, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 29076 + }, + { + "epoch": 1.4482406015037594, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 29077 + }, + { + "epoch": 1.4483007518796993, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 29078 + }, + { + "epoch": 1.4483609022556392, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 29079 + }, + { + "epoch": 1.4484210526315788, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 29080 + }, + { + "epoch": 1.448481203007519, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 29081 + }, + { + "epoch": 1.4485413533834586, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 29082 + }, + { + "epoch": 1.4486015037593984, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 29083 + }, + { + "epoch": 1.4486616541353383, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 29084 + }, + { + "epoch": 1.4487218045112782, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 29085 + }, + { + "epoch": 1.448781954887218, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 29086 + }, + { + "epoch": 1.448842105263158, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 29087 + }, + { + "epoch": 1.4489022556390978, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 29088 + }, + { + "epoch": 1.4489624060150377, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9393, + "step": 29089 + }, + { + "epoch": 1.4490225563909775, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 29090 + }, + { + "epoch": 1.4490827067669172, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 29091 + }, + { + "epoch": 1.449142857142857, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 29092 + }, + { + "epoch": 1.449203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 29093 + }, + { + "epoch": 1.4492631578947368, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 29094 + }, + { + "epoch": 1.4493233082706767, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 29095 + }, + { + "epoch": 1.4493834586466166, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 29096 + }, + { + "epoch": 1.4494436090225564, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 29097 + }, + { + "epoch": 1.4495037593984963, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 29098 + }, + { + "epoch": 1.4495639097744362, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 29099 + }, + { + "epoch": 1.4496240601503758, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 29100 + }, + { + "epoch": 1.449684210526316, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 29101 + }, + { + "epoch": 1.4497443609022556, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 29102 + }, + { + "epoch": 1.4498045112781954, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 29103 + }, + { + "epoch": 1.4498646616541353, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 29104 + }, + { + "epoch": 1.4499248120300752, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 29105 + }, + { + "epoch": 1.449984962406015, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 29106 + }, + { + "epoch": 1.450045112781955, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 29107 + }, + { + "epoch": 1.4501052631578948, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1339, + "step": 29108 + }, + { + "epoch": 1.4501654135338347, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 29109 + }, + { + "epoch": 1.4502255639097745, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 29110 + }, + { + "epoch": 1.4502857142857142, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 29111 + }, + { + "epoch": 1.4503458646616543, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 29112 + }, + { + "epoch": 1.450406015037594, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 29113 + }, + { + "epoch": 1.4504661654135338, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 29114 + }, + { + "epoch": 1.4505263157894737, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 29115 + }, + { + "epoch": 1.4505864661654135, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 29116 + }, + { + "epoch": 1.4506466165413534, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 29117 + }, + { + "epoch": 1.4507067669172933, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 29118 + }, + { + "epoch": 1.4507669172932331, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9316, + "step": 29119 + }, + { + "epoch": 1.4508270676691728, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 29120 + }, + { + "epoch": 1.450887218045113, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 29121 + }, + { + "epoch": 1.4509473684210525, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 29122 + }, + { + "epoch": 1.4510075187969924, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 29123 + }, + { + "epoch": 1.4510676691729323, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 29124 + }, + { + "epoch": 1.4511278195488722, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2685, + "step": 29125 + }, + { + "epoch": 1.451187969924812, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 29126 + }, + { + "epoch": 1.451248120300752, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 29127 + }, + { + "epoch": 1.4513082706766918, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 29128 + }, + { + "epoch": 1.4513684210526316, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 29129 + }, + { + "epoch": 1.4514285714285715, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 29130 + }, + { + "epoch": 1.4514887218045112, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 29131 + }, + { + "epoch": 1.4515488721804513, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 29132 + }, + { + "epoch": 1.451609022556391, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 29133 + }, + { + "epoch": 1.4516691729323308, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 29134 + }, + { + "epoch": 1.4517293233082706, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 29135 + }, + { + "epoch": 1.4517894736842105, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 29136 + }, + { + "epoch": 1.4518496240601504, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 29137 + }, + { + "epoch": 1.4519097744360903, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 29138 + }, + { + "epoch": 1.4519699248120301, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 29139 + }, + { + "epoch": 1.45203007518797, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 29140 + }, + { + "epoch": 1.4520902255639099, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 29141 + }, + { + "epoch": 1.4521503759398495, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 29142 + }, + { + "epoch": 1.4522105263157894, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 29143 + }, + { + "epoch": 1.4522706766917293, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 29144 + }, + { + "epoch": 1.4523308270676691, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 29145 + }, + { + "epoch": 1.452390977443609, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 29146 + }, + { + "epoch": 1.4524511278195489, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 29147 + }, + { + "epoch": 1.4525112781954888, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 29148 + }, + { + "epoch": 1.4525714285714286, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 29149 + }, + { + "epoch": 1.4526315789473685, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 29150 + }, + { + "epoch": 1.4526917293233081, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 29151 + }, + { + "epoch": 1.4527518796992482, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 29152 + }, + { + "epoch": 1.4528120300751879, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 29153 + }, + { + "epoch": 1.4528721804511278, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 29154 + }, + { + "epoch": 1.4529323308270676, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 29155 + }, + { + "epoch": 1.4529924812030075, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 29156 + }, + { + "epoch": 1.4530526315789474, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 29157 + }, + { + "epoch": 1.4531127819548872, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 29158 + }, + { + "epoch": 1.4531729323308271, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 29159 + }, + { + "epoch": 1.453233082706767, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 29160 + }, + { + "epoch": 1.4532932330827069, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 29161 + }, + { + "epoch": 1.4533533834586465, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 29162 + }, + { + "epoch": 1.4534135338345866, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 29163 + }, + { + "epoch": 1.4534736842105263, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 29164 + }, + { + "epoch": 1.4535338345864661, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 29165 + }, + { + "epoch": 1.453593984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 29166 + }, + { + "epoch": 1.4536541353383459, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 29167 + }, + { + "epoch": 1.4537142857142857, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 29168 + }, + { + "epoch": 1.4537744360902256, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 29169 + }, + { + "epoch": 1.4538345864661655, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 29170 + }, + { + "epoch": 1.4538947368421051, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 29171 + }, + { + "epoch": 1.4539548872180452, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 29172 + }, + { + "epoch": 1.4540150375939849, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 29173 + }, + { + "epoch": 1.4540751879699247, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1512, + "step": 29174 + }, + { + "epoch": 1.4541353383458646, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 29175 + }, + { + "epoch": 1.4541954887218045, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 29176 + }, + { + "epoch": 1.4542556390977444, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 29177 + }, + { + "epoch": 1.4543157894736842, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 29178 + }, + { + "epoch": 1.454375939849624, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 29179 + }, + { + "epoch": 1.454436090225564, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 29180 + }, + { + "epoch": 1.4544962406015038, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 29181 + }, + { + "epoch": 1.4545563909774435, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 29182 + }, + { + "epoch": 1.4546165413533836, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 29183 + }, + { + "epoch": 1.4546766917293232, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 29184 + }, + { + "epoch": 1.454736842105263, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 29185 + }, + { + "epoch": 1.454796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 29186 + }, + { + "epoch": 1.4548571428571428, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 29187 + }, + { + "epoch": 1.4549172932330827, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 29188 + }, + { + "epoch": 1.4549774436090226, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 29189 + }, + { + "epoch": 1.4550375939849625, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 29190 + }, + { + "epoch": 1.4550977443609023, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 29191 + }, + { + "epoch": 1.4551578947368422, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 29192 + }, + { + "epoch": 1.4552180451127819, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 29193 + }, + { + "epoch": 1.4552781954887217, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 29194 + }, + { + "epoch": 1.4553383458646616, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 29195 + }, + { + "epoch": 1.4553984962406015, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 29196 + }, + { + "epoch": 1.4554586466165413, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 29197 + }, + { + "epoch": 1.4555187969924812, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 29198 + }, + { + "epoch": 1.455578947368421, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 29199 + }, + { + "epoch": 1.455639097744361, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.166, + "step": 29200 + }, + { + "epoch": 1.4556992481203008, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 29201 + }, + { + "epoch": 1.4557593984962405, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 29202 + }, + { + "epoch": 1.4558195488721806, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 29203 + }, + { + "epoch": 1.4558796992481202, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 29204 + }, + { + "epoch": 1.45593984962406, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 29205 + }, + { + "epoch": 1.456, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 29206 + }, + { + "epoch": 1.4560601503759398, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 29207 + }, + { + "epoch": 1.4561203007518797, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 29208 + }, + { + "epoch": 1.4561804511278196, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 29209 + }, + { + "epoch": 1.4562406015037594, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 29210 + }, + { + "epoch": 1.4563007518796993, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 29211 + }, + { + "epoch": 1.4563609022556392, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 29212 + }, + { + "epoch": 1.4564210526315788, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 29213 + }, + { + "epoch": 1.456481203007519, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 29214 + }, + { + "epoch": 1.4565413533834586, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 29215 + }, + { + "epoch": 1.4566015037593985, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 29216 + }, + { + "epoch": 1.4566616541353383, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 29217 + }, + { + "epoch": 1.4567218045112782, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 29218 + }, + { + "epoch": 1.456781954887218, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 29219 + }, + { + "epoch": 1.456842105263158, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 29220 + }, + { + "epoch": 1.4569022556390978, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 29221 + }, + { + "epoch": 1.4569624060150375, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 29222 + }, + { + "epoch": 1.4570225563909776, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 29223 + }, + { + "epoch": 1.4570827067669172, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 29224 + }, + { + "epoch": 1.457142857142857, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 29225 + }, + { + "epoch": 1.457203007518797, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 29226 + }, + { + "epoch": 1.4572631578947368, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 29227 + }, + { + "epoch": 1.4573233082706767, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 29228 + }, + { + "epoch": 1.4573834586466166, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 29229 + }, + { + "epoch": 1.4574436090225564, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 29230 + }, + { + "epoch": 1.4575037593984963, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 29231 + }, + { + "epoch": 1.4575639097744362, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 29232 + }, + { + "epoch": 1.4576240601503758, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 29233 + }, + { + "epoch": 1.457684210526316, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 29234 + }, + { + "epoch": 1.4577443609022556, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 29235 + }, + { + "epoch": 1.4578045112781954, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 29236 + }, + { + "epoch": 1.4578646616541353, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 29237 + }, + { + "epoch": 1.4579248120300752, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 29238 + }, + { + "epoch": 1.457984962406015, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 29239 + }, + { + "epoch": 1.458045112781955, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 29240 + }, + { + "epoch": 1.4581052631578948, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 29241 + }, + { + "epoch": 1.4581654135338347, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 29242 + }, + { + "epoch": 1.4582255639097745, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 29243 + }, + { + "epoch": 1.4582857142857142, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 29244 + }, + { + "epoch": 1.458345864661654, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 29245 + }, + { + "epoch": 1.458406015037594, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 29246 + }, + { + "epoch": 1.4584661654135338, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 29247 + }, + { + "epoch": 1.4585263157894737, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 29248 + }, + { + "epoch": 1.4585864661654135, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 29249 + }, + { + "epoch": 1.4586466165413534, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 29250 + }, + { + "epoch": 1.4587067669172933, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 29251 + }, + { + "epoch": 1.4587669172932332, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 29252 + }, + { + "epoch": 1.4588270676691728, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 29253 + }, + { + "epoch": 1.458887218045113, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 29254 + }, + { + "epoch": 1.4589473684210525, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 29255 + }, + { + "epoch": 1.4590075187969924, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1724, + "step": 29256 + }, + { + "epoch": 1.4590676691729323, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 29257 + }, + { + "epoch": 1.4591278195488722, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 29258 + }, + { + "epoch": 1.459187969924812, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 29259 + }, + { + "epoch": 1.459248120300752, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 29260 + }, + { + "epoch": 1.4593082706766918, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 29261 + }, + { + "epoch": 1.4593684210526316, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 29262 + }, + { + "epoch": 1.4594285714285715, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 29263 + }, + { + "epoch": 1.4594887218045112, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 29264 + }, + { + "epoch": 1.4595488721804513, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 29265 + }, + { + "epoch": 1.459609022556391, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 29266 + }, + { + "epoch": 1.4596691729323308, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 29267 + }, + { + "epoch": 1.4597293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 29268 + }, + { + "epoch": 1.4597894736842105, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 29269 + }, + { + "epoch": 1.4598496240601504, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 29270 + }, + { + "epoch": 1.4599097744360903, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 29271 + }, + { + "epoch": 1.4599699248120301, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 29272 + }, + { + "epoch": 1.4600300751879698, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 29273 + }, + { + "epoch": 1.4600902255639099, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 29274 + }, + { + "epoch": 1.4601503759398495, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 29275 + }, + { + "epoch": 1.4602105263157894, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 29276 + }, + { + "epoch": 1.4602706766917293, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 29277 + }, + { + "epoch": 1.4603308270676691, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 29278 + }, + { + "epoch": 1.460390977443609, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 29279 + }, + { + "epoch": 1.4604511278195489, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 29280 + }, + { + "epoch": 1.4605112781954888, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 29281 + }, + { + "epoch": 1.4605714285714286, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 29282 + }, + { + "epoch": 1.4606315789473685, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 29283 + }, + { + "epoch": 1.4606917293233082, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 29284 + }, + { + "epoch": 1.4607518796992482, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 29285 + }, + { + "epoch": 1.460812030075188, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 29286 + }, + { + "epoch": 1.4608721804511278, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 29287 + }, + { + "epoch": 1.4609323308270676, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 29288 + }, + { + "epoch": 1.4609924812030075, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 29289 + }, + { + "epoch": 1.4610526315789474, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1438, + "step": 29290 + }, + { + "epoch": 1.4611127819548873, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 29291 + }, + { + "epoch": 1.4611729323308271, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 29292 + }, + { + "epoch": 1.461233082706767, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 29293 + }, + { + "epoch": 1.4612932330827069, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 29294 + }, + { + "epoch": 1.4613533834586465, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 29295 + }, + { + "epoch": 1.4614135338345864, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 29296 + }, + { + "epoch": 1.4614736842105263, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 29297 + }, + { + "epoch": 1.4615338345864661, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 29298 + }, + { + "epoch": 1.461593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 29299 + }, + { + "epoch": 1.4616541353383459, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 29300 + }, + { + "epoch": 1.4617142857142857, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 29301 + }, + { + "epoch": 1.4617744360902256, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 29302 + }, + { + "epoch": 1.4618345864661655, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 29303 + }, + { + "epoch": 1.4618947368421051, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 29304 + }, + { + "epoch": 1.4619548872180452, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 29305 + }, + { + "epoch": 1.4620150375939849, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 29306 + }, + { + "epoch": 1.4620751879699247, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 29307 + }, + { + "epoch": 1.4621353383458646, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 29308 + }, + { + "epoch": 1.4621954887218045, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 29309 + }, + { + "epoch": 1.4622556390977444, + "grad_norm": 1.46875, + "learning_rate": 3e-05, + "loss": 1.1541, + "step": 29310 + }, + { + "epoch": 1.4623157894736842, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 29311 + }, + { + "epoch": 1.462375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 29312 + }, + { + "epoch": 1.462436090225564, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 29313 + }, + { + "epoch": 1.4624962406015038, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 29314 + }, + { + "epoch": 1.4625563909774435, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 29315 + }, + { + "epoch": 1.4626165413533836, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 29316 + }, + { + "epoch": 1.4626766917293232, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 29317 + }, + { + "epoch": 1.4627368421052631, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 29318 + }, + { + "epoch": 1.462796992481203, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 29319 + }, + { + "epoch": 1.4628571428571429, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 29320 + }, + { + "epoch": 1.4629172932330827, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 29321 + }, + { + "epoch": 1.4629774436090226, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 29322 + }, + { + "epoch": 1.4630375939849625, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 29323 + }, + { + "epoch": 1.4630977443609021, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 29324 + }, + { + "epoch": 1.4631578947368422, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 29325 + }, + { + "epoch": 1.4632180451127819, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 29326 + }, + { + "epoch": 1.4632781954887217, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 29327 + }, + { + "epoch": 1.4633383458646616, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 29328 + }, + { + "epoch": 1.4633984962406015, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 29329 + }, + { + "epoch": 1.4634586466165413, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 29330 + }, + { + "epoch": 1.4635187969924812, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1822, + "step": 29331 + }, + { + "epoch": 1.463578947368421, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 29332 + }, + { + "epoch": 1.463639097744361, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 29333 + }, + { + "epoch": 1.4636992481203008, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 29334 + }, + { + "epoch": 1.4637593984962405, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 29335 + }, + { + "epoch": 1.4638195488721806, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 29336 + }, + { + "epoch": 1.4638796992481202, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 29337 + }, + { + "epoch": 1.46393984962406, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 29338 + }, + { + "epoch": 1.464, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 29339 + }, + { + "epoch": 1.4640601503759398, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 29340 + }, + { + "epoch": 1.4641203007518797, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 29341 + }, + { + "epoch": 1.4641804511278196, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 29342 + }, + { + "epoch": 1.4642406015037595, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 29343 + }, + { + "epoch": 1.4643007518796993, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 29344 + }, + { + "epoch": 1.4643609022556392, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 29345 + }, + { + "epoch": 1.4644210526315788, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 29346 + }, + { + "epoch": 1.4644812030075187, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 29347 + }, + { + "epoch": 1.4645413533834586, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 29348 + }, + { + "epoch": 1.4646015037593985, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 29349 + }, + { + "epoch": 1.4646616541353383, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 29350 + }, + { + "epoch": 1.4647218045112782, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 29351 + }, + { + "epoch": 1.464781954887218, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1956, + "step": 29352 + }, + { + "epoch": 1.464842105263158, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 29353 + }, + { + "epoch": 1.4649022556390978, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 29354 + }, + { + "epoch": 1.4649624060150375, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 29355 + }, + { + "epoch": 1.4650225563909776, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 29356 + }, + { + "epoch": 1.4650827067669172, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 29357 + }, + { + "epoch": 1.465142857142857, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 29358 + }, + { + "epoch": 1.465203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 29359 + }, + { + "epoch": 1.4652631578947368, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1418, + "step": 29360 + }, + { + "epoch": 1.4653233082706767, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 29361 + }, + { + "epoch": 1.4653834586466166, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 29362 + }, + { + "epoch": 1.4654436090225564, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 29363 + }, + { + "epoch": 1.4655037593984963, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 29364 + }, + { + "epoch": 1.4655639097744362, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1518, + "step": 29365 + }, + { + "epoch": 1.4656240601503758, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 29366 + }, + { + "epoch": 1.465684210526316, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 29367 + }, + { + "epoch": 1.4657443609022556, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 29368 + }, + { + "epoch": 1.4658045112781954, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 29369 + }, + { + "epoch": 1.4658646616541353, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 29370 + }, + { + "epoch": 1.4659248120300752, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 29371 + }, + { + "epoch": 1.465984962406015, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 29372 + }, + { + "epoch": 1.466045112781955, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 29373 + }, + { + "epoch": 1.4661052631578948, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 29374 + }, + { + "epoch": 1.4661654135338344, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 29375 + }, + { + "epoch": 1.4662255639097745, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 29376 + }, + { + "epoch": 1.4662857142857142, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 29377 + }, + { + "epoch": 1.466345864661654, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 29378 + }, + { + "epoch": 1.466406015037594, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 29379 + }, + { + "epoch": 1.4664661654135338, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 29380 + }, + { + "epoch": 1.4665263157894737, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 29381 + }, + { + "epoch": 1.4665864661654135, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 29382 + }, + { + "epoch": 1.4666466165413534, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 29383 + }, + { + "epoch": 1.4667067669172933, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 29384 + }, + { + "epoch": 1.4667669172932332, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 29385 + }, + { + "epoch": 1.4668270676691728, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 29386 + }, + { + "epoch": 1.466887218045113, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 29387 + }, + { + "epoch": 1.4669473684210526, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 29388 + }, + { + "epoch": 1.4670075187969924, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 29389 + }, + { + "epoch": 1.4670676691729323, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 29390 + }, + { + "epoch": 1.4671278195488722, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 29391 + }, + { + "epoch": 1.467187969924812, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 29392 + }, + { + "epoch": 1.467248120300752, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 29393 + }, + { + "epoch": 1.4673082706766918, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 29394 + }, + { + "epoch": 1.4673684210526317, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 29395 + }, + { + "epoch": 1.4674285714285715, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 29396 + }, + { + "epoch": 1.4674887218045112, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 29397 + }, + { + "epoch": 1.467548872180451, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 29398 + }, + { + "epoch": 1.467609022556391, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 29399 + }, + { + "epoch": 1.4676691729323308, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 29400 + }, + { + "epoch": 1.4677293233082707, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 29401 + }, + { + "epoch": 1.4677894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 29402 + }, + { + "epoch": 1.4678496240601504, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 29403 + }, + { + "epoch": 1.4679097744360903, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 29404 + }, + { + "epoch": 1.4679699248120301, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 29405 + }, + { + "epoch": 1.4680300751879698, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 29406 + }, + { + "epoch": 1.46809022556391, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 29407 + }, + { + "epoch": 1.4681503759398495, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1733, + "step": 29408 + }, + { + "epoch": 1.4682105263157894, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 29409 + }, + { + "epoch": 1.4682706766917293, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 29410 + }, + { + "epoch": 1.4683308270676692, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 29411 + }, + { + "epoch": 1.468390977443609, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 29412 + }, + { + "epoch": 1.468451127819549, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 29413 + }, + { + "epoch": 1.4685112781954888, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 29414 + }, + { + "epoch": 1.4685714285714286, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 29415 + }, + { + "epoch": 1.4686315789473685, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1854, + "step": 29416 + }, + { + "epoch": 1.4686917293233082, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 29417 + }, + { + "epoch": 1.4687518796992483, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 29418 + }, + { + "epoch": 1.468812030075188, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 29419 + }, + { + "epoch": 1.4688721804511278, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 29420 + }, + { + "epoch": 1.4689323308270676, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 29421 + }, + { + "epoch": 1.4689924812030075, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 29422 + }, + { + "epoch": 1.4690526315789474, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 29423 + }, + { + "epoch": 1.4691127819548873, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2969, + "step": 29424 + }, + { + "epoch": 1.4691729323308271, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 29425 + }, + { + "epoch": 1.4692330827067668, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 29426 + }, + { + "epoch": 1.4692932330827069, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.928, + "step": 29427 + }, + { + "epoch": 1.4693533834586465, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 29428 + }, + { + "epoch": 1.4694135338345864, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 29429 + }, + { + "epoch": 1.4694736842105263, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 29430 + }, + { + "epoch": 1.4695338345864661, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 29431 + }, + { + "epoch": 1.469593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 29432 + }, + { + "epoch": 1.4696541353383459, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 29433 + }, + { + "epoch": 1.4697142857142858, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 29434 + }, + { + "epoch": 1.4697744360902256, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 29435 + }, + { + "epoch": 1.4698345864661655, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 29436 + }, + { + "epoch": 1.4698947368421051, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 29437 + }, + { + "epoch": 1.4699548872180452, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 29438 + }, + { + "epoch": 1.4700150375939849, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 29439 + }, + { + "epoch": 1.4700751879699248, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 29440 + }, + { + "epoch": 1.4701353383458646, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 29441 + }, + { + "epoch": 1.4701954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 29442 + }, + { + "epoch": 1.4702556390977444, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 29443 + }, + { + "epoch": 1.4703157894736842, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 29444 + }, + { + "epoch": 1.4703759398496241, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 29445 + }, + { + "epoch": 1.470436090225564, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 29446 + }, + { + "epoch": 1.4704962406015039, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 29447 + }, + { + "epoch": 1.4705563909774435, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 29448 + }, + { + "epoch": 1.4706165413533834, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 29449 + }, + { + "epoch": 1.4706766917293232, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 29450 + }, + { + "epoch": 1.4707368421052631, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 29451 + }, + { + "epoch": 1.470796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 29452 + }, + { + "epoch": 1.4708571428571429, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 29453 + }, + { + "epoch": 1.4709172932330827, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 29454 + }, + { + "epoch": 1.4709774436090226, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.2081, + "step": 29455 + }, + { + "epoch": 1.4710375939849625, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2419, + "step": 29456 + }, + { + "epoch": 1.4710977443609021, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 29457 + }, + { + "epoch": 1.4711578947368422, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 29458 + }, + { + "epoch": 1.4712180451127819, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 29459 + }, + { + "epoch": 1.4712781954887217, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.271, + "step": 29460 + }, + { + "epoch": 1.4713383458646616, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 29461 + }, + { + "epoch": 1.4713984962406015, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 29462 + }, + { + "epoch": 1.4714586466165414, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 29463 + }, + { + "epoch": 1.4715187969924812, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 29464 + }, + { + "epoch": 1.471578947368421, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 29465 + }, + { + "epoch": 1.471639097744361, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 29466 + }, + { + "epoch": 1.4716992481203008, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 29467 + }, + { + "epoch": 1.4717593984962405, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 29468 + }, + { + "epoch": 1.4718195488721806, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 29469 + }, + { + "epoch": 1.4718796992481202, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 29470 + }, + { + "epoch": 1.47193984962406, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 29471 + }, + { + "epoch": 1.472, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 29472 + }, + { + "epoch": 1.4720601503759398, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 29473 + }, + { + "epoch": 1.4721203007518797, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 29474 + }, + { + "epoch": 1.4721804511278196, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 29475 + }, + { + "epoch": 1.4722406015037595, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 29476 + }, + { + "epoch": 1.472300751879699, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 29477 + }, + { + "epoch": 1.4723609022556392, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 29478 + }, + { + "epoch": 1.4724210526315789, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 29479 + }, + { + "epoch": 1.4724812030075187, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 29480 + }, + { + "epoch": 1.4725413533834586, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 29481 + }, + { + "epoch": 1.4726015037593985, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 29482 + }, + { + "epoch": 1.4726616541353383, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 29483 + }, + { + "epoch": 1.4727218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 29484 + }, + { + "epoch": 1.472781954887218, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 29485 + }, + { + "epoch": 1.472842105263158, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1932, + "step": 29486 + }, + { + "epoch": 1.4729022556390978, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 29487 + }, + { + "epoch": 1.4729624060150375, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 29488 + }, + { + "epoch": 1.4730225563909776, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 29489 + }, + { + "epoch": 1.4730827067669172, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 29490 + }, + { + "epoch": 1.473142857142857, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 29491 + }, + { + "epoch": 1.473203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 29492 + }, + { + "epoch": 1.4732631578947368, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 29493 + }, + { + "epoch": 1.4733233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 29494 + }, + { + "epoch": 1.4733834586466166, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 29495 + }, + { + "epoch": 1.4734436090225564, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 29496 + }, + { + "epoch": 1.4735037593984963, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 29497 + }, + { + "epoch": 1.4735639097744362, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 29498 + }, + { + "epoch": 1.4736240601503758, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 29499 + }, + { + "epoch": 1.4736842105263157, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 29500 + }, + { + "epoch": 1.4737443609022556, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 29501 + }, + { + "epoch": 1.4738045112781955, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 29502 + }, + { + "epoch": 1.4738646616541353, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 29503 + }, + { + "epoch": 1.4739248120300752, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 29504 + }, + { + "epoch": 1.473984962406015, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 29505 + }, + { + "epoch": 1.474045112781955, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 29506 + }, + { + "epoch": 1.4741052631578948, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 29507 + }, + { + "epoch": 1.4741654135338345, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 29508 + }, + { + "epoch": 1.4742255639097746, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 29509 + }, + { + "epoch": 1.4742857142857142, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 29510 + }, + { + "epoch": 1.474345864661654, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 29511 + }, + { + "epoch": 1.474406015037594, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 29512 + }, + { + "epoch": 1.4744661654135338, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 29513 + }, + { + "epoch": 1.4745263157894737, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9274, + "step": 29514 + }, + { + "epoch": 1.4745864661654136, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 29515 + }, + { + "epoch": 1.4746466165413534, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 29516 + }, + { + "epoch": 1.4747067669172933, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 29517 + }, + { + "epoch": 1.4747669172932332, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 29518 + }, + { + "epoch": 1.4748270676691728, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 29519 + }, + { + "epoch": 1.474887218045113, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 29520 + }, + { + "epoch": 1.4749473684210526, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 29521 + }, + { + "epoch": 1.4750075187969924, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 29522 + }, + { + "epoch": 1.4750676691729323, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9323, + "step": 29523 + }, + { + "epoch": 1.4751278195488722, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 29524 + }, + { + "epoch": 1.475187969924812, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 29525 + }, + { + "epoch": 1.475248120300752, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 29526 + }, + { + "epoch": 1.4753082706766918, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 29527 + }, + { + "epoch": 1.4753684210526314, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 29528 + }, + { + "epoch": 1.4754285714285715, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 29529 + }, + { + "epoch": 1.4754887218045112, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 29530 + }, + { + "epoch": 1.475548872180451, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 29531 + }, + { + "epoch": 1.475609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 29532 + }, + { + "epoch": 1.4756691729323308, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 29533 + }, + { + "epoch": 1.4757293233082707, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 29534 + }, + { + "epoch": 1.4757894736842105, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 29535 + }, + { + "epoch": 1.4758496240601504, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 29536 + }, + { + "epoch": 1.4759097744360903, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 29537 + }, + { + "epoch": 1.4759699248120302, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 29538 + }, + { + "epoch": 1.4760300751879698, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 29539 + }, + { + "epoch": 1.47609022556391, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 29540 + }, + { + "epoch": 1.4761503759398495, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 29541 + }, + { + "epoch": 1.4762105263157894, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 29542 + }, + { + "epoch": 1.4762706766917293, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 29543 + }, + { + "epoch": 1.4763308270676692, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 29544 + }, + { + "epoch": 1.476390977443609, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 29545 + }, + { + "epoch": 1.476451127819549, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 29546 + }, + { + "epoch": 1.4765112781954888, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 29547 + }, + { + "epoch": 1.4765714285714286, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 29548 + }, + { + "epoch": 1.4766315789473685, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 29549 + }, + { + "epoch": 1.4766917293233082, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 29550 + }, + { + "epoch": 1.476751879699248, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 29551 + }, + { + "epoch": 1.476812030075188, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 29552 + }, + { + "epoch": 1.4768721804511278, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 29553 + }, + { + "epoch": 1.4769323308270677, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 29554 + }, + { + "epoch": 1.4769924812030075, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 29555 + }, + { + "epoch": 1.4770526315789474, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 29556 + }, + { + "epoch": 1.4771127819548873, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 29557 + }, + { + "epoch": 1.4771729323308271, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 29558 + }, + { + "epoch": 1.4772330827067668, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 29559 + }, + { + "epoch": 1.4772932330827069, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 29560 + }, + { + "epoch": 1.4773533834586465, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.2148, + "step": 29561 + }, + { + "epoch": 1.4774135338345864, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 29562 + }, + { + "epoch": 1.4774736842105263, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 29563 + }, + { + "epoch": 1.4775338345864661, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 29564 + }, + { + "epoch": 1.477593984962406, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 29565 + }, + { + "epoch": 1.4776541353383459, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 29566 + }, + { + "epoch": 1.4777142857142858, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 29567 + }, + { + "epoch": 1.4777744360902256, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 29568 + }, + { + "epoch": 1.4778345864661655, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 29569 + }, + { + "epoch": 1.4778947368421052, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 29570 + }, + { + "epoch": 1.4779548872180452, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 29571 + }, + { + "epoch": 1.478015037593985, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 29572 + }, + { + "epoch": 1.4780751879699248, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 29573 + }, + { + "epoch": 1.4781353383458646, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 29574 + }, + { + "epoch": 1.4781954887218045, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 29575 + }, + { + "epoch": 1.4782556390977444, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 29576 + }, + { + "epoch": 1.4783157894736842, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 29577 + }, + { + "epoch": 1.4783759398496241, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 29578 + }, + { + "epoch": 1.478436090225564, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 29579 + }, + { + "epoch": 1.4784962406015039, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 29580 + }, + { + "epoch": 1.4785563909774435, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 29581 + }, + { + "epoch": 1.4786165413533834, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 29582 + }, + { + "epoch": 1.4786766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 29583 + }, + { + "epoch": 1.4787368421052631, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 29584 + }, + { + "epoch": 1.478796992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 29585 + }, + { + "epoch": 1.4788571428571429, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 29586 + }, + { + "epoch": 1.4789172932330827, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 29587 + }, + { + "epoch": 1.4789774436090226, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.2253, + "step": 29588 + }, + { + "epoch": 1.4790375939849625, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 29589 + }, + { + "epoch": 1.4790977443609021, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 29590 + }, + { + "epoch": 1.4791578947368422, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 29591 + }, + { + "epoch": 1.4792180451127819, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 29592 + }, + { + "epoch": 1.4792781954887217, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 29593 + }, + { + "epoch": 1.4793383458646616, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 29594 + }, + { + "epoch": 1.4793984962406015, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 29595 + }, + { + "epoch": 1.4794586466165414, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 29596 + }, + { + "epoch": 1.4795187969924812, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 29597 + }, + { + "epoch": 1.479578947368421, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 29598 + }, + { + "epoch": 1.479639097744361, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 29599 + }, + { + "epoch": 1.4796992481203008, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 29600 + }, + { + "epoch": 1.4797593984962405, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 29601 + }, + { + "epoch": 1.4798195488721804, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 29602 + }, + { + "epoch": 1.4798796992481202, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 29603 + }, + { + "epoch": 1.47993984962406, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 29604 + }, + { + "epoch": 1.48, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 29605 + }, + { + "epoch": 1.4800601503759399, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 29606 + }, + { + "epoch": 1.4801203007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 29607 + }, + { + "epoch": 1.4801804511278196, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 29608 + }, + { + "epoch": 1.4802406015037595, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 29609 + }, + { + "epoch": 1.4803007518796991, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 29610 + }, + { + "epoch": 1.4803609022556392, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 29611 + }, + { + "epoch": 1.4804210526315789, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 29612 + }, + { + "epoch": 1.4804812030075187, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 29613 + }, + { + "epoch": 1.4805413533834586, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 29614 + }, + { + "epoch": 1.4806015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1332, + "step": 29615 + }, + { + "epoch": 1.4806616541353383, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 29616 + }, + { + "epoch": 1.4807218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 29617 + }, + { + "epoch": 1.480781954887218, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 29618 + }, + { + "epoch": 1.480842105263158, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 29619 + }, + { + "epoch": 1.4809022556390978, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 29620 + }, + { + "epoch": 1.4809624060150375, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 29621 + }, + { + "epoch": 1.4810225563909776, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 29622 + }, + { + "epoch": 1.4810827067669172, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 29623 + }, + { + "epoch": 1.481142857142857, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 29624 + }, + { + "epoch": 1.481203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 29625 + }, + { + "epoch": 1.4812631578947368, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 29626 + }, + { + "epoch": 1.4813233082706767, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 29627 + }, + { + "epoch": 1.4813834586466166, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 29628 + }, + { + "epoch": 1.4814436090225565, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 29629 + }, + { + "epoch": 1.4815037593984963, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 29630 + }, + { + "epoch": 1.4815639097744362, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 29631 + }, + { + "epoch": 1.4816240601503758, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1847, + "step": 29632 + }, + { + "epoch": 1.4816842105263157, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.236, + "step": 29633 + }, + { + "epoch": 1.4817443609022556, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 29634 + }, + { + "epoch": 1.4818045112781955, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 29635 + }, + { + "epoch": 1.4818646616541353, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 29636 + }, + { + "epoch": 1.4819248120300752, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 29637 + }, + { + "epoch": 1.481984962406015, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 29638 + }, + { + "epoch": 1.482045112781955, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 29639 + }, + { + "epoch": 1.4821052631578948, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 29640 + }, + { + "epoch": 1.4821654135338345, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 29641 + }, + { + "epoch": 1.4822255639097746, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 29642 + }, + { + "epoch": 1.4822857142857142, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 29643 + }, + { + "epoch": 1.482345864661654, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 29644 + }, + { + "epoch": 1.482406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 29645 + }, + { + "epoch": 1.4824661654135338, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 29646 + }, + { + "epoch": 1.4825263157894737, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 29647 + }, + { + "epoch": 1.4825864661654136, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 29648 + }, + { + "epoch": 1.4826466165413534, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 29649 + }, + { + "epoch": 1.4827067669172933, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 29650 + }, + { + "epoch": 1.4827669172932332, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 29651 + }, + { + "epoch": 1.4828270676691728, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 29652 + }, + { + "epoch": 1.482887218045113, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 29653 + }, + { + "epoch": 1.4829473684210526, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2692, + "step": 29654 + }, + { + "epoch": 1.4830075187969924, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 29655 + }, + { + "epoch": 1.4830676691729323, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 29656 + }, + { + "epoch": 1.4831278195488722, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 29657 + }, + { + "epoch": 1.483187969924812, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 29658 + }, + { + "epoch": 1.483248120300752, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 29659 + }, + { + "epoch": 1.4833082706766918, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 29660 + }, + { + "epoch": 1.4833684210526314, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 29661 + }, + { + "epoch": 1.4834285714285715, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 29662 + }, + { + "epoch": 1.4834887218045112, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 29663 + }, + { + "epoch": 1.483548872180451, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 29664 + }, + { + "epoch": 1.483609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 29665 + }, + { + "epoch": 1.4836691729323308, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 29666 + }, + { + "epoch": 1.4837293233082707, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 29667 + }, + { + "epoch": 1.4837894736842105, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 29668 + }, + { + "epoch": 1.4838496240601504, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 29669 + }, + { + "epoch": 1.4839097744360903, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 29670 + }, + { + "epoch": 1.4839699248120302, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 29671 + }, + { + "epoch": 1.4840300751879698, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 29672 + }, + { + "epoch": 1.48409022556391, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 29673 + }, + { + "epoch": 1.4841503759398496, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 29674 + }, + { + "epoch": 1.4842105263157894, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 29675 + }, + { + "epoch": 1.4842706766917293, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 29676 + }, + { + "epoch": 1.4843308270676692, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 29677 + }, + { + "epoch": 1.484390977443609, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 29678 + }, + { + "epoch": 1.484451127819549, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1496, + "step": 29679 + }, + { + "epoch": 1.4845112781954888, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 29680 + }, + { + "epoch": 1.4845714285714287, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 29681 + }, + { + "epoch": 1.4846315789473685, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 29682 + }, + { + "epoch": 1.4846917293233082, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 29683 + }, + { + "epoch": 1.484751879699248, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 29684 + }, + { + "epoch": 1.484812030075188, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 29685 + }, + { + "epoch": 1.4848721804511278, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 29686 + }, + { + "epoch": 1.4849323308270677, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 29687 + }, + { + "epoch": 1.4849924812030075, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.2698, + "step": 29688 + }, + { + "epoch": 1.4850526315789474, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1751, + "step": 29689 + }, + { + "epoch": 1.4851127819548873, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 29690 + }, + { + "epoch": 1.4851729323308271, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2281, + "step": 29691 + }, + { + "epoch": 1.4852330827067668, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 29692 + }, + { + "epoch": 1.4852932330827069, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 29693 + }, + { + "epoch": 1.4853533834586465, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 29694 + }, + { + "epoch": 1.4854135338345864, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 29695 + }, + { + "epoch": 1.4854736842105263, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 29696 + }, + { + "epoch": 1.4855338345864662, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 29697 + }, + { + "epoch": 1.485593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 29698 + }, + { + "epoch": 1.485654135338346, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 29699 + }, + { + "epoch": 1.4857142857142858, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 29700 + }, + { + "epoch": 1.4857744360902256, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 29701 + }, + { + "epoch": 1.4858345864661655, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 29702 + }, + { + "epoch": 1.4858947368421052, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 29703 + }, + { + "epoch": 1.4859548872180453, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 29704 + }, + { + "epoch": 1.486015037593985, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.1693, + "step": 29705 + }, + { + "epoch": 1.4860751879699248, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 29706 + }, + { + "epoch": 1.4861353383458646, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 29707 + }, + { + "epoch": 1.4861954887218045, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 29708 + }, + { + "epoch": 1.4862556390977444, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 29709 + }, + { + "epoch": 1.4863157894736843, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 29710 + }, + { + "epoch": 1.4863759398496241, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 29711 + }, + { + "epoch": 1.4864360902255638, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 29712 + }, + { + "epoch": 1.4864962406015039, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 29713 + }, + { + "epoch": 1.4865563909774435, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 29714 + }, + { + "epoch": 1.4866165413533834, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 29715 + }, + { + "epoch": 1.4866766917293233, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 29716 + }, + { + "epoch": 1.4867368421052631, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 29717 + }, + { + "epoch": 1.486796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 29718 + }, + { + "epoch": 1.4868571428571429, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 29719 + }, + { + "epoch": 1.4869172932330827, + "grad_norm": 1.703125, + "learning_rate": 3e-05, + "loss": 1.1586, + "step": 29720 + }, + { + "epoch": 1.4869774436090226, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 29721 + }, + { + "epoch": 1.4870375939849625, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.8875, + "step": 29722 + }, + { + "epoch": 1.4870977443609021, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 29723 + }, + { + "epoch": 1.4871578947368422, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 29724 + }, + { + "epoch": 1.4872180451127819, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 29725 + }, + { + "epoch": 1.4872781954887218, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 29726 + }, + { + "epoch": 1.4873383458646616, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 29727 + }, + { + "epoch": 1.4873984962406015, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 29728 + }, + { + "epoch": 1.4874586466165414, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 29729 + }, + { + "epoch": 1.4875187969924812, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 29730 + }, + { + "epoch": 1.4875789473684211, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 29731 + }, + { + "epoch": 1.487639097744361, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 29732 + }, + { + "epoch": 1.4876992481203009, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 29733 + }, + { + "epoch": 1.4877593984962405, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 29734 + }, + { + "epoch": 1.4878195488721804, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 29735 + }, + { + "epoch": 1.4878796992481202, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 29736 + }, + { + "epoch": 1.4879398496240601, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 29737 + }, + { + "epoch": 1.488, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 29738 + }, + { + "epoch": 1.4880601503759399, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 29739 + }, + { + "epoch": 1.4881203007518797, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 29740 + }, + { + "epoch": 1.4881804511278196, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 29741 + }, + { + "epoch": 1.4882406015037595, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 29742 + }, + { + "epoch": 1.4883007518796991, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 29743 + }, + { + "epoch": 1.4883609022556392, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 29744 + }, + { + "epoch": 1.4884210526315789, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 29745 + }, + { + "epoch": 1.4884812030075187, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 29746 + }, + { + "epoch": 1.4885413533834586, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.3537, + "step": 29747 + }, + { + "epoch": 1.4886015037593985, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 29748 + }, + { + "epoch": 1.4886616541353384, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 29749 + }, + { + "epoch": 1.4887218045112782, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 29750 + }, + { + "epoch": 1.488781954887218, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 29751 + }, + { + "epoch": 1.488842105263158, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 29752 + }, + { + "epoch": 1.4889022556390978, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 29753 + }, + { + "epoch": 1.4889624060150375, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 29754 + }, + { + "epoch": 1.4890225563909776, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 29755 + }, + { + "epoch": 1.4890827067669172, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 29756 + }, + { + "epoch": 1.489142857142857, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 29757 + }, + { + "epoch": 1.489203007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 29758 + }, + { + "epoch": 1.4892631578947368, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 29759 + }, + { + "epoch": 1.4893233082706767, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 29760 + }, + { + "epoch": 1.4893834586466166, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 29761 + }, + { + "epoch": 1.4894436090225565, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 29762 + }, + { + "epoch": 1.489503759398496, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 29763 + }, + { + "epoch": 1.4895639097744362, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 29764 + }, + { + "epoch": 1.4896240601503759, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 29765 + }, + { + "epoch": 1.4896842105263157, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 29766 + }, + { + "epoch": 1.4897443609022556, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 29767 + }, + { + "epoch": 1.4898045112781955, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 29768 + }, + { + "epoch": 1.4898646616541353, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 29769 + }, + { + "epoch": 1.4899248120300752, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 29770 + }, + { + "epoch": 1.489984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 29771 + }, + { + "epoch": 1.490045112781955, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 29772 + }, + { + "epoch": 1.4901052631578948, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 29773 + }, + { + "epoch": 1.4901654135338345, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 29774 + }, + { + "epoch": 1.4902255639097746, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 29775 + }, + { + "epoch": 1.4902857142857142, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 29776 + }, + { + "epoch": 1.490345864661654, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 29777 + }, + { + "epoch": 1.490406015037594, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 29778 + }, + { + "epoch": 1.4904661654135338, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 29779 + }, + { + "epoch": 1.4905263157894737, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 29780 + }, + { + "epoch": 1.4905864661654136, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 29781 + }, + { + "epoch": 1.4906466165413534, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 29782 + }, + { + "epoch": 1.4907067669172933, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 29783 + }, + { + "epoch": 1.4907669172932332, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 29784 + }, + { + "epoch": 1.4908270676691728, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 29785 + }, + { + "epoch": 1.4908872180451127, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 29786 + }, + { + "epoch": 1.4909473684210526, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 29787 + }, + { + "epoch": 1.4910075187969924, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 29788 + }, + { + "epoch": 1.4910676691729323, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 29789 + }, + { + "epoch": 1.4911278195488722, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 29790 + }, + { + "epoch": 1.491187969924812, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 29791 + }, + { + "epoch": 1.491248120300752, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 29792 + }, + { + "epoch": 1.4913082706766918, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9482, + "step": 29793 + }, + { + "epoch": 1.4913684210526315, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 29794 + }, + { + "epoch": 1.4914285714285715, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 29795 + }, + { + "epoch": 1.4914887218045112, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 29796 + }, + { + "epoch": 1.491548872180451, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 29797 + }, + { + "epoch": 1.491609022556391, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 29798 + }, + { + "epoch": 1.4916691729323308, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 29799 + }, + { + "epoch": 1.4917293233082707, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 29800 + }, + { + "epoch": 1.4917894736842106, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 29801 + }, + { + "epoch": 1.4918496240601504, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 29802 + }, + { + "epoch": 1.4919097744360903, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 29803 + }, + { + "epoch": 1.4919699248120302, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.2161, + "step": 29804 + }, + { + "epoch": 1.4920300751879698, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 29805 + }, + { + "epoch": 1.49209022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 29806 + }, + { + "epoch": 1.4921503759398496, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 29807 + }, + { + "epoch": 1.4922105263157894, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 29808 + }, + { + "epoch": 1.4922706766917293, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 29809 + }, + { + "epoch": 1.4923308270676692, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 29810 + }, + { + "epoch": 1.492390977443609, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 29811 + }, + { + "epoch": 1.492451127819549, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 29812 + }, + { + "epoch": 1.4925112781954888, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 29813 + }, + { + "epoch": 1.4925714285714284, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 29814 + }, + { + "epoch": 1.4926315789473685, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 29815 + }, + { + "epoch": 1.4926917293233082, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 29816 + }, + { + "epoch": 1.492751879699248, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 29817 + }, + { + "epoch": 1.492812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 29818 + }, + { + "epoch": 1.4928721804511278, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 29819 + }, + { + "epoch": 1.4929323308270677, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 29820 + }, + { + "epoch": 1.4929924812030075, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 29821 + }, + { + "epoch": 1.4930526315789474, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 29822 + }, + { + "epoch": 1.4931127819548873, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 29823 + }, + { + "epoch": 1.4931729323308272, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 29824 + }, + { + "epoch": 1.4932330827067668, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 29825 + }, + { + "epoch": 1.493293233082707, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 29826 + }, + { + "epoch": 1.4933533834586465, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 29827 + }, + { + "epoch": 1.4934135338345864, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 29828 + }, + { + "epoch": 1.4934736842105263, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 29829 + }, + { + "epoch": 1.4935338345864662, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 29830 + }, + { + "epoch": 1.493593984962406, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 29831 + }, + { + "epoch": 1.493654135338346, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 29832 + }, + { + "epoch": 1.4937142857142858, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 29833 + }, + { + "epoch": 1.4937744360902256, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 29834 + }, + { + "epoch": 1.4938345864661655, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 29835 + }, + { + "epoch": 1.4938947368421052, + "grad_norm": 1.40625, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 29836 + }, + { + "epoch": 1.493954887218045, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 29837 + }, + { + "epoch": 1.494015037593985, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 29838 + }, + { + "epoch": 1.4940751879699248, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 29839 + }, + { + "epoch": 1.4941353383458646, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 29840 + }, + { + "epoch": 1.4941954887218045, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 29841 + }, + { + "epoch": 1.4942556390977444, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 29842 + }, + { + "epoch": 1.4943157894736843, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 29843 + }, + { + "epoch": 1.4943759398496241, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 29844 + }, + { + "epoch": 1.4944360902255638, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 29845 + }, + { + "epoch": 1.4944962406015039, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 29846 + }, + { + "epoch": 1.4945563909774435, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 29847 + }, + { + "epoch": 1.4946165413533834, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 29848 + }, + { + "epoch": 1.4946766917293233, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 29849 + }, + { + "epoch": 1.4947368421052631, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 29850 + }, + { + "epoch": 1.494796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 29851 + }, + { + "epoch": 1.4948571428571429, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 29852 + }, + { + "epoch": 1.4949172932330828, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 29853 + }, + { + "epoch": 1.4949774436090226, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 29854 + }, + { + "epoch": 1.4950375939849625, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 29855 + }, + { + "epoch": 1.4950977443609021, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 29856 + }, + { + "epoch": 1.4951578947368422, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 29857 + }, + { + "epoch": 1.495218045112782, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.197, + "step": 29858 + }, + { + "epoch": 1.4952781954887218, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 29859 + }, + { + "epoch": 1.4953383458646616, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 29860 + }, + { + "epoch": 1.4953984962406015, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 29861 + }, + { + "epoch": 1.4954586466165414, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 29862 + }, + { + "epoch": 1.4955187969924812, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 29863 + }, + { + "epoch": 1.4955789473684211, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 29864 + }, + { + "epoch": 1.4956390977443608, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 29865 + }, + { + "epoch": 1.4956992481203009, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 29866 + }, + { + "epoch": 1.4957593984962405, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 29867 + }, + { + "epoch": 1.4958195488721804, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 29868 + }, + { + "epoch": 1.4958796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 29869 + }, + { + "epoch": 1.4959398496240601, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 29870 + }, + { + "epoch": 1.496, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 29871 + }, + { + "epoch": 1.4960601503759399, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 29872 + }, + { + "epoch": 1.4961203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 29873 + }, + { + "epoch": 1.4961804511278196, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 29874 + }, + { + "epoch": 1.4962406015037595, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 29875 + }, + { + "epoch": 1.4963007518796991, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 29876 + }, + { + "epoch": 1.4963609022556392, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 29877 + }, + { + "epoch": 1.4964210526315789, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 29878 + }, + { + "epoch": 1.4964812030075187, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 29879 + }, + { + "epoch": 1.4965413533834586, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 29880 + }, + { + "epoch": 1.4966015037593985, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 29881 + }, + { + "epoch": 1.4966616541353384, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 29882 + }, + { + "epoch": 1.4967218045112782, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 29883 + }, + { + "epoch": 1.496781954887218, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 29884 + }, + { + "epoch": 1.496842105263158, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 29885 + }, + { + "epoch": 1.4969022556390978, + "grad_norm": 1.4140625, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 29886 + }, + { + "epoch": 1.4969624060150375, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 29887 + }, + { + "epoch": 1.4970225563909774, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 29888 + }, + { + "epoch": 1.4970827067669172, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 29889 + }, + { + "epoch": 1.497142857142857, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9183, + "step": 29890 + }, + { + "epoch": 1.497203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 29891 + }, + { + "epoch": 1.4972631578947369, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 29892 + }, + { + "epoch": 1.4973233082706767, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 29893 + }, + { + "epoch": 1.4973834586466166, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 29894 + }, + { + "epoch": 1.4974436090225565, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 29895 + }, + { + "epoch": 1.4975037593984961, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 29896 + }, + { + "epoch": 1.4975639097744362, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 29897 + }, + { + "epoch": 1.4976240601503759, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 29898 + }, + { + "epoch": 1.4976842105263157, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 29899 + }, + { + "epoch": 1.4977443609022556, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 29900 + }, + { + "epoch": 1.4978045112781955, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 29901 + }, + { + "epoch": 1.4978646616541353, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 29902 + }, + { + "epoch": 1.4979248120300752, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 29903 + }, + { + "epoch": 1.497984962406015, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 29904 + }, + { + "epoch": 1.498045112781955, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 29905 + }, + { + "epoch": 1.4981052631578948, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 29906 + }, + { + "epoch": 1.4981654135338345, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 29907 + }, + { + "epoch": 1.4982255639097746, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 29908 + }, + { + "epoch": 1.4982857142857142, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 29909 + }, + { + "epoch": 1.498345864661654, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 29910 + }, + { + "epoch": 1.498406015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 29911 + }, + { + "epoch": 1.4984661654135338, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 29912 + }, + { + "epoch": 1.4985263157894737, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2244, + "step": 29913 + }, + { + "epoch": 1.4985864661654136, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 29914 + }, + { + "epoch": 1.4986466165413534, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 29915 + }, + { + "epoch": 1.498706766917293, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 29916 + }, + { + "epoch": 1.4987669172932332, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2142, + "step": 29917 + }, + { + "epoch": 1.4988270676691728, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1746, + "step": 29918 + }, + { + "epoch": 1.4988872180451127, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 29919 + }, + { + "epoch": 1.4989473684210526, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 29920 + }, + { + "epoch": 1.4990075187969925, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 29921 + }, + { + "epoch": 1.4990676691729323, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 29922 + }, + { + "epoch": 1.4991278195488722, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 29923 + }, + { + "epoch": 1.499187969924812, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 29924 + }, + { + "epoch": 1.499248120300752, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 29925 + }, + { + "epoch": 1.4993082706766918, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 29926 + }, + { + "epoch": 1.4993684210526315, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 29927 + }, + { + "epoch": 1.4994285714285716, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 29928 + }, + { + "epoch": 1.4994887218045112, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 29929 + }, + { + "epoch": 1.499548872180451, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 29930 + }, + { + "epoch": 1.499609022556391, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 29931 + }, + { + "epoch": 1.4996691729323308, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 29932 + }, + { + "epoch": 1.4997293233082707, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 29933 + }, + { + "epoch": 1.4997894736842106, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 29934 + }, + { + "epoch": 1.4998496240601504, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 29935 + }, + { + "epoch": 1.4999097744360903, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 29936 + }, + { + "epoch": 1.4999699248120302, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 29937 + }, + { + "epoch": 1.5000300751879698, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 29938 + }, + { + "epoch": 1.50009022556391, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 29939 + }, + { + "epoch": 1.5001503759398496, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 29940 + }, + { + "epoch": 1.5002105263157894, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 29941 + }, + { + "epoch": 1.5002706766917293, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 29942 + }, + { + "epoch": 1.5003308270676692, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 29943 + }, + { + "epoch": 1.500390977443609, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 29944 + }, + { + "epoch": 1.500451127819549, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 29945 + }, + { + "epoch": 1.5005112781954888, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 29946 + }, + { + "epoch": 1.5005714285714284, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 29947 + }, + { + "epoch": 1.5006315789473685, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 29948 + }, + { + "epoch": 1.5006917293233082, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 29949 + }, + { + "epoch": 1.5007518796992483, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 29950 + }, + { + "epoch": 1.500812030075188, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 29951 + }, + { + "epoch": 1.5008721804511278, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 29952 + }, + { + "epoch": 1.5009323308270677, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 29953 + }, + { + "epoch": 1.5009924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 29954 + }, + { + "epoch": 1.5010526315789474, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 29955 + }, + { + "epoch": 1.501112781954887, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 29956 + }, + { + "epoch": 1.5011729323308272, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 29957 + }, + { + "epoch": 1.5012330827067668, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 29958 + }, + { + "epoch": 1.501293233082707, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 29959 + }, + { + "epoch": 1.5013533834586466, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 29960 + }, + { + "epoch": 1.5014135338345864, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 29961 + }, + { + "epoch": 1.5014736842105263, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 29962 + }, + { + "epoch": 1.5015338345864662, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 29963 + }, + { + "epoch": 1.501593984962406, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 29964 + }, + { + "epoch": 1.501654135338346, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 29965 + }, + { + "epoch": 1.5017142857142858, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 29966 + }, + { + "epoch": 1.5017744360902254, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 29967 + }, + { + "epoch": 1.5018345864661655, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 29968 + }, + { + "epoch": 1.5018947368421052, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 29969 + }, + { + "epoch": 1.5019548872180453, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 29970 + }, + { + "epoch": 1.502015037593985, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 29971 + }, + { + "epoch": 1.5020751879699248, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 29972 + }, + { + "epoch": 1.5021353383458647, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 29973 + }, + { + "epoch": 1.5021954887218045, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2989, + "step": 29974 + }, + { + "epoch": 1.5022556390977444, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 29975 + }, + { + "epoch": 1.502315789473684, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 29976 + }, + { + "epoch": 1.5023759398496241, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 29977 + }, + { + "epoch": 1.5024360902255638, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 29978 + }, + { + "epoch": 1.5024962406015039, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 29979 + }, + { + "epoch": 1.5025563909774435, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 29980 + }, + { + "epoch": 1.5026165413533834, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 29981 + }, + { + "epoch": 1.5026766917293233, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 29982 + }, + { + "epoch": 1.5027368421052631, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 29983 + }, + { + "epoch": 1.502796992481203, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 29984 + }, + { + "epoch": 1.502857142857143, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 29985 + }, + { + "epoch": 1.5029172932330828, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 29986 + }, + { + "epoch": 1.5029774436090224, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 29987 + }, + { + "epoch": 1.5030375939849625, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 29988 + }, + { + "epoch": 1.5030977443609022, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 29989 + }, + { + "epoch": 1.5031578947368422, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1441, + "step": 29990 + }, + { + "epoch": 1.503218045112782, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 29991 + }, + { + "epoch": 1.5032781954887218, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 29992 + }, + { + "epoch": 1.5033383458646616, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.3164, + "step": 29993 + }, + { + "epoch": 1.5033984962406015, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 29994 + }, + { + "epoch": 1.5034586466165414, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 29995 + }, + { + "epoch": 1.5035187969924813, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 29996 + }, + { + "epoch": 1.5035789473684211, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 29997 + }, + { + "epoch": 1.5036390977443608, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 29998 + }, + { + "epoch": 1.5036992481203009, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 29999 + }, + { + "epoch": 1.5037593984962405, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 30000 + }, + { + "epoch": 1.5038195488721806, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 30001 + }, + { + "epoch": 1.5038796992481203, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 30002 + }, + { + "epoch": 1.5039398496240601, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 30003 + }, + { + "epoch": 1.504, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 30004 + }, + { + "epoch": 1.5040601503759399, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 30005 + }, + { + "epoch": 1.5041203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 30006 + }, + { + "epoch": 1.5041804511278194, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 30007 + }, + { + "epoch": 1.5042406015037595, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 30008 + }, + { + "epoch": 1.5043007518796991, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 30009 + }, + { + "epoch": 1.5043609022556392, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 30010 + }, + { + "epoch": 1.5044210526315789, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 30011 + }, + { + "epoch": 1.5044812030075188, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 30012 + }, + { + "epoch": 1.5045413533834586, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 30013 + }, + { + "epoch": 1.5046015037593985, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 30014 + }, + { + "epoch": 1.5046616541353384, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 30015 + }, + { + "epoch": 1.5047218045112782, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 30016 + }, + { + "epoch": 1.504781954887218, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 30017 + }, + { + "epoch": 1.5048421052631578, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 30018 + }, + { + "epoch": 1.5049022556390979, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1879, + "step": 30019 + }, + { + "epoch": 1.5049624060150375, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 30020 + }, + { + "epoch": 1.5050225563909776, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 30021 + }, + { + "epoch": 1.5050827067669172, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 30022 + }, + { + "epoch": 1.5051428571428571, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 30023 + }, + { + "epoch": 1.505203007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 30024 + }, + { + "epoch": 1.5052631578947369, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 30025 + }, + { + "epoch": 1.5053233082706767, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2007, + "step": 30026 + }, + { + "epoch": 1.5053834586466164, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 30027 + }, + { + "epoch": 1.5054436090225565, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1464, + "step": 30028 + }, + { + "epoch": 1.5055037593984961, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 30029 + }, + { + "epoch": 1.5055639097744362, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 30030 + }, + { + "epoch": 1.5056240601503759, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 30031 + }, + { + "epoch": 1.5056842105263157, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 30032 + }, + { + "epoch": 1.5057443609022556, + "grad_norm": 1.484375, + "learning_rate": 3e-05, + "loss": 1.1956, + "step": 30033 + }, + { + "epoch": 1.5058045112781955, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 30034 + }, + { + "epoch": 1.5058646616541354, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 30035 + }, + { + "epoch": 1.5059248120300752, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 30036 + }, + { + "epoch": 1.505984962406015, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 30037 + }, + { + "epoch": 1.5060451127819547, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 30038 + }, + { + "epoch": 1.5061052631578948, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 30039 + }, + { + "epoch": 1.5061654135338345, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 30040 + }, + { + "epoch": 1.5062255639097746, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 30041 + }, + { + "epoch": 1.5062857142857142, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 30042 + }, + { + "epoch": 1.506345864661654, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 30043 + }, + { + "epoch": 1.506406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 30044 + }, + { + "epoch": 1.5064661654135338, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 30045 + }, + { + "epoch": 1.5065263157894737, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 30046 + }, + { + "epoch": 1.5065864661654136, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 30047 + }, + { + "epoch": 1.5066466165413535, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 30048 + }, + { + "epoch": 1.506706766917293, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 30049 + }, + { + "epoch": 1.5067669172932332, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 30050 + }, + { + "epoch": 1.5068270676691728, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 30051 + }, + { + "epoch": 1.506887218045113, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 30052 + }, + { + "epoch": 1.5069473684210526, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 30053 + }, + { + "epoch": 1.5070075187969925, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 30054 + }, + { + "epoch": 1.5070676691729323, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 30055 + }, + { + "epoch": 1.5071278195488722, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 30056 + }, + { + "epoch": 1.507187969924812, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 30057 + }, + { + "epoch": 1.5072481203007517, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 30058 + }, + { + "epoch": 1.5073082706766918, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 30059 + }, + { + "epoch": 1.5073684210526315, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 30060 + }, + { + "epoch": 1.5074285714285716, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 30061 + }, + { + "epoch": 1.5074887218045112, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 30062 + }, + { + "epoch": 1.507548872180451, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 30063 + }, + { + "epoch": 1.507609022556391, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 30064 + }, + { + "epoch": 1.5076691729323308, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 30065 + }, + { + "epoch": 1.5077293233082707, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 30066 + }, + { + "epoch": 1.5077894736842106, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 30067 + }, + { + "epoch": 1.5078496240601504, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 30068 + }, + { + "epoch": 1.50790977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 30069 + }, + { + "epoch": 1.5079699248120302, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 30070 + }, + { + "epoch": 1.5080300751879698, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 30071 + }, + { + "epoch": 1.50809022556391, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 30072 + }, + { + "epoch": 1.5081503759398496, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 30073 + }, + { + "epoch": 1.5082105263157894, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 30074 + }, + { + "epoch": 1.5082706766917293, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 30075 + }, + { + "epoch": 1.5083308270676692, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 30076 + }, + { + "epoch": 1.508390977443609, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 30077 + }, + { + "epoch": 1.5084511278195487, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 30078 + }, + { + "epoch": 1.5085112781954888, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 30079 + }, + { + "epoch": 1.5085714285714285, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 30080 + }, + { + "epoch": 1.5086315789473685, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 30081 + }, + { + "epoch": 1.5086917293233082, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 30082 + }, + { + "epoch": 1.508751879699248, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 30083 + }, + { + "epoch": 1.508812030075188, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 30084 + }, + { + "epoch": 1.5088721804511278, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 30085 + }, + { + "epoch": 1.5089323308270677, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 30086 + }, + { + "epoch": 1.5089924812030076, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 30087 + }, + { + "epoch": 1.5090526315789474, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 30088 + }, + { + "epoch": 1.509112781954887, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 30089 + }, + { + "epoch": 1.5091729323308272, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 30090 + }, + { + "epoch": 1.5092330827067668, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 30091 + }, + { + "epoch": 1.509293233082707, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 30092 + }, + { + "epoch": 1.5093533834586466, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 30093 + }, + { + "epoch": 1.5094135338345864, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 30094 + }, + { + "epoch": 1.5094736842105263, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 30095 + }, + { + "epoch": 1.5095338345864662, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 30096 + }, + { + "epoch": 1.509593984962406, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 30097 + }, + { + "epoch": 1.509654135338346, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 30098 + }, + { + "epoch": 1.5097142857142858, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 30099 + }, + { + "epoch": 1.5097744360902254, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 30100 + }, + { + "epoch": 1.5098345864661655, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 30101 + }, + { + "epoch": 1.5098947368421052, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 30102 + }, + { + "epoch": 1.5099548872180453, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 30103 + }, + { + "epoch": 1.510015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 30104 + }, + { + "epoch": 1.5100751879699248, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 30105 + }, + { + "epoch": 1.5101353383458647, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 30106 + }, + { + "epoch": 1.5101954887218045, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 30107 + }, + { + "epoch": 1.5102556390977444, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 30108 + }, + { + "epoch": 1.510315789473684, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 30109 + }, + { + "epoch": 1.5103759398496241, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 30110 + }, + { + "epoch": 1.5104360902255638, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9032, + "step": 30111 + }, + { + "epoch": 1.510496240601504, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 30112 + }, + { + "epoch": 1.5105563909774435, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 30113 + }, + { + "epoch": 1.5106165413533834, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2616, + "step": 30114 + }, + { + "epoch": 1.5106766917293233, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 30115 + }, + { + "epoch": 1.5107368421052632, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 30116 + }, + { + "epoch": 1.510796992481203, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 30117 + }, + { + "epoch": 1.510857142857143, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 30118 + }, + { + "epoch": 1.5109172932330828, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 30119 + }, + { + "epoch": 1.5109774436090224, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 30120 + }, + { + "epoch": 1.5110375939849625, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 30121 + }, + { + "epoch": 1.5110977443609022, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 30122 + }, + { + "epoch": 1.5111578947368423, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 30123 + }, + { + "epoch": 1.511218045112782, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 30124 + }, + { + "epoch": 1.5112781954887218, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 30125 + }, + { + "epoch": 1.5113383458646616, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 30126 + }, + { + "epoch": 1.5113984962406015, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 30127 + }, + { + "epoch": 1.5114586466165414, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 30128 + }, + { + "epoch": 1.511518796992481, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 30129 + }, + { + "epoch": 1.5115789473684211, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 30130 + }, + { + "epoch": 1.5116390977443608, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.2466, + "step": 30131 + }, + { + "epoch": 1.5116992481203009, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 30132 + }, + { + "epoch": 1.5117593984962405, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 30133 + }, + { + "epoch": 1.5118195488721804, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 30134 + }, + { + "epoch": 1.5118796992481203, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 30135 + }, + { + "epoch": 1.5119398496240601, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 30136 + }, + { + "epoch": 1.512, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 30137 + }, + { + "epoch": 1.5120601503759399, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 30138 + }, + { + "epoch": 1.5121203007518798, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 30139 + }, + { + "epoch": 1.5121804511278194, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 30140 + }, + { + "epoch": 1.5122406015037595, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1682, + "step": 30141 + }, + { + "epoch": 1.5123007518796991, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 30142 + }, + { + "epoch": 1.5123609022556392, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 30143 + }, + { + "epoch": 1.5124210526315789, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 30144 + }, + { + "epoch": 1.5124812030075188, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 30145 + }, + { + "epoch": 1.5125413533834586, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 30146 + }, + { + "epoch": 1.5126015037593985, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 30147 + }, + { + "epoch": 1.5126616541353384, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 30148 + }, + { + "epoch": 1.5127218045112782, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.3137, + "step": 30149 + }, + { + "epoch": 1.5127819548872181, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 30150 + }, + { + "epoch": 1.5128421052631578, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 30151 + }, + { + "epoch": 1.5129022556390979, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 30152 + }, + { + "epoch": 1.5129624060150375, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 30153 + }, + { + "epoch": 1.5130225563909776, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 30154 + }, + { + "epoch": 1.5130827067669173, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 30155 + }, + { + "epoch": 1.5131428571428571, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 30156 + }, + { + "epoch": 1.513203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 30157 + }, + { + "epoch": 1.5132631578947369, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 30158 + }, + { + "epoch": 1.5133233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9481, + "step": 30159 + }, + { + "epoch": 1.5133834586466164, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 30160 + }, + { + "epoch": 1.5134436090225565, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 30161 + }, + { + "epoch": 1.5135037593984961, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1664, + "step": 30162 + }, + { + "epoch": 1.5135639097744362, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 30163 + }, + { + "epoch": 1.5136240601503759, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 30164 + }, + { + "epoch": 1.5136842105263157, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 30165 + }, + { + "epoch": 1.5137443609022556, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 30166 + }, + { + "epoch": 1.5138045112781955, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 30167 + }, + { + "epoch": 1.5138646616541354, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 30168 + }, + { + "epoch": 1.5139248120300752, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 30169 + }, + { + "epoch": 1.513984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 30170 + }, + { + "epoch": 1.5140451127819547, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 30171 + }, + { + "epoch": 1.5141052631578948, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 30172 + }, + { + "epoch": 1.5141654135338345, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 30173 + }, + { + "epoch": 1.5142255639097746, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 30174 + }, + { + "epoch": 1.5142857142857142, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 30175 + }, + { + "epoch": 1.514345864661654, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 30176 + }, + { + "epoch": 1.514406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 30177 + }, + { + "epoch": 1.5144661654135338, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 30178 + }, + { + "epoch": 1.5145263157894737, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 30179 + }, + { + "epoch": 1.5145864661654134, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 30180 + }, + { + "epoch": 1.5146466165413535, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 30181 + }, + { + "epoch": 1.5147067669172931, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 30182 + }, + { + "epoch": 1.5147669172932332, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 30183 + }, + { + "epoch": 1.5148270676691729, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 30184 + }, + { + "epoch": 1.5148872180451127, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 30185 + }, + { + "epoch": 1.5149473684210526, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9239, + "step": 30186 + }, + { + "epoch": 1.5150075187969925, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 30187 + }, + { + "epoch": 1.5150676691729323, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 30188 + }, + { + "epoch": 1.5151278195488722, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 30189 + }, + { + "epoch": 1.515187969924812, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 30190 + }, + { + "epoch": 1.5152481203007517, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 30191 + }, + { + "epoch": 1.5153082706766918, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 30192 + }, + { + "epoch": 1.5153684210526315, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 30193 + }, + { + "epoch": 1.5154285714285716, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 30194 + }, + { + "epoch": 1.5154887218045112, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 30195 + }, + { + "epoch": 1.515548872180451, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 30196 + }, + { + "epoch": 1.515609022556391, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 30197 + }, + { + "epoch": 1.5156691729323308, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 30198 + }, + { + "epoch": 1.5157293233082707, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 30199 + }, + { + "epoch": 1.5157894736842106, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 30200 + }, + { + "epoch": 1.5158496240601504, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 30201 + }, + { + "epoch": 1.51590977443609, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 30202 + }, + { + "epoch": 1.5159699248120302, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1579, + "step": 30203 + }, + { + "epoch": 1.5160300751879698, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 30204 + }, + { + "epoch": 1.51609022556391, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 30205 + }, + { + "epoch": 1.5161503759398496, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 30206 + }, + { + "epoch": 1.5162105263157895, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 30207 + }, + { + "epoch": 1.5162706766917293, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 30208 + }, + { + "epoch": 1.5163308270676692, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 30209 + }, + { + "epoch": 1.516390977443609, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 30210 + }, + { + "epoch": 1.5164511278195487, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 30211 + }, + { + "epoch": 1.5165112781954888, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 30212 + }, + { + "epoch": 1.5165714285714285, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 30213 + }, + { + "epoch": 1.5166315789473686, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 30214 + }, + { + "epoch": 1.5166917293233082, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 30215 + }, + { + "epoch": 1.516751879699248, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 30216 + }, + { + "epoch": 1.516812030075188, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 30217 + }, + { + "epoch": 1.5168721804511278, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 30218 + }, + { + "epoch": 1.5169323308270677, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 30219 + }, + { + "epoch": 1.5169924812030076, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 30220 + }, + { + "epoch": 1.5170526315789474, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 30221 + }, + { + "epoch": 1.517112781954887, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 30222 + }, + { + "epoch": 1.5171729323308272, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 30223 + }, + { + "epoch": 1.5172330827067668, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 30224 + }, + { + "epoch": 1.517293233082707, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 30225 + }, + { + "epoch": 1.5173533834586466, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 30226 + }, + { + "epoch": 1.5174135338345864, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 30227 + }, + { + "epoch": 1.5174736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 30228 + }, + { + "epoch": 1.5175338345864662, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 30229 + }, + { + "epoch": 1.517593984962406, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 30230 + }, + { + "epoch": 1.5176541353383457, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 30231 + }, + { + "epoch": 1.5177142857142858, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 30232 + }, + { + "epoch": 1.5177744360902254, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 30233 + }, + { + "epoch": 1.5178345864661655, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 30234 + }, + { + "epoch": 1.5178947368421052, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 30235 + }, + { + "epoch": 1.5179548872180453, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 30236 + }, + { + "epoch": 1.518015037593985, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 30237 + }, + { + "epoch": 1.5180751879699248, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 30238 + }, + { + "epoch": 1.5181353383458647, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 30239 + }, + { + "epoch": 1.5181954887218045, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 30240 + }, + { + "epoch": 1.5182556390977444, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 30241 + }, + { + "epoch": 1.518315789473684, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 30242 + }, + { + "epoch": 1.5183759398496242, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 30243 + }, + { + "epoch": 1.5184360902255638, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 30244 + }, + { + "epoch": 1.518496240601504, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 30245 + }, + { + "epoch": 1.5185563909774435, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 30246 + }, + { + "epoch": 1.5186165413533834, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2306, + "step": 30247 + }, + { + "epoch": 1.5186766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 30248 + }, + { + "epoch": 1.5187368421052632, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 30249 + }, + { + "epoch": 1.518796992481203, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 30250 + }, + { + "epoch": 1.518857142857143, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 30251 + }, + { + "epoch": 1.5189172932330828, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 30252 + }, + { + "epoch": 1.5189774436090224, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 30253 + }, + { + "epoch": 1.5190375939849625, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 30254 + }, + { + "epoch": 1.5190977443609022, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 30255 + }, + { + "epoch": 1.5191578947368423, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 30256 + }, + { + "epoch": 1.519218045112782, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 30257 + }, + { + "epoch": 1.5192781954887218, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 30258 + }, + { + "epoch": 1.5193383458646617, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 30259 + }, + { + "epoch": 1.5193984962406015, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 30260 + }, + { + "epoch": 1.5194586466165414, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 30261 + }, + { + "epoch": 1.519518796992481, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 30262 + }, + { + "epoch": 1.5195789473684211, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 30263 + }, + { + "epoch": 1.5196390977443608, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 30264 + }, + { + "epoch": 1.5196992481203009, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 30265 + }, + { + "epoch": 1.5197593984962405, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 30266 + }, + { + "epoch": 1.5198195488721804, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 30267 + }, + { + "epoch": 1.5198796992481203, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 30268 + }, + { + "epoch": 1.5199398496240601, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 30269 + }, + { + "epoch": 1.52, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 30270 + }, + { + "epoch": 1.52006015037594, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 30271 + }, + { + "epoch": 1.5201203007518798, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 30272 + }, + { + "epoch": 1.5201804511278194, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 30273 + }, + { + "epoch": 1.5202406015037595, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 30274 + }, + { + "epoch": 1.5203007518796992, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 30275 + }, + { + "epoch": 1.5203609022556392, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 30276 + }, + { + "epoch": 1.520421052631579, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 30277 + }, + { + "epoch": 1.5204812030075188, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 30278 + }, + { + "epoch": 1.5205413533834586, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 30279 + }, + { + "epoch": 1.5206015037593985, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.189, + "step": 30280 + }, + { + "epoch": 1.5206616541353384, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 30281 + }, + { + "epoch": 1.520721804511278, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 30282 + }, + { + "epoch": 1.5207819548872181, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 30283 + }, + { + "epoch": 1.5208421052631578, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 30284 + }, + { + "epoch": 1.5209022556390979, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 30285 + }, + { + "epoch": 1.5209624060150375, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 30286 + }, + { + "epoch": 1.5210225563909776, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 30287 + }, + { + "epoch": 1.5210827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 30288 + }, + { + "epoch": 1.5211428571428571, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 30289 + }, + { + "epoch": 1.521203007518797, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 30290 + }, + { + "epoch": 1.5212631578947369, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 30291 + }, + { + "epoch": 1.5213233082706767, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 30292 + }, + { + "epoch": 1.5213834586466164, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 30293 + }, + { + "epoch": 1.5214436090225565, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 30294 + }, + { + "epoch": 1.5215037593984961, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 30295 + }, + { + "epoch": 1.5215639097744362, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 30296 + }, + { + "epoch": 1.5216240601503759, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 30297 + }, + { + "epoch": 1.5216842105263158, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 30298 + }, + { + "epoch": 1.5217443609022556, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 30299 + }, + { + "epoch": 1.5218045112781955, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 30300 + }, + { + "epoch": 1.5218646616541354, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 30301 + }, + { + "epoch": 1.5219248120300752, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 30302 + }, + { + "epoch": 1.521984962406015, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 30303 + }, + { + "epoch": 1.5220451127819548, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 30304 + }, + { + "epoch": 1.5221052631578948, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 30305 + }, + { + "epoch": 1.5221654135338345, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.2375, + "step": 30306 + }, + { + "epoch": 1.5222255639097746, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 30307 + }, + { + "epoch": 1.5222857142857142, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 30308 + }, + { + "epoch": 1.5223458646616541, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 30309 + }, + { + "epoch": 1.522406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 30310 + }, + { + "epoch": 1.5224661654135339, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 30311 + }, + { + "epoch": 1.5225263157894737, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 30312 + }, + { + "epoch": 1.5225864661654134, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 30313 + }, + { + "epoch": 1.5226466165413535, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 30314 + }, + { + "epoch": 1.5227067669172931, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 30315 + }, + { + "epoch": 1.5227669172932332, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9342, + "step": 30316 + }, + { + "epoch": 1.5228270676691729, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 30317 + }, + { + "epoch": 1.5228872180451127, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 30318 + }, + { + "epoch": 1.5229473684210526, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 30319 + }, + { + "epoch": 1.5230075187969925, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 30320 + }, + { + "epoch": 1.5230676691729323, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 30321 + }, + { + "epoch": 1.5231278195488722, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 30322 + }, + { + "epoch": 1.523187969924812, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 30323 + }, + { + "epoch": 1.5232481203007517, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 30324 + }, + { + "epoch": 1.5233082706766918, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 30325 + }, + { + "epoch": 1.5233684210526315, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 30326 + }, + { + "epoch": 1.5234285714285716, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 30327 + }, + { + "epoch": 1.5234887218045112, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 30328 + }, + { + "epoch": 1.523548872180451, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 30329 + }, + { + "epoch": 1.523609022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 30330 + }, + { + "epoch": 1.5236691729323308, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 30331 + }, + { + "epoch": 1.5237293233082707, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 30332 + }, + { + "epoch": 1.5237894736842104, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 30333 + }, + { + "epoch": 1.5238496240601505, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 30334 + }, + { + "epoch": 1.52390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 30335 + }, + { + "epoch": 1.5239699248120302, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 30336 + }, + { + "epoch": 1.5240300751879698, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1629, + "step": 30337 + }, + { + "epoch": 1.52409022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 30338 + }, + { + "epoch": 1.5241503759398496, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 30339 + }, + { + "epoch": 1.5242105263157895, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 30340 + }, + { + "epoch": 1.5242706766917293, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 30341 + }, + { + "epoch": 1.5243308270676692, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 30342 + }, + { + "epoch": 1.524390977443609, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 30343 + }, + { + "epoch": 1.5244511278195487, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 30344 + }, + { + "epoch": 1.5245112781954888, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9396, + "step": 30345 + }, + { + "epoch": 1.5245714285714285, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 30346 + }, + { + "epoch": 1.5246315789473686, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 30347 + }, + { + "epoch": 1.5246917293233082, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 30348 + }, + { + "epoch": 1.524751879699248, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 30349 + }, + { + "epoch": 1.524812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 30350 + }, + { + "epoch": 1.5248721804511278, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 30351 + }, + { + "epoch": 1.5249323308270677, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 30352 + }, + { + "epoch": 1.5249924812030076, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 30353 + }, + { + "epoch": 1.5250526315789474, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 30354 + }, + { + "epoch": 1.525112781954887, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 30355 + }, + { + "epoch": 1.5251729323308272, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 30356 + }, + { + "epoch": 1.5252330827067668, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 30357 + }, + { + "epoch": 1.525293233082707, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1933, + "step": 30358 + }, + { + "epoch": 1.5253533834586466, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 30359 + }, + { + "epoch": 1.5254135338345864, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 30360 + }, + { + "epoch": 1.5254736842105263, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 30361 + }, + { + "epoch": 1.5255338345864662, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 30362 + }, + { + "epoch": 1.525593984962406, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 30363 + }, + { + "epoch": 1.5256541353383457, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 30364 + }, + { + "epoch": 1.5257142857142858, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 30365 + }, + { + "epoch": 1.5257744360902254, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 30366 + }, + { + "epoch": 1.5258345864661655, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 30367 + }, + { + "epoch": 1.5258947368421052, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 30368 + }, + { + "epoch": 1.525954887218045, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 30369 + }, + { + "epoch": 1.526015037593985, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 30370 + }, + { + "epoch": 1.5260751879699248, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1737, + "step": 30371 + }, + { + "epoch": 1.5261353383458647, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 30372 + }, + { + "epoch": 1.5261954887218045, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 30373 + }, + { + "epoch": 1.5262556390977444, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 30374 + }, + { + "epoch": 1.526315789473684, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 30375 + }, + { + "epoch": 1.5263759398496242, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 30376 + }, + { + "epoch": 1.5264360902255638, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 30377 + }, + { + "epoch": 1.526496240601504, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 30378 + }, + { + "epoch": 1.5265563909774436, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 30379 + }, + { + "epoch": 1.5266165413533834, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 30380 + }, + { + "epoch": 1.5266766917293233, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 30381 + }, + { + "epoch": 1.5267368421052632, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 30382 + }, + { + "epoch": 1.526796992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 30383 + }, + { + "epoch": 1.5268571428571427, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.174, + "step": 30384 + }, + { + "epoch": 1.5269172932330828, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 30385 + }, + { + "epoch": 1.5269774436090224, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 30386 + }, + { + "epoch": 1.5270375939849625, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 30387 + }, + { + "epoch": 1.5270977443609022, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 30388 + }, + { + "epoch": 1.5271578947368423, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 30389 + }, + { + "epoch": 1.527218045112782, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 30390 + }, + { + "epoch": 1.5272781954887218, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 30391 + }, + { + "epoch": 1.5273383458646617, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 30392 + }, + { + "epoch": 1.5273984962406015, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 30393 + }, + { + "epoch": 1.5274586466165414, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 30394 + }, + { + "epoch": 1.527518796992481, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 30395 + }, + { + "epoch": 1.5275789473684211, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 30396 + }, + { + "epoch": 1.5276390977443608, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 30397 + }, + { + "epoch": 1.527699248120301, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 30398 + }, + { + "epoch": 1.5277593984962405, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 30399 + }, + { + "epoch": 1.5278195488721804, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 30400 + }, + { + "epoch": 1.5278796992481203, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 30401 + }, + { + "epoch": 1.5279398496240602, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 30402 + }, + { + "epoch": 1.528, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 30403 + }, + { + "epoch": 1.52806015037594, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 30404 + }, + { + "epoch": 1.5281203007518798, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 30405 + }, + { + "epoch": 1.5281804511278194, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 30406 + }, + { + "epoch": 1.5282406015037595, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 30407 + }, + { + "epoch": 1.5283007518796992, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 30408 + }, + { + "epoch": 1.5283609022556393, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 30409 + }, + { + "epoch": 1.528421052631579, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 30410 + }, + { + "epoch": 1.5284812030075188, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 30411 + }, + { + "epoch": 1.5285413533834586, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 30412 + }, + { + "epoch": 1.5286015037593985, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 30413 + }, + { + "epoch": 1.5286616541353384, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 30414 + }, + { + "epoch": 1.528721804511278, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.2404, + "step": 30415 + }, + { + "epoch": 1.5287819548872181, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 30416 + }, + { + "epoch": 1.5288421052631578, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 30417 + }, + { + "epoch": 1.5289022556390979, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 30418 + }, + { + "epoch": 1.5289624060150375, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 30419 + }, + { + "epoch": 1.5290225563909774, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 30420 + }, + { + "epoch": 1.5290827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1863, + "step": 30421 + }, + { + "epoch": 1.5291428571428571, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 30422 + }, + { + "epoch": 1.529203007518797, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 30423 + }, + { + "epoch": 1.5292631578947369, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 30424 + }, + { + "epoch": 1.5293233082706768, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 30425 + }, + { + "epoch": 1.5293834586466164, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 30426 + }, + { + "epoch": 1.5294436090225565, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 30427 + }, + { + "epoch": 1.5295037593984961, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 30428 + }, + { + "epoch": 1.5295639097744362, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 30429 + }, + { + "epoch": 1.5296240601503759, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 30430 + }, + { + "epoch": 1.5296842105263158, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 30431 + }, + { + "epoch": 1.5297443609022556, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 30432 + }, + { + "epoch": 1.5298045112781955, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 30433 + }, + { + "epoch": 1.5298646616541354, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 30434 + }, + { + "epoch": 1.5299248120300752, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 30435 + }, + { + "epoch": 1.5299849624060151, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 30436 + }, + { + "epoch": 1.5300451127819548, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 30437 + }, + { + "epoch": 1.5301052631578949, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 30438 + }, + { + "epoch": 1.5301654135338345, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 30439 + }, + { + "epoch": 1.5302255639097746, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 30440 + }, + { + "epoch": 1.5302857142857142, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 30441 + }, + { + "epoch": 1.5303458646616541, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 30442 + }, + { + "epoch": 1.530406015037594, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 30443 + }, + { + "epoch": 1.5304661654135339, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 30444 + }, + { + "epoch": 1.5305263157894737, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 30445 + }, + { + "epoch": 1.5305864661654134, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1241, + "step": 30446 + }, + { + "epoch": 1.5306466165413535, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 30447 + }, + { + "epoch": 1.5307067669172931, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 30448 + }, + { + "epoch": 1.5307669172932332, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 30449 + }, + { + "epoch": 1.5308270676691729, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 30450 + }, + { + "epoch": 1.5308872180451127, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 30451 + }, + { + "epoch": 1.5309473684210526, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 30452 + }, + { + "epoch": 1.5310075187969925, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 30453 + }, + { + "epoch": 1.5310676691729324, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 30454 + }, + { + "epoch": 1.5311278195488722, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 30455 + }, + { + "epoch": 1.531187969924812, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1698, + "step": 30456 + }, + { + "epoch": 1.5312481203007517, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 30457 + }, + { + "epoch": 1.5313082706766918, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 30458 + }, + { + "epoch": 1.5313684210526315, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 30459 + }, + { + "epoch": 1.5314285714285716, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 30460 + }, + { + "epoch": 1.5314887218045112, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 30461 + }, + { + "epoch": 1.531548872180451, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 30462 + }, + { + "epoch": 1.531609022556391, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 30463 + }, + { + "epoch": 1.5316691729323308, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 30464 + }, + { + "epoch": 1.5317293233082707, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 30465 + }, + { + "epoch": 1.5317894736842104, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 30466 + }, + { + "epoch": 1.5318496240601505, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 30467 + }, + { + "epoch": 1.53190977443609, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 30468 + }, + { + "epoch": 1.5319699248120302, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 30469 + }, + { + "epoch": 1.5320300751879699, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 30470 + }, + { + "epoch": 1.5320902255639097, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 30471 + }, + { + "epoch": 1.5321503759398496, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 30472 + }, + { + "epoch": 1.5322105263157895, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 30473 + }, + { + "epoch": 1.5322706766917293, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 30474 + }, + { + "epoch": 1.5323308270676692, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 30475 + }, + { + "epoch": 1.532390977443609, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 30476 + }, + { + "epoch": 1.5324511278195487, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 30477 + }, + { + "epoch": 1.5325112781954888, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 30478 + }, + { + "epoch": 1.5325714285714285, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.169, + "step": 30479 + }, + { + "epoch": 1.5326315789473686, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 30480 + }, + { + "epoch": 1.5326917293233082, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 30481 + }, + { + "epoch": 1.532751879699248, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 30482 + }, + { + "epoch": 1.532812030075188, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 30483 + }, + { + "epoch": 1.5328721804511278, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 30484 + }, + { + "epoch": 1.5329323308270677, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 30485 + }, + { + "epoch": 1.5329924812030076, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 30486 + }, + { + "epoch": 1.5330526315789474, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 30487 + }, + { + "epoch": 1.533112781954887, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 30488 + }, + { + "epoch": 1.5331729323308272, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 30489 + }, + { + "epoch": 1.5332330827067668, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 30490 + }, + { + "epoch": 1.533293233082707, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 30491 + }, + { + "epoch": 1.5333533834586466, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 30492 + }, + { + "epoch": 1.5334135338345865, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 30493 + }, + { + "epoch": 1.5334736842105263, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 30494 + }, + { + "epoch": 1.5335338345864662, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 30495 + }, + { + "epoch": 1.533593984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 30496 + }, + { + "epoch": 1.5336541353383457, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 30497 + }, + { + "epoch": 1.5337142857142858, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 30498 + }, + { + "epoch": 1.5337744360902255, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 30499 + }, + { + "epoch": 1.5338345864661656, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 30500 + }, + { + "epoch": 1.5338947368421052, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 30501 + }, + { + "epoch": 1.533954887218045, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 30502 + }, + { + "epoch": 1.534015037593985, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 30503 + }, + { + "epoch": 1.5340751879699248, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 30504 + }, + { + "epoch": 1.5341353383458647, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 30505 + }, + { + "epoch": 1.5341954887218046, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 30506 + }, + { + "epoch": 1.5342556390977444, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 30507 + }, + { + "epoch": 1.534315789473684, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 30508 + }, + { + "epoch": 1.5343759398496242, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 30509 + }, + { + "epoch": 1.5344360902255638, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 30510 + }, + { + "epoch": 1.534496240601504, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 30511 + }, + { + "epoch": 1.5345563909774436, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 30512 + }, + { + "epoch": 1.5346165413533834, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 30513 + }, + { + "epoch": 1.5346766917293233, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 30514 + }, + { + "epoch": 1.5347368421052632, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 30515 + }, + { + "epoch": 1.534796992481203, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 30516 + }, + { + "epoch": 1.5348571428571427, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 30517 + }, + { + "epoch": 1.5349172932330828, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 30518 + }, + { + "epoch": 1.5349774436090224, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 30519 + }, + { + "epoch": 1.5350375939849625, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 30520 + }, + { + "epoch": 1.5350977443609022, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 30521 + }, + { + "epoch": 1.535157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 30522 + }, + { + "epoch": 1.535218045112782, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 30523 + }, + { + "epoch": 1.5352781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 30524 + }, + { + "epoch": 1.5353383458646617, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 30525 + }, + { + "epoch": 1.5353984962406015, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 30526 + }, + { + "epoch": 1.5354586466165414, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 30527 + }, + { + "epoch": 1.535518796992481, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 30528 + }, + { + "epoch": 1.5355789473684212, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 30529 + }, + { + "epoch": 1.5356390977443608, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 30530 + }, + { + "epoch": 1.535699248120301, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 30531 + }, + { + "epoch": 1.5357593984962405, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 30532 + }, + { + "epoch": 1.5358195488721804, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 30533 + }, + { + "epoch": 1.5358796992481203, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 30534 + }, + { + "epoch": 1.5359398496240602, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 30535 + }, + { + "epoch": 1.536, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 30536 + }, + { + "epoch": 1.53606015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 30537 + }, + { + "epoch": 1.5361203007518798, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 30538 + }, + { + "epoch": 1.5361804511278194, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 30539 + }, + { + "epoch": 1.5362406015037595, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 30540 + }, + { + "epoch": 1.5363007518796992, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 30541 + }, + { + "epoch": 1.5363609022556393, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 30542 + }, + { + "epoch": 1.536421052631579, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 30543 + }, + { + "epoch": 1.5364812030075188, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 30544 + }, + { + "epoch": 1.5365413533834587, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 30545 + }, + { + "epoch": 1.5366015037593985, + "grad_norm": 1.625, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 30546 + }, + { + "epoch": 1.5366616541353384, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 30547 + }, + { + "epoch": 1.536721804511278, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 30548 + }, + { + "epoch": 1.5367819548872181, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 30549 + }, + { + "epoch": 1.5368421052631578, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.8662, + "step": 30550 + }, + { + "epoch": 1.5369022556390979, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 30551 + }, + { + "epoch": 1.5369624060150375, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.175, + "step": 30552 + }, + { + "epoch": 1.5370225563909774, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 30553 + }, + { + "epoch": 1.5370827067669173, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 30554 + }, + { + "epoch": 1.5371428571428571, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 30555 + }, + { + "epoch": 1.537203007518797, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9269, + "step": 30556 + }, + { + "epoch": 1.5372631578947369, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 30557 + }, + { + "epoch": 1.5373233082706768, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 30558 + }, + { + "epoch": 1.5373834586466164, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 30559 + }, + { + "epoch": 1.5374436090225565, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 30560 + }, + { + "epoch": 1.5375037593984962, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 30561 + }, + { + "epoch": 1.5375639097744362, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 30562 + }, + { + "epoch": 1.537624060150376, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 30563 + }, + { + "epoch": 1.5376842105263158, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 30564 + }, + { + "epoch": 1.5377443609022556, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 30565 + }, + { + "epoch": 1.5378045112781955, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 30566 + }, + { + "epoch": 1.5378646616541354, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 30567 + }, + { + "epoch": 1.537924812030075, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 30568 + }, + { + "epoch": 1.5379849624060151, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 30569 + }, + { + "epoch": 1.5380451127819548, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 30570 + }, + { + "epoch": 1.5381052631578949, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 30571 + }, + { + "epoch": 1.5381654135338345, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1805, + "step": 30572 + }, + { + "epoch": 1.5382255639097744, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 30573 + }, + { + "epoch": 1.5382857142857143, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 30574 + }, + { + "epoch": 1.5383458646616541, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 30575 + }, + { + "epoch": 1.538406015037594, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 30576 + }, + { + "epoch": 1.5384661654135339, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 30577 + }, + { + "epoch": 1.5385263157894737, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 30578 + }, + { + "epoch": 1.5385864661654134, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 30579 + }, + { + "epoch": 1.5386466165413535, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 30580 + }, + { + "epoch": 1.5387067669172931, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 30581 + }, + { + "epoch": 1.5387669172932332, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 30582 + }, + { + "epoch": 1.5388270676691729, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 30583 + }, + { + "epoch": 1.5388872180451127, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 30584 + }, + { + "epoch": 1.5389473684210526, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 30585 + }, + { + "epoch": 1.5390075187969925, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 30586 + }, + { + "epoch": 1.5390676691729324, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 30587 + }, + { + "epoch": 1.5391278195488722, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 30588 + }, + { + "epoch": 1.539187969924812, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 30589 + }, + { + "epoch": 1.5392481203007518, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 30590 + }, + { + "epoch": 1.5393082706766918, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 30591 + }, + { + "epoch": 1.5393684210526315, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 30592 + }, + { + "epoch": 1.5394285714285716, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 30593 + }, + { + "epoch": 1.5394887218045112, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 30594 + }, + { + "epoch": 1.5395488721804511, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 30595 + }, + { + "epoch": 1.539609022556391, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 30596 + }, + { + "epoch": 1.5396691729323309, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 30597 + }, + { + "epoch": 1.5397293233082707, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 30598 + }, + { + "epoch": 1.5397894736842104, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 30599 + }, + { + "epoch": 1.5398496240601505, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 30600 + }, + { + "epoch": 1.5399097744360901, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 30601 + }, + { + "epoch": 1.5399699248120302, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9394, + "step": 30602 + }, + { + "epoch": 1.5400300751879699, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 30603 + }, + { + "epoch": 1.5400902255639097, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 30604 + }, + { + "epoch": 1.5401503759398496, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 30605 + }, + { + "epoch": 1.5402105263157895, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 30606 + }, + { + "epoch": 1.5402706766917293, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 30607 + }, + { + "epoch": 1.5403308270676692, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 30608 + }, + { + "epoch": 1.540390977443609, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 30609 + }, + { + "epoch": 1.5404511278195487, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 30610 + }, + { + "epoch": 1.5405112781954888, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 30611 + }, + { + "epoch": 1.5405714285714285, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 30612 + }, + { + "epoch": 1.5406315789473686, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 30613 + }, + { + "epoch": 1.5406917293233082, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 30614 + }, + { + "epoch": 1.540751879699248, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 30615 + }, + { + "epoch": 1.540812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 30616 + }, + { + "epoch": 1.5408721804511278, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 30617 + }, + { + "epoch": 1.5409323308270677, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 30618 + }, + { + "epoch": 1.5409924812030074, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 30619 + }, + { + "epoch": 1.5410526315789475, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 30620 + }, + { + "epoch": 1.541112781954887, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 30621 + }, + { + "epoch": 1.5411729323308272, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 30622 + }, + { + "epoch": 1.5412330827067668, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 30623 + }, + { + "epoch": 1.5412932330827067, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 30624 + }, + { + "epoch": 1.5413533834586466, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 30625 + }, + { + "epoch": 1.5414135338345865, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 30626 + }, + { + "epoch": 1.5414736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 30627 + }, + { + "epoch": 1.5415338345864662, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 30628 + }, + { + "epoch": 1.541593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 30629 + }, + { + "epoch": 1.5416541353383457, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 30630 + }, + { + "epoch": 1.5417142857142858, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 30631 + }, + { + "epoch": 1.5417744360902255, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 30632 + }, + { + "epoch": 1.5418345864661656, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 30633 + }, + { + "epoch": 1.5418947368421052, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.2808, + "step": 30634 + }, + { + "epoch": 1.541954887218045, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 30635 + }, + { + "epoch": 1.542015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 30636 + }, + { + "epoch": 1.5420751879699248, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 30637 + }, + { + "epoch": 1.5421353383458647, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 30638 + }, + { + "epoch": 1.5421954887218046, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 30639 + }, + { + "epoch": 1.5422556390977444, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 30640 + }, + { + "epoch": 1.542315789473684, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 30641 + }, + { + "epoch": 1.5423759398496242, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 30642 + }, + { + "epoch": 1.5424360902255638, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 30643 + }, + { + "epoch": 1.542496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 30644 + }, + { + "epoch": 1.5425563909774436, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 30645 + }, + { + "epoch": 1.5426165413533834, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 30646 + }, + { + "epoch": 1.5426766917293233, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 30647 + }, + { + "epoch": 1.5427368421052632, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 30648 + }, + { + "epoch": 1.542796992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 30649 + }, + { + "epoch": 1.5428571428571427, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 30650 + }, + { + "epoch": 1.5429172932330828, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 30651 + }, + { + "epoch": 1.5429774436090224, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 30652 + }, + { + "epoch": 1.5430375939849625, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 30653 + }, + { + "epoch": 1.5430977443609022, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 30654 + }, + { + "epoch": 1.543157894736842, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 30655 + }, + { + "epoch": 1.543218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 30656 + }, + { + "epoch": 1.5432781954887218, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 30657 + }, + { + "epoch": 1.5433383458646617, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 30658 + }, + { + "epoch": 1.5433984962406015, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 30659 + }, + { + "epoch": 1.5434586466165414, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 30660 + }, + { + "epoch": 1.543518796992481, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 30661 + }, + { + "epoch": 1.5435789473684212, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 30662 + }, + { + "epoch": 1.5436390977443608, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 30663 + }, + { + "epoch": 1.543699248120301, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 30664 + }, + { + "epoch": 1.5437593984962406, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 30665 + }, + { + "epoch": 1.5438195488721804, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 30666 + }, + { + "epoch": 1.5438796992481203, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 30667 + }, + { + "epoch": 1.5439398496240602, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 30668 + }, + { + "epoch": 1.544, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 30669 + }, + { + "epoch": 1.5440601503759397, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 30670 + }, + { + "epoch": 1.5441203007518798, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 30671 + }, + { + "epoch": 1.5441804511278194, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.3723, + "step": 30672 + }, + { + "epoch": 1.5442406015037595, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 30673 + }, + { + "epoch": 1.5443007518796992, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 30674 + }, + { + "epoch": 1.544360902255639, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 30675 + }, + { + "epoch": 1.544421052631579, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 30676 + }, + { + "epoch": 1.5444812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1717, + "step": 30677 + }, + { + "epoch": 1.5445413533834587, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 30678 + }, + { + "epoch": 1.5446015037593985, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 30679 + }, + { + "epoch": 1.5446616541353384, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9268, + "step": 30680 + }, + { + "epoch": 1.544721804511278, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 30681 + }, + { + "epoch": 1.5447819548872181, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 30682 + }, + { + "epoch": 1.5448421052631578, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 30683 + }, + { + "epoch": 1.5449022556390979, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 30684 + }, + { + "epoch": 1.5449624060150375, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 30685 + }, + { + "epoch": 1.5450225563909774, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 30686 + }, + { + "epoch": 1.5450827067669173, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 30687 + }, + { + "epoch": 1.5451428571428572, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2177, + "step": 30688 + }, + { + "epoch": 1.545203007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 30689 + }, + { + "epoch": 1.545263157894737, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 30690 + }, + { + "epoch": 1.5453233082706768, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 30691 + }, + { + "epoch": 1.5453834586466164, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 30692 + }, + { + "epoch": 1.5454436090225565, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 30693 + }, + { + "epoch": 1.5455037593984962, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 30694 + }, + { + "epoch": 1.5455639097744363, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 30695 + }, + { + "epoch": 1.545624060150376, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 30696 + }, + { + "epoch": 1.5456842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 30697 + }, + { + "epoch": 1.5457443609022556, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 30698 + }, + { + "epoch": 1.5458045112781955, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 30699 + }, + { + "epoch": 1.5458646616541354, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 30700 + }, + { + "epoch": 1.545924812030075, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 30701 + }, + { + "epoch": 1.5459849624060151, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 30702 + }, + { + "epoch": 1.5460451127819548, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 30703 + }, + { + "epoch": 1.5461052631578949, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 30704 + }, + { + "epoch": 1.5461654135338345, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 30705 + }, + { + "epoch": 1.5462255639097744, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 30706 + }, + { + "epoch": 1.5462857142857143, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 30707 + }, + { + "epoch": 1.5463458646616541, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 30708 + }, + { + "epoch": 1.546406015037594, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 30709 + }, + { + "epoch": 1.5464661654135339, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 30710 + }, + { + "epoch": 1.5465263157894737, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 30711 + }, + { + "epoch": 1.5465864661654134, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 30712 + }, + { + "epoch": 1.5466466165413535, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 30713 + }, + { + "epoch": 1.5467067669172931, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 30714 + }, + { + "epoch": 1.5467669172932332, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 30715 + }, + { + "epoch": 1.5468270676691729, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 30716 + }, + { + "epoch": 1.5468872180451128, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 30717 + }, + { + "epoch": 1.5469473684210526, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 30718 + }, + { + "epoch": 1.5470075187969925, + "grad_norm": 1.4765625, + "learning_rate": 3e-05, + "loss": 1.1395, + "step": 30719 + }, + { + "epoch": 1.5470676691729324, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 30720 + }, + { + "epoch": 1.547127819548872, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 30721 + }, + { + "epoch": 1.5471879699248121, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 30722 + }, + { + "epoch": 1.5472481203007518, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 30723 + }, + { + "epoch": 1.5473082706766919, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 30724 + }, + { + "epoch": 1.5473684210526315, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 30725 + }, + { + "epoch": 1.5474285714285714, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 30726 + }, + { + "epoch": 1.5474887218045112, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 30727 + }, + { + "epoch": 1.5475488721804511, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 30728 + }, + { + "epoch": 1.547609022556391, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 30729 + }, + { + "epoch": 1.5476691729323309, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 30730 + }, + { + "epoch": 1.5477293233082707, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 30731 + }, + { + "epoch": 1.5477894736842104, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 30732 + }, + { + "epoch": 1.5478496240601505, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 30733 + }, + { + "epoch": 1.5479097744360901, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 30734 + }, + { + "epoch": 1.5479699248120302, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 30735 + }, + { + "epoch": 1.5480300751879699, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 30736 + }, + { + "epoch": 1.5480902255639097, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 30737 + }, + { + "epoch": 1.5481503759398496, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 30738 + }, + { + "epoch": 1.5482105263157895, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 30739 + }, + { + "epoch": 1.5482706766917294, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 30740 + }, + { + "epoch": 1.5483308270676692, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 30741 + }, + { + "epoch": 1.548390977443609, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 30742 + }, + { + "epoch": 1.5484511278195487, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 30743 + }, + { + "epoch": 1.5485112781954888, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 30744 + }, + { + "epoch": 1.5485714285714285, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 30745 + }, + { + "epoch": 1.5486315789473686, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 30746 + }, + { + "epoch": 1.5486917293233082, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 30747 + }, + { + "epoch": 1.548751879699248, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 30748 + }, + { + "epoch": 1.548812030075188, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 30749 + }, + { + "epoch": 1.5488721804511278, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 30750 + }, + { + "epoch": 1.5489323308270677, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 30751 + }, + { + "epoch": 1.5489924812030074, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 30752 + }, + { + "epoch": 1.5490526315789475, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 30753 + }, + { + "epoch": 1.549112781954887, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 30754 + }, + { + "epoch": 1.5491729323308272, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 30755 + }, + { + "epoch": 1.5492330827067669, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 30756 + }, + { + "epoch": 1.5492932330827067, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 30757 + }, + { + "epoch": 1.5493533834586466, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 30758 + }, + { + "epoch": 1.5494135338345865, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 30759 + }, + { + "epoch": 1.5494736842105263, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 30760 + }, + { + "epoch": 1.5495338345864662, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 30761 + }, + { + "epoch": 1.549593984962406, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 30762 + }, + { + "epoch": 1.5496541353383457, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 30763 + }, + { + "epoch": 1.5497142857142858, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 30764 + }, + { + "epoch": 1.5497744360902255, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 30765 + }, + { + "epoch": 1.5498345864661656, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 30766 + }, + { + "epoch": 1.5498947368421052, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 30767 + }, + { + "epoch": 1.549954887218045, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 30768 + }, + { + "epoch": 1.550015037593985, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 30769 + }, + { + "epoch": 1.5500751879699248, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 30770 + }, + { + "epoch": 1.5501353383458647, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 30771 + }, + { + "epoch": 1.5501954887218043, + "grad_norm": 1.4375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 30772 + }, + { + "epoch": 1.5502556390977444, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 30773 + }, + { + "epoch": 1.550315789473684, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 30774 + }, + { + "epoch": 1.5503759398496242, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 30775 + }, + { + "epoch": 1.5504360902255638, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 30776 + }, + { + "epoch": 1.5504962406015037, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 30777 + }, + { + "epoch": 1.5505563909774436, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 30778 + }, + { + "epoch": 1.5506165413533834, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 30779 + }, + { + "epoch": 1.5506766917293233, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 30780 + }, + { + "epoch": 1.5507368421052632, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 30781 + }, + { + "epoch": 1.550796992481203, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 30782 + }, + { + "epoch": 1.5508571428571427, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 30783 + }, + { + "epoch": 1.5509172932330828, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 30784 + }, + { + "epoch": 1.5509774436090225, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 30785 + }, + { + "epoch": 1.5510375939849625, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 30786 + }, + { + "epoch": 1.5510977443609022, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 30787 + }, + { + "epoch": 1.551157894736842, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 30788 + }, + { + "epoch": 1.551218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 30789 + }, + { + "epoch": 1.5512781954887218, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 30790 + }, + { + "epoch": 1.5513383458646617, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 30791 + }, + { + "epoch": 1.5513984962406016, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 30792 + }, + { + "epoch": 1.5514586466165414, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 30793 + }, + { + "epoch": 1.551518796992481, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 30794 + }, + { + "epoch": 1.5515789473684212, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 30795 + }, + { + "epoch": 1.5516390977443608, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 30796 + }, + { + "epoch": 1.551699248120301, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 30797 + }, + { + "epoch": 1.5517593984962406, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 30798 + }, + { + "epoch": 1.5518195488721804, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 30799 + }, + { + "epoch": 1.5518796992481203, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 30800 + }, + { + "epoch": 1.5519398496240602, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 30801 + }, + { + "epoch": 1.552, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 30802 + }, + { + "epoch": 1.5520601503759397, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 30803 + }, + { + "epoch": 1.5521203007518798, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 30804 + }, + { + "epoch": 1.5521804511278194, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 30805 + }, + { + "epoch": 1.5522406015037595, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1559, + "step": 30806 + }, + { + "epoch": 1.5523007518796992, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 30807 + }, + { + "epoch": 1.552360902255639, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 30808 + }, + { + "epoch": 1.552421052631579, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 30809 + }, + { + "epoch": 1.5524812030075188, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 30810 + }, + { + "epoch": 1.5525413533834587, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 30811 + }, + { + "epoch": 1.5526015037593985, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 30812 + }, + { + "epoch": 1.5526616541353384, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 30813 + }, + { + "epoch": 1.552721804511278, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 30814 + }, + { + "epoch": 1.5527819548872182, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 30815 + }, + { + "epoch": 1.5528421052631578, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1908, + "step": 30816 + }, + { + "epoch": 1.552902255639098, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 30817 + }, + { + "epoch": 1.5529624060150375, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 30818 + }, + { + "epoch": 1.5530225563909774, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 30819 + }, + { + "epoch": 1.5530827067669173, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 30820 + }, + { + "epoch": 1.5531428571428572, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 30821 + }, + { + "epoch": 1.553203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 30822 + }, + { + "epoch": 1.5532631578947367, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 30823 + }, + { + "epoch": 1.5533233082706768, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 30824 + }, + { + "epoch": 1.5533834586466164, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 30825 + }, + { + "epoch": 1.5534436090225565, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 30826 + }, + { + "epoch": 1.5535037593984962, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 30827 + }, + { + "epoch": 1.553563909774436, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 30828 + }, + { + "epoch": 1.553624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 30829 + }, + { + "epoch": 1.5536842105263158, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 30830 + }, + { + "epoch": 1.5537443609022557, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 30831 + }, + { + "epoch": 1.5538045112781955, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 30832 + }, + { + "epoch": 1.5538646616541354, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 30833 + }, + { + "epoch": 1.553924812030075, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 30834 + }, + { + "epoch": 1.5539849624060151, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 30835 + }, + { + "epoch": 1.5540451127819548, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 30836 + }, + { + "epoch": 1.5541052631578949, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 30837 + }, + { + "epoch": 1.5541654135338345, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 30838 + }, + { + "epoch": 1.5542255639097744, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 30839 + }, + { + "epoch": 1.5542857142857143, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 30840 + }, + { + "epoch": 1.5543458646616541, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 30841 + }, + { + "epoch": 1.554406015037594, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 30842 + }, + { + "epoch": 1.5544661654135339, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 30843 + }, + { + "epoch": 1.5545263157894738, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 30844 + }, + { + "epoch": 1.5545864661654134, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 30845 + }, + { + "epoch": 1.5546466165413535, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 30846 + }, + { + "epoch": 1.5547067669172931, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 30847 + }, + { + "epoch": 1.5547669172932332, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 30848 + }, + { + "epoch": 1.554827067669173, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 30849 + }, + { + "epoch": 1.5548872180451128, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 30850 + }, + { + "epoch": 1.5549473684210526, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 30851 + }, + { + "epoch": 1.5550075187969925, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 30852 + }, + { + "epoch": 1.5550676691729324, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 30853 + }, + { + "epoch": 1.555127819548872, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 30854 + }, + { + "epoch": 1.5551879699248121, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 30855 + }, + { + "epoch": 1.5552481203007518, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 30856 + }, + { + "epoch": 1.5553082706766919, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 30857 + }, + { + "epoch": 1.5553684210526315, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 30858 + }, + { + "epoch": 1.5554285714285714, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 30859 + }, + { + "epoch": 1.5554887218045113, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 30860 + }, + { + "epoch": 1.5555488721804511, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1446, + "step": 30861 + }, + { + "epoch": 1.555609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 30862 + }, + { + "epoch": 1.5556691729323309, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 30863 + }, + { + "epoch": 1.5557293233082707, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 30864 + }, + { + "epoch": 1.5557894736842104, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 30865 + }, + { + "epoch": 1.5558496240601505, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 30866 + }, + { + "epoch": 1.5559097744360901, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 30867 + }, + { + "epoch": 1.5559699248120302, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 30868 + }, + { + "epoch": 1.5560300751879699, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 30869 + }, + { + "epoch": 1.5560902255639097, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 30870 + }, + { + "epoch": 1.5561503759398496, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 30871 + }, + { + "epoch": 1.5562105263157895, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 30872 + }, + { + "epoch": 1.5562706766917294, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 30873 + }, + { + "epoch": 1.556330827067669, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 30874 + }, + { + "epoch": 1.556390977443609, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 30875 + }, + { + "epoch": 1.5564511278195488, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 30876 + }, + { + "epoch": 1.5565112781954888, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 30877 + }, + { + "epoch": 1.5565714285714285, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 30878 + }, + { + "epoch": 1.5566315789473684, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 30879 + }, + { + "epoch": 1.5566917293233082, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 30880 + }, + { + "epoch": 1.556751879699248, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 30881 + }, + { + "epoch": 1.556812030075188, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 30882 + }, + { + "epoch": 1.5568721804511279, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 30883 + }, + { + "epoch": 1.5569323308270677, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 30884 + }, + { + "epoch": 1.5569924812030074, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 30885 + }, + { + "epoch": 1.5570526315789475, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 30886 + }, + { + "epoch": 1.5571127819548871, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 30887 + }, + { + "epoch": 1.5571729323308272, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 30888 + }, + { + "epoch": 1.5572330827067669, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 30889 + }, + { + "epoch": 1.5572932330827067, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 30890 + }, + { + "epoch": 1.5573533834586466, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 30891 + }, + { + "epoch": 1.5574135338345865, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 30892 + }, + { + "epoch": 1.5574736842105263, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 30893 + }, + { + "epoch": 1.5575338345864662, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 30894 + }, + { + "epoch": 1.557593984962406, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 30895 + }, + { + "epoch": 1.5576541353383457, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 30896 + }, + { + "epoch": 1.5577142857142858, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 30897 + }, + { + "epoch": 1.5577744360902255, + "grad_norm": 1.5234375, + "learning_rate": 3e-05, + "loss": 0.8941, + "step": 30898 + }, + { + "epoch": 1.5578345864661656, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 30899 + }, + { + "epoch": 1.5578947368421052, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 30900 + }, + { + "epoch": 1.557954887218045, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.2861, + "step": 30901 + }, + { + "epoch": 1.558015037593985, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 30902 + }, + { + "epoch": 1.5580751879699248, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 30903 + }, + { + "epoch": 1.5581353383458647, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 30904 + }, + { + "epoch": 1.5581954887218044, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 30905 + }, + { + "epoch": 1.5582556390977444, + "grad_norm": 1.484375, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 30906 + }, + { + "epoch": 1.558315789473684, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 30907 + }, + { + "epoch": 1.5583759398496242, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 30908 + }, + { + "epoch": 1.5584360902255638, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 30909 + }, + { + "epoch": 1.5584962406015037, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 30910 + }, + { + "epoch": 1.5585563909774436, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 30911 + }, + { + "epoch": 1.5586165413533835, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 30912 + }, + { + "epoch": 1.5586766917293233, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2982, + "step": 30913 + }, + { + "epoch": 1.5587368421052632, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 30914 + }, + { + "epoch": 1.558796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 30915 + }, + { + "epoch": 1.5588571428571427, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 30916 + }, + { + "epoch": 1.5589172932330828, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 30917 + }, + { + "epoch": 1.5589774436090225, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 30918 + }, + { + "epoch": 1.5590375939849626, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 30919 + }, + { + "epoch": 1.5590977443609022, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 30920 + }, + { + "epoch": 1.559157894736842, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 30921 + }, + { + "epoch": 1.559218045112782, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 30922 + }, + { + "epoch": 1.5592781954887218, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2065, + "step": 30923 + }, + { + "epoch": 1.5593383458646617, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 30924 + }, + { + "epoch": 1.5593984962406013, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 30925 + }, + { + "epoch": 1.5594586466165414, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 30926 + }, + { + "epoch": 1.559518796992481, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 30927 + }, + { + "epoch": 1.5595789473684212, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 30928 + }, + { + "epoch": 1.5596390977443608, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 30929 + }, + { + "epoch": 1.5596992481203007, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 30930 + }, + { + "epoch": 1.5597593984962406, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 30931 + }, + { + "epoch": 1.5598195488721804, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 30932 + }, + { + "epoch": 1.5598796992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 30933 + }, + { + "epoch": 1.5599398496240602, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 30934 + }, + { + "epoch": 1.56, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 30935 + }, + { + "epoch": 1.5600601503759397, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 30936 + }, + { + "epoch": 1.5601203007518798, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 30937 + }, + { + "epoch": 1.5601804511278194, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 30938 + }, + { + "epoch": 1.5602406015037595, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 30939 + }, + { + "epoch": 1.5603007518796992, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1764, + "step": 30940 + }, + { + "epoch": 1.560360902255639, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 30941 + }, + { + "epoch": 1.560421052631579, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 30942 + }, + { + "epoch": 1.5604812030075188, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 30943 + }, + { + "epoch": 1.5605413533834587, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 30944 + }, + { + "epoch": 1.5606015037593985, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 30945 + }, + { + "epoch": 1.5606616541353384, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 30946 + }, + { + "epoch": 1.560721804511278, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 30947 + }, + { + "epoch": 1.5607819548872182, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 30948 + }, + { + "epoch": 1.5608421052631578, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 30949 + }, + { + "epoch": 1.560902255639098, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 30950 + }, + { + "epoch": 1.5609624060150376, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 30951 + }, + { + "epoch": 1.5610225563909774, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 30952 + }, + { + "epoch": 1.5610827067669173, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 30953 + }, + { + "epoch": 1.5611428571428572, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 30954 + }, + { + "epoch": 1.561203007518797, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 30955 + }, + { + "epoch": 1.5612631578947367, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 30956 + }, + { + "epoch": 1.5613233082706768, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 30957 + }, + { + "epoch": 1.5613834586466164, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 30958 + }, + { + "epoch": 1.5614436090225565, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 30959 + }, + { + "epoch": 1.5615037593984962, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 30960 + }, + { + "epoch": 1.561563909774436, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 30961 + }, + { + "epoch": 1.561624060150376, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 30962 + }, + { + "epoch": 1.5616842105263158, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 30963 + }, + { + "epoch": 1.5617443609022557, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 30964 + }, + { + "epoch": 1.5618045112781955, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 30965 + }, + { + "epoch": 1.5618646616541354, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 30966 + }, + { + "epoch": 1.561924812030075, + "grad_norm": 1.921875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 30967 + }, + { + "epoch": 1.5619849624060151, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1916, + "step": 30968 + }, + { + "epoch": 1.5620451127819548, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 30969 + }, + { + "epoch": 1.5621052631578949, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 30970 + }, + { + "epoch": 1.5621654135338345, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 30971 + }, + { + "epoch": 1.5622255639097744, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 30972 + }, + { + "epoch": 1.5622857142857143, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 30973 + }, + { + "epoch": 1.5623458646616541, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 30974 + }, + { + "epoch": 1.562406015037594, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1392, + "step": 30975 + }, + { + "epoch": 1.5624661654135337, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 30976 + }, + { + "epoch": 1.5625263157894738, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1444, + "step": 30977 + }, + { + "epoch": 1.5625864661654134, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 30978 + }, + { + "epoch": 1.5626466165413535, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 30979 + }, + { + "epoch": 1.5627067669172932, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 30980 + }, + { + "epoch": 1.5627669172932332, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 30981 + }, + { + "epoch": 1.562827067669173, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 30982 + }, + { + "epoch": 1.5628872180451128, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 30983 + }, + { + "epoch": 1.5629473684210526, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 30984 + }, + { + "epoch": 1.5630075187969925, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 30985 + }, + { + "epoch": 1.5630676691729324, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 30986 + }, + { + "epoch": 1.563127819548872, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 30987 + }, + { + "epoch": 1.5631879699248121, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 30988 + }, + { + "epoch": 1.5632481203007518, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 30989 + }, + { + "epoch": 1.5633082706766919, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 30990 + }, + { + "epoch": 1.5633684210526315, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 30991 + }, + { + "epoch": 1.5634285714285714, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 30992 + }, + { + "epoch": 1.5634887218045113, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 30993 + }, + { + "epoch": 1.5635488721804511, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 30994 + }, + { + "epoch": 1.563609022556391, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 30995 + }, + { + "epoch": 1.5636691729323309, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 30996 + }, + { + "epoch": 1.5637293233082707, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.947, + "step": 30997 + }, + { + "epoch": 1.5637894736842104, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 30998 + }, + { + "epoch": 1.5638496240601505, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 30999 + }, + { + "epoch": 1.5639097744360901, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 31000 + }, + { + "epoch": 1.5639699248120302, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 31001 + }, + { + "epoch": 1.5640300751879699, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 31002 + }, + { + "epoch": 1.5640902255639098, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 31003 + }, + { + "epoch": 1.5641503759398496, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1855, + "step": 31004 + }, + { + "epoch": 1.5642105263157895, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 31005 + }, + { + "epoch": 1.5642706766917294, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 31006 + }, + { + "epoch": 1.564330827067669, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 31007 + }, + { + "epoch": 1.564390977443609, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 31008 + }, + { + "epoch": 1.5644511278195488, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 31009 + }, + { + "epoch": 1.5645112781954889, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 31010 + }, + { + "epoch": 1.5645714285714285, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 31011 + }, + { + "epoch": 1.5646315789473684, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 31012 + }, + { + "epoch": 1.5646917293233082, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 31013 + }, + { + "epoch": 1.5647518796992481, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 31014 + }, + { + "epoch": 1.564812030075188, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 31015 + }, + { + "epoch": 1.5648721804511279, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 31016 + }, + { + "epoch": 1.5649323308270677, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 31017 + }, + { + "epoch": 1.5649924812030074, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 31018 + }, + { + "epoch": 1.5650526315789475, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 31019 + }, + { + "epoch": 1.5651127819548871, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 31020 + }, + { + "epoch": 1.5651729323308272, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 31021 + }, + { + "epoch": 1.5652330827067669, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 31022 + }, + { + "epoch": 1.5652932330827067, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 31023 + }, + { + "epoch": 1.5653533834586466, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 31024 + }, + { + "epoch": 1.5654135338345865, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 31025 + }, + { + "epoch": 1.5654736842105264, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 31026 + }, + { + "epoch": 1.565533834586466, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 31027 + }, + { + "epoch": 1.565593984962406, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 31028 + }, + { + "epoch": 1.5656541353383457, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 31029 + }, + { + "epoch": 1.5657142857142858, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 31030 + }, + { + "epoch": 1.5657744360902255, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 31031 + }, + { + "epoch": 1.5658345864661656, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 31032 + }, + { + "epoch": 1.5658947368421052, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 31033 + }, + { + "epoch": 1.565954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 31034 + }, + { + "epoch": 1.566015037593985, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 31035 + }, + { + "epoch": 1.5660751879699248, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 31036 + }, + { + "epoch": 1.5661353383458647, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1848, + "step": 31037 + }, + { + "epoch": 1.5661954887218044, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 31038 + }, + { + "epoch": 1.5662556390977445, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 31039 + }, + { + "epoch": 1.566315789473684, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 31040 + }, + { + "epoch": 1.5663759398496242, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 31041 + }, + { + "epoch": 1.5664360902255638, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 31042 + }, + { + "epoch": 1.5664962406015037, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 31043 + }, + { + "epoch": 1.5665563909774436, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 31044 + }, + { + "epoch": 1.5666165413533835, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 31045 + }, + { + "epoch": 1.5666766917293233, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 31046 + }, + { + "epoch": 1.5667368421052632, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 31047 + }, + { + "epoch": 1.566796992481203, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 31048 + }, + { + "epoch": 1.5668571428571427, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 31049 + }, + { + "epoch": 1.5669172932330828, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 31050 + }, + { + "epoch": 1.5669774436090225, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 31051 + }, + { + "epoch": 1.5670375939849626, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 31052 + }, + { + "epoch": 1.5670977443609022, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 31053 + }, + { + "epoch": 1.567157894736842, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 31054 + }, + { + "epoch": 1.567218045112782, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 31055 + }, + { + "epoch": 1.5672781954887218, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 31056 + }, + { + "epoch": 1.5673383458646617, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 31057 + }, + { + "epoch": 1.5673984962406013, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 31058 + }, + { + "epoch": 1.5674586466165414, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 31059 + }, + { + "epoch": 1.567518796992481, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 31060 + }, + { + "epoch": 1.5675789473684212, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 31061 + }, + { + "epoch": 1.5676390977443608, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 31062 + }, + { + "epoch": 1.5676992481203007, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 31063 + }, + { + "epoch": 1.5677593984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 31064 + }, + { + "epoch": 1.5678195488721804, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 31065 + }, + { + "epoch": 1.5678796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 31066 + }, + { + "epoch": 1.5679398496240602, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 31067 + }, + { + "epoch": 1.568, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 31068 + }, + { + "epoch": 1.5680601503759397, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 31069 + }, + { + "epoch": 1.5681203007518798, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 31070 + }, + { + "epoch": 1.5681804511278195, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 31071 + }, + { + "epoch": 1.5682406015037595, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 31072 + }, + { + "epoch": 1.5683007518796992, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 31073 + }, + { + "epoch": 1.568360902255639, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1355, + "step": 31074 + }, + { + "epoch": 1.568421052631579, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 31075 + }, + { + "epoch": 1.5684812030075188, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 31076 + }, + { + "epoch": 1.5685413533834587, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 31077 + }, + { + "epoch": 1.5686015037593983, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 31078 + }, + { + "epoch": 1.5686616541353384, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 31079 + }, + { + "epoch": 1.568721804511278, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 31080 + }, + { + "epoch": 1.5687819548872182, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 31081 + }, + { + "epoch": 1.5688421052631578, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 31082 + }, + { + "epoch": 1.568902255639098, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.205, + "step": 31083 + }, + { + "epoch": 1.5689624060150376, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 31084 + }, + { + "epoch": 1.5690225563909774, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 31085 + }, + { + "epoch": 1.5690827067669173, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 31086 + }, + { + "epoch": 1.5691428571428572, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 31087 + }, + { + "epoch": 1.569203007518797, + "grad_norm": 1.4921875, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 31088 + }, + { + "epoch": 1.5692631578947367, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 31089 + }, + { + "epoch": 1.5693233082706768, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 31090 + }, + { + "epoch": 1.5693834586466164, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 31091 + }, + { + "epoch": 1.5694436090225565, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 31092 + }, + { + "epoch": 1.5695037593984962, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 31093 + }, + { + "epoch": 1.569563909774436, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 31094 + }, + { + "epoch": 1.569624060150376, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 31095 + }, + { + "epoch": 1.5696842105263158, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 31096 + }, + { + "epoch": 1.5697443609022557, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 31097 + }, + { + "epoch": 1.5698045112781955, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 31098 + }, + { + "epoch": 1.5698646616541354, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 31099 + }, + { + "epoch": 1.569924812030075, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 31100 + }, + { + "epoch": 1.5699849624060151, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 31101 + }, + { + "epoch": 1.5700451127819548, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.2376, + "step": 31102 + }, + { + "epoch": 1.570105263157895, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 31103 + }, + { + "epoch": 1.5701654135338345, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 31104 + }, + { + "epoch": 1.5702255639097744, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 31105 + }, + { + "epoch": 1.5702857142857143, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 31106 + }, + { + "epoch": 1.5703458646616542, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 31107 + }, + { + "epoch": 1.570406015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 31108 + }, + { + "epoch": 1.5704661654135337, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 31109 + }, + { + "epoch": 1.5705263157894738, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 31110 + }, + { + "epoch": 1.5705864661654134, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 31111 + }, + { + "epoch": 1.5706466165413535, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 31112 + }, + { + "epoch": 1.5707067669172932, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 31113 + }, + { + "epoch": 1.570766917293233, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 31114 + }, + { + "epoch": 1.570827067669173, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 31115 + }, + { + "epoch": 1.5708872180451128, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 31116 + }, + { + "epoch": 1.5709473684210526, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 31117 + }, + { + "epoch": 1.5710075187969925, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 31118 + }, + { + "epoch": 1.5710676691729324, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 31119 + }, + { + "epoch": 1.571127819548872, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 31120 + }, + { + "epoch": 1.5711879699248121, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 31121 + }, + { + "epoch": 1.5712481203007518, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 31122 + }, + { + "epoch": 1.5713082706766919, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 31123 + }, + { + "epoch": 1.5713684210526315, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 31124 + }, + { + "epoch": 1.5714285714285714, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 31125 + }, + { + "epoch": 1.5714887218045113, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 31126 + }, + { + "epoch": 1.5715488721804511, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 31127 + }, + { + "epoch": 1.571609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 31128 + }, + { + "epoch": 1.5716691729323309, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.248, + "step": 31129 + }, + { + "epoch": 1.5717293233082708, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 31130 + }, + { + "epoch": 1.5717894736842104, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 31131 + }, + { + "epoch": 1.5718496240601505, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 31132 + }, + { + "epoch": 1.5719097744360901, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 31133 + }, + { + "epoch": 1.5719699248120302, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1254, + "step": 31134 + }, + { + "epoch": 1.5720300751879699, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 31135 + }, + { + "epoch": 1.5720902255639098, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 31136 + }, + { + "epoch": 1.5721503759398496, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 31137 + }, + { + "epoch": 1.5722105263157895, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 31138 + }, + { + "epoch": 1.5722706766917294, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 31139 + }, + { + "epoch": 1.572330827067669, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 31140 + }, + { + "epoch": 1.5723909774436091, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 31141 + }, + { + "epoch": 1.5724511278195488, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 31142 + }, + { + "epoch": 1.5725112781954889, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 31143 + }, + { + "epoch": 1.5725714285714285, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1167, + "step": 31144 + }, + { + "epoch": 1.5726315789473684, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 31145 + }, + { + "epoch": 1.5726917293233083, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 31146 + }, + { + "epoch": 1.5727518796992481, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 31147 + }, + { + "epoch": 1.572812030075188, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 31148 + }, + { + "epoch": 1.5728721804511279, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 31149 + }, + { + "epoch": 1.5729323308270677, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 31150 + }, + { + "epoch": 1.5729924812030074, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 31151 + }, + { + "epoch": 1.5730526315789475, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 31152 + }, + { + "epoch": 1.5731127819548871, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 31153 + }, + { + "epoch": 1.5731729323308272, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 31154 + }, + { + "epoch": 1.5732330827067669, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9392, + "step": 31155 + }, + { + "epoch": 1.5732932330827067, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 31156 + }, + { + "epoch": 1.5733533834586466, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 31157 + }, + { + "epoch": 1.5734135338345865, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 31158 + }, + { + "epoch": 1.5734736842105264, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 31159 + }, + { + "epoch": 1.573533834586466, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 31160 + }, + { + "epoch": 1.573593984962406, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 31161 + }, + { + "epoch": 1.5736541353383457, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 31162 + }, + { + "epoch": 1.5737142857142858, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 31163 + }, + { + "epoch": 1.5737744360902255, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 31164 + }, + { + "epoch": 1.5738345864661654, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 31165 + }, + { + "epoch": 1.5738947368421052, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 31166 + }, + { + "epoch": 1.573954887218045, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 31167 + }, + { + "epoch": 1.574015037593985, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 31168 + }, + { + "epoch": 1.5740751879699248, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 31169 + }, + { + "epoch": 1.5741353383458647, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 31170 + }, + { + "epoch": 1.5741954887218044, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 31171 + }, + { + "epoch": 1.5742556390977445, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 31172 + }, + { + "epoch": 1.5743157894736841, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 31173 + }, + { + "epoch": 1.5743759398496242, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 31174 + }, + { + "epoch": 1.5744360902255639, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 31175 + }, + { + "epoch": 1.5744962406015037, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 31176 + }, + { + "epoch": 1.5745563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 31177 + }, + { + "epoch": 1.5746165413533835, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 31178 + }, + { + "epoch": 1.5746766917293233, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 31179 + }, + { + "epoch": 1.5747368421052632, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 31180 + }, + { + "epoch": 1.574796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 31181 + }, + { + "epoch": 1.5748571428571427, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 31182 + }, + { + "epoch": 1.5749172932330828, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 31183 + }, + { + "epoch": 1.5749774436090225, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 31184 + }, + { + "epoch": 1.5750375939849626, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 31185 + }, + { + "epoch": 1.5750977443609022, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 31186 + }, + { + "epoch": 1.575157894736842, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 31187 + }, + { + "epoch": 1.575218045112782, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 31188 + }, + { + "epoch": 1.5752781954887218, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 31189 + }, + { + "epoch": 1.5753383458646617, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 31190 + }, + { + "epoch": 1.5753984962406014, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 31191 + }, + { + "epoch": 1.5754586466165414, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 31192 + }, + { + "epoch": 1.575518796992481, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 31193 + }, + { + "epoch": 1.5755789473684212, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 31194 + }, + { + "epoch": 1.5756390977443608, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 31195 + }, + { + "epoch": 1.5756992481203007, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1895, + "step": 31196 + }, + { + "epoch": 1.5757593984962406, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 31197 + }, + { + "epoch": 1.5758195488721805, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 31198 + }, + { + "epoch": 1.5758796992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 31199 + }, + { + "epoch": 1.5759398496240602, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 31200 + }, + { + "epoch": 1.576, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 31201 + }, + { + "epoch": 1.5760601503759397, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 31202 + }, + { + "epoch": 1.5761203007518798, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 31203 + }, + { + "epoch": 1.5761804511278195, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 31204 + }, + { + "epoch": 1.5762406015037596, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 31205 + }, + { + "epoch": 1.5763007518796992, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 31206 + }, + { + "epoch": 1.576360902255639, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 31207 + }, + { + "epoch": 1.576421052631579, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 31208 + }, + { + "epoch": 1.5764812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 31209 + }, + { + "epoch": 1.5765413533834587, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 31210 + }, + { + "epoch": 1.5766015037593983, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 31211 + }, + { + "epoch": 1.5766616541353384, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2996, + "step": 31212 + }, + { + "epoch": 1.576721804511278, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 31213 + }, + { + "epoch": 1.5767819548872182, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 31214 + }, + { + "epoch": 1.5768421052631578, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 31215 + }, + { + "epoch": 1.5769022556390977, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 31216 + }, + { + "epoch": 1.5769624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 31217 + }, + { + "epoch": 1.5770225563909774, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 31218 + }, + { + "epoch": 1.5770827067669173, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 31219 + }, + { + "epoch": 1.5771428571428572, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 31220 + }, + { + "epoch": 1.577203007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 31221 + }, + { + "epoch": 1.5772631578947367, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2263, + "step": 31222 + }, + { + "epoch": 1.5773233082706768, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 31223 + }, + { + "epoch": 1.5773834586466164, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 31224 + }, + { + "epoch": 1.5774436090225565, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 31225 + }, + { + "epoch": 1.5775037593984962, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 31226 + }, + { + "epoch": 1.577563909774436, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 31227 + }, + { + "epoch": 1.577624060150376, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 31228 + }, + { + "epoch": 1.5776842105263158, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 31229 + }, + { + "epoch": 1.5777443609022557, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 31230 + }, + { + "epoch": 1.5778045112781955, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 31231 + }, + { + "epoch": 1.5778646616541354, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 31232 + }, + { + "epoch": 1.577924812030075, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 31233 + }, + { + "epoch": 1.5779849624060152, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.178, + "step": 31234 + }, + { + "epoch": 1.5780451127819548, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 31235 + }, + { + "epoch": 1.578105263157895, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 31236 + }, + { + "epoch": 1.5781654135338345, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 31237 + }, + { + "epoch": 1.5782255639097744, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 31238 + }, + { + "epoch": 1.5782857142857143, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.2682, + "step": 31239 + }, + { + "epoch": 1.5783458646616542, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 31240 + }, + { + "epoch": 1.578406015037594, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 31241 + }, + { + "epoch": 1.5784661654135337, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 31242 + }, + { + "epoch": 1.5785263157894738, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 31243 + }, + { + "epoch": 1.5785864661654134, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 31244 + }, + { + "epoch": 1.5786466165413535, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 31245 + }, + { + "epoch": 1.5787067669172932, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 31246 + }, + { + "epoch": 1.578766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 31247 + }, + { + "epoch": 1.578827067669173, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 31248 + }, + { + "epoch": 1.5788872180451128, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 31249 + }, + { + "epoch": 1.5789473684210527, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 31250 + }, + { + "epoch": 1.5790075187969925, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 31251 + }, + { + "epoch": 1.5790676691729324, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 31252 + }, + { + "epoch": 1.579127819548872, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 31253 + }, + { + "epoch": 1.5791879699248121, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 31254 + }, + { + "epoch": 1.5792481203007518, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 31255 + }, + { + "epoch": 1.5793082706766919, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 31256 + }, + { + "epoch": 1.5793684210526315, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 31257 + }, + { + "epoch": 1.5794285714285714, + "grad_norm": 1.3984375, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 31258 + }, + { + "epoch": 1.5794887218045113, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 31259 + }, + { + "epoch": 1.5795488721804511, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 31260 + }, + { + "epoch": 1.579609022556391, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 31261 + }, + { + "epoch": 1.5796691729323307, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 31262 + }, + { + "epoch": 1.5797293233082708, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 31263 + }, + { + "epoch": 1.5797894736842104, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 31264 + }, + { + "epoch": 1.5798496240601505, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 31265 + }, + { + "epoch": 1.5799097744360902, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 31266 + }, + { + "epoch": 1.57996992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 31267 + }, + { + "epoch": 1.58003007518797, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2787, + "step": 31268 + }, + { + "epoch": 1.5800902255639098, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 31269 + }, + { + "epoch": 1.5801503759398496, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 31270 + }, + { + "epoch": 1.5802105263157895, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 31271 + }, + { + "epoch": 1.5802706766917294, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 31272 + }, + { + "epoch": 1.580330827067669, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 31273 + }, + { + "epoch": 1.5803909774436091, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 31274 + }, + { + "epoch": 1.5804511278195488, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 31275 + }, + { + "epoch": 1.5805112781954889, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 31276 + }, + { + "epoch": 1.5805714285714285, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 31277 + }, + { + "epoch": 1.5806315789473684, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 31278 + }, + { + "epoch": 1.5806917293233083, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 31279 + }, + { + "epoch": 1.5807518796992481, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 31280 + }, + { + "epoch": 1.580812030075188, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1754, + "step": 31281 + }, + { + "epoch": 1.5808721804511279, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 31282 + }, + { + "epoch": 1.5809323308270677, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 31283 + }, + { + "epoch": 1.5809924812030074, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9182, + "step": 31284 + }, + { + "epoch": 1.5810526315789475, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 31285 + }, + { + "epoch": 1.5811127819548871, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 31286 + }, + { + "epoch": 1.5811729323308272, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 31287 + }, + { + "epoch": 1.5812330827067669, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 31288 + }, + { + "epoch": 1.5812932330827068, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 31289 + }, + { + "epoch": 1.5813533834586466, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 31290 + }, + { + "epoch": 1.5814135338345865, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 31291 + }, + { + "epoch": 1.5814736842105264, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 31292 + }, + { + "epoch": 1.581533834586466, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 31293 + }, + { + "epoch": 1.581593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 31294 + }, + { + "epoch": 1.5816541353383458, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 31295 + }, + { + "epoch": 1.5817142857142859, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 31296 + }, + { + "epoch": 1.5817744360902255, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 31297 + }, + { + "epoch": 1.5818345864661654, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 31298 + }, + { + "epoch": 1.5818947368421052, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 31299 + }, + { + "epoch": 1.5819548872180451, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 31300 + }, + { + "epoch": 1.582015037593985, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 31301 + }, + { + "epoch": 1.5820751879699249, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 31302 + }, + { + "epoch": 1.5821353383458647, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 31303 + }, + { + "epoch": 1.5821954887218044, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 31304 + }, + { + "epoch": 1.5822556390977445, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 31305 + }, + { + "epoch": 1.5823157894736841, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 31306 + }, + { + "epoch": 1.5823759398496242, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 31307 + }, + { + "epoch": 1.5824360902255639, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 31308 + }, + { + "epoch": 1.5824962406015037, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 31309 + }, + { + "epoch": 1.5825563909774436, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 31310 + }, + { + "epoch": 1.5826165413533835, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 31311 + }, + { + "epoch": 1.5826766917293233, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 31312 + }, + { + "epoch": 1.582736842105263, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 31313 + }, + { + "epoch": 1.582796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 31314 + }, + { + "epoch": 1.5828571428571427, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 31315 + }, + { + "epoch": 1.5829172932330828, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 31316 + }, + { + "epoch": 1.5829774436090225, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 31317 + }, + { + "epoch": 1.5830375939849624, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 31318 + }, + { + "epoch": 1.5830977443609022, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 31319 + }, + { + "epoch": 1.583157894736842, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 31320 + }, + { + "epoch": 1.583218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 31321 + }, + { + "epoch": 1.5832781954887218, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 31322 + }, + { + "epoch": 1.5833383458646617, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 31323 + }, + { + "epoch": 1.5833984962406014, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 31324 + }, + { + "epoch": 1.5834586466165415, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 31325 + }, + { + "epoch": 1.583518796992481, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 31326 + }, + { + "epoch": 1.5835789473684212, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 31327 + }, + { + "epoch": 1.5836390977443608, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 31328 + }, + { + "epoch": 1.5836992481203007, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 31329 + }, + { + "epoch": 1.5837593984962406, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 31330 + }, + { + "epoch": 1.5838195488721805, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 31331 + }, + { + "epoch": 1.5838796992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 31332 + }, + { + "epoch": 1.5839398496240602, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 31333 + }, + { + "epoch": 1.584, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 31334 + }, + { + "epoch": 1.5840601503759397, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 31335 + }, + { + "epoch": 1.5841203007518798, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 31336 + }, + { + "epoch": 1.5841804511278195, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 31337 + }, + { + "epoch": 1.5842406015037596, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 31338 + }, + { + "epoch": 1.5843007518796992, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 31339 + }, + { + "epoch": 1.584360902255639, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 31340 + }, + { + "epoch": 1.584421052631579, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 31341 + }, + { + "epoch": 1.5844812030075188, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 31342 + }, + { + "epoch": 1.5845413533834587, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 31343 + }, + { + "epoch": 1.5846015037593983, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 31344 + }, + { + "epoch": 1.5846616541353384, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 31345 + }, + { + "epoch": 1.584721804511278, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 31346 + }, + { + "epoch": 1.5847819548872182, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 31347 + }, + { + "epoch": 1.5848421052631578, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1792, + "step": 31348 + }, + { + "epoch": 1.5849022556390977, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.3398, + "step": 31349 + }, + { + "epoch": 1.5849624060150376, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 31350 + }, + { + "epoch": 1.5850225563909774, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 31351 + }, + { + "epoch": 1.5850827067669173, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 31352 + }, + { + "epoch": 1.5851428571428572, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 31353 + }, + { + "epoch": 1.585203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1789, + "step": 31354 + }, + { + "epoch": 1.5852631578947367, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 31355 + }, + { + "epoch": 1.5853233082706768, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 31356 + }, + { + "epoch": 1.5853834586466165, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 31357 + }, + { + "epoch": 1.5854436090225565, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2176, + "step": 31358 + }, + { + "epoch": 1.5855037593984962, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 31359 + }, + { + "epoch": 1.585563909774436, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 31360 + }, + { + "epoch": 1.585624060150376, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 31361 + }, + { + "epoch": 1.5856842105263158, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 31362 + }, + { + "epoch": 1.5857443609022557, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 31363 + }, + { + "epoch": 1.5858045112781953, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 31364 + }, + { + "epoch": 1.5858646616541354, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 31365 + }, + { + "epoch": 1.585924812030075, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 31366 + }, + { + "epoch": 1.5859849624060152, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 31367 + }, + { + "epoch": 1.5860451127819548, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 31368 + }, + { + "epoch": 1.5861052631578947, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 31369 + }, + { + "epoch": 1.5861654135338346, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 31370 + }, + { + "epoch": 1.5862255639097744, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 31371 + }, + { + "epoch": 1.5862857142857143, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 31372 + }, + { + "epoch": 1.5863458646616542, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 31373 + }, + { + "epoch": 1.586406015037594, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 31374 + }, + { + "epoch": 1.5864661654135337, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 31375 + }, + { + "epoch": 1.5865263157894738, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 31376 + }, + { + "epoch": 1.5865864661654134, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 31377 + }, + { + "epoch": 1.5866466165413535, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 31378 + }, + { + "epoch": 1.5867067669172932, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 31379 + }, + { + "epoch": 1.586766917293233, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 31380 + }, + { + "epoch": 1.586827067669173, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 31381 + }, + { + "epoch": 1.5868872180451128, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 31382 + }, + { + "epoch": 1.5869473684210527, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 31383 + }, + { + "epoch": 1.5870075187969925, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 31384 + }, + { + "epoch": 1.5870676691729324, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 31385 + }, + { + "epoch": 1.587127819548872, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 31386 + }, + { + "epoch": 1.5871879699248121, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 31387 + }, + { + "epoch": 1.5872481203007518, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 31388 + }, + { + "epoch": 1.587308270676692, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 31389 + }, + { + "epoch": 1.5873684210526315, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 31390 + }, + { + "epoch": 1.5874285714285714, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 31391 + }, + { + "epoch": 1.5874887218045113, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 31392 + }, + { + "epoch": 1.5875488721804512, + "grad_norm": 3.9375, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 31393 + }, + { + "epoch": 1.587609022556391, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 31394 + }, + { + "epoch": 1.5876691729323307, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 31395 + }, + { + "epoch": 1.5877293233082708, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 31396 + }, + { + "epoch": 1.5877894736842104, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 31397 + }, + { + "epoch": 1.5878496240601505, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 31398 + }, + { + "epoch": 1.5879097744360902, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 31399 + }, + { + "epoch": 1.58796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 31400 + }, + { + "epoch": 1.58803007518797, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2449, + "step": 31401 + }, + { + "epoch": 1.5880902255639098, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 31402 + }, + { + "epoch": 1.5881503759398496, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 31403 + }, + { + "epoch": 1.5882105263157895, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 31404 + }, + { + "epoch": 1.5882706766917294, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 31405 + }, + { + "epoch": 1.588330827067669, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 31406 + }, + { + "epoch": 1.5883909774436091, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 31407 + }, + { + "epoch": 1.5884511278195488, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 31408 + }, + { + "epoch": 1.5885112781954889, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.243, + "step": 31409 + }, + { + "epoch": 1.5885714285714285, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 31410 + }, + { + "epoch": 1.5886315789473684, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 31411 + }, + { + "epoch": 1.5886917293233083, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 31412 + }, + { + "epoch": 1.5887518796992481, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 31413 + }, + { + "epoch": 1.588812030075188, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 31414 + }, + { + "epoch": 1.5888721804511277, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 31415 + }, + { + "epoch": 1.5889323308270678, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 31416 + }, + { + "epoch": 1.5889924812030074, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.2058, + "step": 31417 + }, + { + "epoch": 1.5890526315789475, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 31418 + }, + { + "epoch": 1.5891127819548871, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 31419 + }, + { + "epoch": 1.589172932330827, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 31420 + }, + { + "epoch": 1.5892330827067669, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 31421 + }, + { + "epoch": 1.5892932330827068, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9491, + "step": 31422 + }, + { + "epoch": 1.5893533834586466, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 31423 + }, + { + "epoch": 1.5894135338345865, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 31424 + }, + { + "epoch": 1.5894736842105264, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 31425 + }, + { + "epoch": 1.589533834586466, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 31426 + }, + { + "epoch": 1.5895939849624061, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 31427 + }, + { + "epoch": 1.5896541353383458, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.3255, + "step": 31428 + }, + { + "epoch": 1.5897142857142859, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 31429 + }, + { + "epoch": 1.5897744360902255, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 31430 + }, + { + "epoch": 1.5898345864661654, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 31431 + }, + { + "epoch": 1.5898947368421052, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 31432 + }, + { + "epoch": 1.5899548872180451, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 31433 + }, + { + "epoch": 1.590015037593985, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 31434 + }, + { + "epoch": 1.5900751879699249, + "grad_norm": 1.8046875, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 31435 + }, + { + "epoch": 1.5901353383458647, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 31436 + }, + { + "epoch": 1.5901954887218044, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 31437 + }, + { + "epoch": 1.5902556390977445, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 31438 + }, + { + "epoch": 1.5903157894736841, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 31439 + }, + { + "epoch": 1.5903759398496242, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 31440 + }, + { + "epoch": 1.5904360902255639, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 31441 + }, + { + "epoch": 1.5904962406015037, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 31442 + }, + { + "epoch": 1.5905563909774436, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 31443 + }, + { + "epoch": 1.5906165413533835, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 31444 + }, + { + "epoch": 1.5906766917293234, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 31445 + }, + { + "epoch": 1.590736842105263, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 31446 + }, + { + "epoch": 1.590796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 31447 + }, + { + "epoch": 1.5908571428571427, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 31448 + }, + { + "epoch": 1.5909172932330828, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 31449 + }, + { + "epoch": 1.5909774436090225, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 31450 + }, + { + "epoch": 1.5910375939849624, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 31451 + }, + { + "epoch": 1.5910977443609022, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 31452 + }, + { + "epoch": 1.591157894736842, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 31453 + }, + { + "epoch": 1.591218045112782, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 31454 + }, + { + "epoch": 1.5912781954887218, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2711, + "step": 31455 + }, + { + "epoch": 1.5913383458646617, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9351, + "step": 31456 + }, + { + "epoch": 1.5913984962406014, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 31457 + }, + { + "epoch": 1.5914586466165415, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 31458 + }, + { + "epoch": 1.591518796992481, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 31459 + }, + { + "epoch": 1.5915789473684212, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 31460 + }, + { + "epoch": 1.5916390977443609, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 31461 + }, + { + "epoch": 1.5916992481203007, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 31462 + }, + { + "epoch": 1.5917593984962406, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 31463 + }, + { + "epoch": 1.5918195488721805, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 31464 + }, + { + "epoch": 1.5918796992481203, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 31465 + }, + { + "epoch": 1.59193984962406, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 31466 + }, + { + "epoch": 1.592, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 31467 + }, + { + "epoch": 1.5920601503759397, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 31468 + }, + { + "epoch": 1.5921203007518798, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 31469 + }, + { + "epoch": 1.5921804511278195, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 31470 + }, + { + "epoch": 1.5922406015037593, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 31471 + }, + { + "epoch": 1.5923007518796992, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 31472 + }, + { + "epoch": 1.592360902255639, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 31473 + }, + { + "epoch": 1.592421052631579, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 31474 + }, + { + "epoch": 1.5924812030075188, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 31475 + }, + { + "epoch": 1.5925413533834587, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 31476 + }, + { + "epoch": 1.5926015037593984, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 31477 + }, + { + "epoch": 1.5926616541353384, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1561, + "step": 31478 + }, + { + "epoch": 1.592721804511278, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 31479 + }, + { + "epoch": 1.5927819548872182, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 31480 + }, + { + "epoch": 1.5928421052631578, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 31481 + }, + { + "epoch": 1.5929022556390977, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 31482 + }, + { + "epoch": 1.5929624060150376, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 31483 + }, + { + "epoch": 1.5930225563909775, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 31484 + }, + { + "epoch": 1.5930827067669173, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 31485 + }, + { + "epoch": 1.5931428571428572, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 31486 + }, + { + "epoch": 1.593203007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 31487 + }, + { + "epoch": 1.5932631578947367, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 31488 + }, + { + "epoch": 1.5933233082706768, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 31489 + }, + { + "epoch": 1.5933834586466165, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 31490 + }, + { + "epoch": 1.5934436090225566, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 31491 + }, + { + "epoch": 1.5935037593984962, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 31492 + }, + { + "epoch": 1.593563909774436, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 31493 + }, + { + "epoch": 1.593624060150376, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 31494 + }, + { + "epoch": 1.5936842105263158, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 31495 + }, + { + "epoch": 1.5937443609022557, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 31496 + }, + { + "epoch": 1.5938045112781953, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 31497 + }, + { + "epoch": 1.5938646616541354, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 31498 + }, + { + "epoch": 1.593924812030075, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 31499 + }, + { + "epoch": 1.5939849624060152, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 31500 + }, + { + "epoch": 1.5940451127819548, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 31501 + }, + { + "epoch": 1.5941052631578947, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 31502 + }, + { + "epoch": 1.5941654135338346, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 31503 + }, + { + "epoch": 1.5942255639097744, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 31504 + }, + { + "epoch": 1.5942857142857143, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 31505 + }, + { + "epoch": 1.5943458646616542, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 31506 + }, + { + "epoch": 1.594406015037594, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 31507 + }, + { + "epoch": 1.5944661654135337, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 31508 + }, + { + "epoch": 1.5945263157894738, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 31509 + }, + { + "epoch": 1.5945864661654134, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 31510 + }, + { + "epoch": 1.5946466165413535, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 31511 + }, + { + "epoch": 1.5947067669172932, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 31512 + }, + { + "epoch": 1.594766917293233, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 31513 + }, + { + "epoch": 1.594827067669173, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 31514 + }, + { + "epoch": 1.5948872180451128, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 31515 + }, + { + "epoch": 1.5949473684210527, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 31516 + }, + { + "epoch": 1.5950075187969923, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 31517 + }, + { + "epoch": 1.5950676691729324, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 31518 + }, + { + "epoch": 1.595127819548872, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 31519 + }, + { + "epoch": 1.5951879699248122, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 31520 + }, + { + "epoch": 1.5952481203007518, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 31521 + }, + { + "epoch": 1.5953082706766917, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 31522 + }, + { + "epoch": 1.5953684210526315, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 31523 + }, + { + "epoch": 1.5954285714285714, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 31524 + }, + { + "epoch": 1.5954887218045113, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 31525 + }, + { + "epoch": 1.5955488721804512, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 31526 + }, + { + "epoch": 1.595609022556391, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 31527 + }, + { + "epoch": 1.5956691729323307, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 31528 + }, + { + "epoch": 1.5957293233082708, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 31529 + }, + { + "epoch": 1.5957894736842104, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 31530 + }, + { + "epoch": 1.5958496240601505, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 31531 + }, + { + "epoch": 1.5959097744360902, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 31532 + }, + { + "epoch": 1.59596992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 31533 + }, + { + "epoch": 1.59603007518797, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 31534 + }, + { + "epoch": 1.5960902255639098, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 31535 + }, + { + "epoch": 1.5961503759398497, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 31536 + }, + { + "epoch": 1.5962105263157895, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 31537 + }, + { + "epoch": 1.5962706766917294, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 31538 + }, + { + "epoch": 1.596330827067669, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 31539 + }, + { + "epoch": 1.5963909774436091, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 31540 + }, + { + "epoch": 1.5964511278195488, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 31541 + }, + { + "epoch": 1.5965112781954889, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2275, + "step": 31542 + }, + { + "epoch": 1.5965714285714285, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 31543 + }, + { + "epoch": 1.5966315789473684, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 31544 + }, + { + "epoch": 1.5966917293233083, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 31545 + }, + { + "epoch": 1.5967518796992481, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 31546 + }, + { + "epoch": 1.596812030075188, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 31547 + }, + { + "epoch": 1.5968721804511277, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 31548 + }, + { + "epoch": 1.5969323308270678, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 31549 + }, + { + "epoch": 1.5969924812030074, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 31550 + }, + { + "epoch": 1.5970526315789475, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 31551 + }, + { + "epoch": 1.5971127819548872, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 31552 + }, + { + "epoch": 1.597172932330827, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 31553 + }, + { + "epoch": 1.597233082706767, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 31554 + }, + { + "epoch": 1.5972932330827068, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 31555 + }, + { + "epoch": 1.5973533834586466, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 31556 + }, + { + "epoch": 1.5974135338345865, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 31557 + }, + { + "epoch": 1.5974736842105264, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 31558 + }, + { + "epoch": 1.597533834586466, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 31559 + }, + { + "epoch": 1.5975939849624061, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 31560 + }, + { + "epoch": 1.5976541353383458, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 31561 + }, + { + "epoch": 1.5977142857142859, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 31562 + }, + { + "epoch": 1.5977744360902255, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 31563 + }, + { + "epoch": 1.5978345864661654, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 31564 + }, + { + "epoch": 1.5978947368421053, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 31565 + }, + { + "epoch": 1.5979548872180451, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 31566 + }, + { + "epoch": 1.598015037593985, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 31567 + }, + { + "epoch": 1.5980751879699246, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 31568 + }, + { + "epoch": 1.5981353383458647, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 31569 + }, + { + "epoch": 1.5981954887218044, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 31570 + }, + { + "epoch": 1.5982556390977445, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 31571 + }, + { + "epoch": 1.5983157894736841, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 31572 + }, + { + "epoch": 1.598375939849624, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 31573 + }, + { + "epoch": 1.5984360902255639, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 31574 + }, + { + "epoch": 1.5984962406015037, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 31575 + }, + { + "epoch": 1.5985563909774436, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 31576 + }, + { + "epoch": 1.5986165413533835, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 31577 + }, + { + "epoch": 1.5986766917293234, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 31578 + }, + { + "epoch": 1.598736842105263, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 31579 + }, + { + "epoch": 1.598796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 31580 + }, + { + "epoch": 1.5988571428571428, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 31581 + }, + { + "epoch": 1.5989172932330828, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 31582 + }, + { + "epoch": 1.5989774436090225, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 31583 + }, + { + "epoch": 1.5990375939849624, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 31584 + }, + { + "epoch": 1.5990977443609022, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 31585 + }, + { + "epoch": 1.5991578947368421, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 31586 + }, + { + "epoch": 1.599218045112782, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 31587 + }, + { + "epoch": 1.5992781954887219, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 31588 + }, + { + "epoch": 1.5993383458646617, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.2057, + "step": 31589 + }, + { + "epoch": 1.5993984962406014, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 31590 + }, + { + "epoch": 1.5994586466165415, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 31591 + }, + { + "epoch": 1.5995187969924811, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 31592 + }, + { + "epoch": 1.5995789473684212, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 31593 + }, + { + "epoch": 1.5996390977443609, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 31594 + }, + { + "epoch": 1.5996992481203007, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 31595 + }, + { + "epoch": 1.5997593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 31596 + }, + { + "epoch": 1.5998195488721805, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 31597 + }, + { + "epoch": 1.5998796992481203, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 31598 + }, + { + "epoch": 1.59993984962406, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 31599 + }, + { + "epoch": 1.6, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 31600 + }, + { + "epoch": 1.6000601503759397, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 31601 + }, + { + "epoch": 1.6001203007518798, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 31602 + }, + { + "epoch": 1.6001804511278195, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 31603 + }, + { + "epoch": 1.6002406015037594, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 31604 + }, + { + "epoch": 1.6003007518796992, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 31605 + }, + { + "epoch": 1.600360902255639, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 31606 + }, + { + "epoch": 1.600421052631579, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 31607 + }, + { + "epoch": 1.6004812030075188, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 31608 + }, + { + "epoch": 1.6005413533834587, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 31609 + }, + { + "epoch": 1.6006015037593984, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 31610 + }, + { + "epoch": 1.6006616541353385, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 31611 + }, + { + "epoch": 1.600721804511278, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 31612 + }, + { + "epoch": 1.6007819548872182, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 31613 + }, + { + "epoch": 1.6008421052631578, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 31614 + }, + { + "epoch": 1.6009022556390977, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 31615 + }, + { + "epoch": 1.6009624060150376, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 31616 + }, + { + "epoch": 1.6010225563909775, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 31617 + }, + { + "epoch": 1.6010827067669173, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 31618 + }, + { + "epoch": 1.601142857142857, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 31619 + }, + { + "epoch": 1.601203007518797, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 31620 + }, + { + "epoch": 1.6012631578947367, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 31621 + }, + { + "epoch": 1.6013233082706768, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 31622 + }, + { + "epoch": 1.6013834586466165, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 31623 + }, + { + "epoch": 1.6014436090225563, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 31624 + }, + { + "epoch": 1.6015037593984962, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 31625 + }, + { + "epoch": 1.601563909774436, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 31626 + }, + { + "epoch": 1.601624060150376, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 31627 + }, + { + "epoch": 1.6016842105263158, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 31628 + }, + { + "epoch": 1.6017443609022557, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 31629 + }, + { + "epoch": 1.6018045112781953, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 31630 + }, + { + "epoch": 1.6018646616541354, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 31631 + }, + { + "epoch": 1.601924812030075, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 31632 + }, + { + "epoch": 1.6019849624060152, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 31633 + }, + { + "epoch": 1.6020451127819548, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 31634 + }, + { + "epoch": 1.6021052631578947, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 31635 + }, + { + "epoch": 1.6021654135338346, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 31636 + }, + { + "epoch": 1.6022255639097744, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 31637 + }, + { + "epoch": 1.6022857142857143, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 31638 + }, + { + "epoch": 1.6023458646616542, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 31639 + }, + { + "epoch": 1.602406015037594, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 31640 + }, + { + "epoch": 1.6024661654135337, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 31641 + }, + { + "epoch": 1.6025263157894738, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 31642 + }, + { + "epoch": 1.6025864661654134, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 31643 + }, + { + "epoch": 1.6026466165413535, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 31644 + }, + { + "epoch": 1.6027067669172932, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 31645 + }, + { + "epoch": 1.602766917293233, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 31646 + }, + { + "epoch": 1.602827067669173, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 31647 + }, + { + "epoch": 1.6028872180451128, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 31648 + }, + { + "epoch": 1.6029473684210527, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 31649 + }, + { + "epoch": 1.6030075187969923, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 31650 + }, + { + "epoch": 1.6030676691729324, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 31651 + }, + { + "epoch": 1.603127819548872, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 31652 + }, + { + "epoch": 1.6031879699248122, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 31653 + }, + { + "epoch": 1.6032481203007518, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 31654 + }, + { + "epoch": 1.6033082706766917, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 31655 + }, + { + "epoch": 1.6033684210526316, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 31656 + }, + { + "epoch": 1.6034285714285714, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 31657 + }, + { + "epoch": 1.6034887218045113, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 31658 + }, + { + "epoch": 1.6035488721804512, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 31659 + }, + { + "epoch": 1.603609022556391, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 31660 + }, + { + "epoch": 1.6036691729323307, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 31661 + }, + { + "epoch": 1.6037293233082708, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 31662 + }, + { + "epoch": 1.6037894736842104, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 31663 + }, + { + "epoch": 1.6038496240601505, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 31664 + }, + { + "epoch": 1.6039097744360902, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 31665 + }, + { + "epoch": 1.60396992481203, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 31666 + }, + { + "epoch": 1.60403007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 31667 + }, + { + "epoch": 1.6040902255639098, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 31668 + }, + { + "epoch": 1.6041503759398497, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 31669 + }, + { + "epoch": 1.6042105263157893, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 31670 + }, + { + "epoch": 1.6042706766917294, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 31671 + }, + { + "epoch": 1.604330827067669, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 31672 + }, + { + "epoch": 1.6043909774436091, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 31673 + }, + { + "epoch": 1.6044511278195488, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 31674 + }, + { + "epoch": 1.6045112781954889, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 31675 + }, + { + "epoch": 1.6045714285714285, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 31676 + }, + { + "epoch": 1.6046315789473684, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 31677 + }, + { + "epoch": 1.6046917293233083, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 31678 + }, + { + "epoch": 1.6047518796992482, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 31679 + }, + { + "epoch": 1.604812030075188, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 31680 + }, + { + "epoch": 1.6048721804511277, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 31681 + }, + { + "epoch": 1.6049323308270678, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 31682 + }, + { + "epoch": 1.6049924812030074, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 31683 + }, + { + "epoch": 1.6050526315789475, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 31684 + }, + { + "epoch": 1.6051127819548872, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 31685 + }, + { + "epoch": 1.605172932330827, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 31686 + }, + { + "epoch": 1.605233082706767, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 31687 + }, + { + "epoch": 1.6052932330827068, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 31688 + }, + { + "epoch": 1.6053533834586466, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 31689 + }, + { + "epoch": 1.6054135338345865, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 31690 + }, + { + "epoch": 1.6054736842105264, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 31691 + }, + { + "epoch": 1.605533834586466, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 31692 + }, + { + "epoch": 1.6055939849624061, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 31693 + }, + { + "epoch": 1.6056541353383458, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 31694 + }, + { + "epoch": 1.6057142857142859, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 31695 + }, + { + "epoch": 1.6057744360902255, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 31696 + }, + { + "epoch": 1.6058345864661654, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 31697 + }, + { + "epoch": 1.6058947368421053, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 31698 + }, + { + "epoch": 1.6059548872180451, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 31699 + }, + { + "epoch": 1.606015037593985, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 31700 + }, + { + "epoch": 1.6060751879699247, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 31701 + }, + { + "epoch": 1.6061353383458647, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 31702 + }, + { + "epoch": 1.6061954887218044, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 31703 + }, + { + "epoch": 1.6062556390977445, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 31704 + }, + { + "epoch": 1.6063157894736841, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 31705 + }, + { + "epoch": 1.606375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 31706 + }, + { + "epoch": 1.6064360902255639, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 31707 + }, + { + "epoch": 1.6064962406015038, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 31708 + }, + { + "epoch": 1.6065563909774436, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 31709 + }, + { + "epoch": 1.6066165413533835, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 31710 + }, + { + "epoch": 1.6066766917293234, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 31711 + }, + { + "epoch": 1.606736842105263, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 31712 + }, + { + "epoch": 1.6067969924812031, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 31713 + }, + { + "epoch": 1.6068571428571428, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 31714 + }, + { + "epoch": 1.6069172932330829, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 31715 + }, + { + "epoch": 1.6069774436090225, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1428, + "step": 31716 + }, + { + "epoch": 1.6070375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 31717 + }, + { + "epoch": 1.6070977443609022, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 31718 + }, + { + "epoch": 1.6071578947368421, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 31719 + }, + { + "epoch": 1.607218045112782, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 31720 + }, + { + "epoch": 1.6072781954887216, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 31721 + }, + { + "epoch": 1.6073383458646617, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 31722 + }, + { + "epoch": 1.6073984962406014, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 31723 + }, + { + "epoch": 1.6074586466165415, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 31724 + }, + { + "epoch": 1.6075187969924811, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 31725 + }, + { + "epoch": 1.6075789473684212, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 31726 + }, + { + "epoch": 1.6076390977443609, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 31727 + }, + { + "epoch": 1.6076992481203007, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 31728 + }, + { + "epoch": 1.6077593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 31729 + }, + { + "epoch": 1.6078195488721805, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 31730 + }, + { + "epoch": 1.6078796992481204, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 31731 + }, + { + "epoch": 1.60793984962406, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 31732 + }, + { + "epoch": 1.608, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 31733 + }, + { + "epoch": 1.6080601503759397, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9495, + "step": 31734 + }, + { + "epoch": 1.6081203007518798, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 31735 + }, + { + "epoch": 1.6081804511278195, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 31736 + }, + { + "epoch": 1.6082406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 31737 + }, + { + "epoch": 1.6083007518796992, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 31738 + }, + { + "epoch": 1.608360902255639, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 31739 + }, + { + "epoch": 1.608421052631579, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 31740 + }, + { + "epoch": 1.6084812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 31741 + }, + { + "epoch": 1.6085413533834587, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 31742 + }, + { + "epoch": 1.6086015037593984, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 31743 + }, + { + "epoch": 1.6086616541353385, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 31744 + }, + { + "epoch": 1.608721804511278, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 31745 + }, + { + "epoch": 1.6087819548872182, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 31746 + }, + { + "epoch": 1.6088421052631579, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 31747 + }, + { + "epoch": 1.6089022556390977, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 31748 + }, + { + "epoch": 1.6089624060150376, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 31749 + }, + { + "epoch": 1.6090225563909775, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 31750 + }, + { + "epoch": 1.6090827067669173, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 31751 + }, + { + "epoch": 1.609142857142857, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 31752 + }, + { + "epoch": 1.609203007518797, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 31753 + }, + { + "epoch": 1.6092631578947367, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 31754 + }, + { + "epoch": 1.6093233082706768, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 31755 + }, + { + "epoch": 1.6093834586466165, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 31756 + }, + { + "epoch": 1.6094436090225563, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 31757 + }, + { + "epoch": 1.6095037593984962, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 31758 + }, + { + "epoch": 1.609563909774436, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 31759 + }, + { + "epoch": 1.609624060150376, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 31760 + }, + { + "epoch": 1.6096842105263158, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 31761 + }, + { + "epoch": 1.6097443609022557, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 31762 + }, + { + "epoch": 1.6098045112781953, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 31763 + }, + { + "epoch": 1.6098646616541354, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 31764 + }, + { + "epoch": 1.609924812030075, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 31765 + }, + { + "epoch": 1.6099849624060152, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 1.201, + "step": 31766 + }, + { + "epoch": 1.6100451127819548, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9595, + "step": 31767 + }, + { + "epoch": 1.6101052631578947, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 31768 + }, + { + "epoch": 1.6101654135338346, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.957, + "step": 31769 + }, + { + "epoch": 1.6102255639097744, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 31770 + }, + { + "epoch": 1.6102857142857143, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 31771 + }, + { + "epoch": 1.610345864661654, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 31772 + }, + { + "epoch": 1.610406015037594, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 31773 + }, + { + "epoch": 1.6104661654135337, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 31774 + }, + { + "epoch": 1.6105263157894738, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 31775 + }, + { + "epoch": 1.6105864661654135, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 31776 + }, + { + "epoch": 1.6106466165413535, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 31777 + }, + { + "epoch": 1.6107067669172932, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 31778 + }, + { + "epoch": 1.610766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 31779 + }, + { + "epoch": 1.610827067669173, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 31780 + }, + { + "epoch": 1.6108872180451128, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 31781 + }, + { + "epoch": 1.6109473684210527, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 31782 + }, + { + "epoch": 1.6110075187969923, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 31783 + }, + { + "epoch": 1.6110676691729324, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 31784 + }, + { + "epoch": 1.611127819548872, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 31785 + }, + { + "epoch": 1.6111879699248122, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 31786 + }, + { + "epoch": 1.6112481203007518, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 31787 + }, + { + "epoch": 1.6113082706766917, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 31788 + }, + { + "epoch": 1.6113684210526316, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 31789 + }, + { + "epoch": 1.6114285714285714, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 31790 + }, + { + "epoch": 1.6114887218045113, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 31791 + }, + { + "epoch": 1.6115488721804512, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 31792 + }, + { + "epoch": 1.611609022556391, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 31793 + }, + { + "epoch": 1.6116691729323307, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 31794 + }, + { + "epoch": 1.6117293233082708, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 31795 + }, + { + "epoch": 1.6117894736842104, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 31796 + }, + { + "epoch": 1.6118496240601505, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 31797 + }, + { + "epoch": 1.6119097744360902, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 31798 + }, + { + "epoch": 1.61196992481203, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 31799 + }, + { + "epoch": 1.61203007518797, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 31800 + }, + { + "epoch": 1.6120902255639098, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 31801 + }, + { + "epoch": 1.6121503759398497, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 31802 + }, + { + "epoch": 1.6122105263157893, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 31803 + }, + { + "epoch": 1.6122706766917294, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 31804 + }, + { + "epoch": 1.612330827067669, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 31805 + }, + { + "epoch": 1.6123909774436092, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 31806 + }, + { + "epoch": 1.6124511278195488, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 31807 + }, + { + "epoch": 1.6125112781954887, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 31808 + }, + { + "epoch": 1.6125714285714285, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 31809 + }, + { + "epoch": 1.6126315789473684, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 31810 + }, + { + "epoch": 1.6126917293233083, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 31811 + }, + { + "epoch": 1.6127518796992482, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 31812 + }, + { + "epoch": 1.612812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 31813 + }, + { + "epoch": 1.6128721804511277, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 31814 + }, + { + "epoch": 1.6129323308270678, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1528, + "step": 31815 + }, + { + "epoch": 1.6129924812030074, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 31816 + }, + { + "epoch": 1.6130526315789475, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 31817 + }, + { + "epoch": 1.6131127819548872, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 31818 + }, + { + "epoch": 1.613172932330827, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 31819 + }, + { + "epoch": 1.613233082706767, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 31820 + }, + { + "epoch": 1.6132932330827068, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 31821 + }, + { + "epoch": 1.6133533834586467, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 31822 + }, + { + "epoch": 1.6134135338345863, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 31823 + }, + { + "epoch": 1.6134736842105264, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 31824 + }, + { + "epoch": 1.613533834586466, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 31825 + }, + { + "epoch": 1.6135939849624061, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 31826 + }, + { + "epoch": 1.6136541353383458, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 31827 + }, + { + "epoch": 1.6137142857142859, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 31828 + }, + { + "epoch": 1.6137744360902255, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 31829 + }, + { + "epoch": 1.6138345864661654, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 31830 + }, + { + "epoch": 1.6138947368421053, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1721, + "step": 31831 + }, + { + "epoch": 1.6139548872180451, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 31832 + }, + { + "epoch": 1.614015037593985, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 31833 + }, + { + "epoch": 1.6140751879699247, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 31834 + }, + { + "epoch": 1.6141353383458648, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 31835 + }, + { + "epoch": 1.6141954887218044, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 31836 + }, + { + "epoch": 1.6142556390977445, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 31837 + }, + { + "epoch": 1.6143157894736841, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 31838 + }, + { + "epoch": 1.614375939849624, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 31839 + }, + { + "epoch": 1.614436090225564, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 31840 + }, + { + "epoch": 1.6144962406015038, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 31841 + }, + { + "epoch": 1.6145563909774436, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 31842 + }, + { + "epoch": 1.6146165413533835, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9333, + "step": 31843 + }, + { + "epoch": 1.6146766917293234, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 31844 + }, + { + "epoch": 1.614736842105263, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 31845 + }, + { + "epoch": 1.6147969924812031, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 31846 + }, + { + "epoch": 1.6148571428571428, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 31847 + }, + { + "epoch": 1.6149172932330829, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 31848 + }, + { + "epoch": 1.6149774436090225, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 31849 + }, + { + "epoch": 1.6150375939849624, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 31850 + }, + { + "epoch": 1.6150977443609023, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 31851 + }, + { + "epoch": 1.6151578947368421, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 31852 + }, + { + "epoch": 1.615218045112782, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 31853 + }, + { + "epoch": 1.6152781954887216, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 31854 + }, + { + "epoch": 1.6153383458646617, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 31855 + }, + { + "epoch": 1.6153984962406014, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 31856 + }, + { + "epoch": 1.6154586466165415, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 31857 + }, + { + "epoch": 1.6155187969924811, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 31858 + }, + { + "epoch": 1.615578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 31859 + }, + { + "epoch": 1.6156390977443609, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 31860 + }, + { + "epoch": 1.6156992481203007, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.171, + "step": 31861 + }, + { + "epoch": 1.6157593984962406, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 31862 + }, + { + "epoch": 1.6158195488721805, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 31863 + }, + { + "epoch": 1.6158796992481204, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 31864 + }, + { + "epoch": 1.61593984962406, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 31865 + }, + { + "epoch": 1.616, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 31866 + }, + { + "epoch": 1.6160601503759398, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 31867 + }, + { + "epoch": 1.6161203007518798, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 31868 + }, + { + "epoch": 1.6161804511278195, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 31869 + }, + { + "epoch": 1.6162406015037594, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 31870 + }, + { + "epoch": 1.6163007518796992, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 31871 + }, + { + "epoch": 1.616360902255639, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 31872 + }, + { + "epoch": 1.616421052631579, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 31873 + }, + { + "epoch": 1.6164812030075189, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 31874 + }, + { + "epoch": 1.6165413533834587, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 31875 + }, + { + "epoch": 1.6166015037593984, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 31876 + }, + { + "epoch": 1.6166616541353385, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.269, + "step": 31877 + }, + { + "epoch": 1.6167218045112781, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 31878 + }, + { + "epoch": 1.6167819548872182, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 31879 + }, + { + "epoch": 1.6168421052631579, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 31880 + }, + { + "epoch": 1.6169022556390977, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 31881 + }, + { + "epoch": 1.6169624060150376, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 31882 + }, + { + "epoch": 1.6170225563909775, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 31883 + }, + { + "epoch": 1.6170827067669173, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 31884 + }, + { + "epoch": 1.617142857142857, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 31885 + }, + { + "epoch": 1.617203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 31886 + }, + { + "epoch": 1.6172631578947367, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 31887 + }, + { + "epoch": 1.6173233082706768, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 31888 + }, + { + "epoch": 1.6173834586466165, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 31889 + }, + { + "epoch": 1.6174436090225564, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 31890 + }, + { + "epoch": 1.6175037593984962, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 31891 + }, + { + "epoch": 1.617563909774436, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 31892 + }, + { + "epoch": 1.617624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 31893 + }, + { + "epoch": 1.6176842105263158, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 31894 + }, + { + "epoch": 1.6177443609022557, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 31895 + }, + { + "epoch": 1.6178045112781954, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 31896 + }, + { + "epoch": 1.6178646616541354, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 31897 + }, + { + "epoch": 1.617924812030075, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 31898 + }, + { + "epoch": 1.6179849624060152, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 31899 + }, + { + "epoch": 1.6180451127819548, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 31900 + }, + { + "epoch": 1.6181052631578947, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 31901 + }, + { + "epoch": 1.6181654135338346, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 31902 + }, + { + "epoch": 1.6182255639097745, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 31903 + }, + { + "epoch": 1.6182857142857143, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 31904 + }, + { + "epoch": 1.618345864661654, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 31905 + }, + { + "epoch": 1.618406015037594, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 31906 + }, + { + "epoch": 1.6184661654135337, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2193, + "step": 31907 + }, + { + "epoch": 1.6185263157894738, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 31908 + }, + { + "epoch": 1.6185864661654135, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 31909 + }, + { + "epoch": 1.6186466165413533, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 31910 + }, + { + "epoch": 1.6187067669172932, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 31911 + }, + { + "epoch": 1.618766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 31912 + }, + { + "epoch": 1.618827067669173, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 31913 + }, + { + "epoch": 1.6188872180451128, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 31914 + }, + { + "epoch": 1.6189473684210527, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 31915 + }, + { + "epoch": 1.6190075187969923, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 31916 + }, + { + "epoch": 1.6190676691729324, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 31917 + }, + { + "epoch": 1.619127819548872, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 31918 + }, + { + "epoch": 1.6191879699248122, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 31919 + }, + { + "epoch": 1.6192481203007518, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 31920 + }, + { + "epoch": 1.6193082706766917, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 31921 + }, + { + "epoch": 1.6193684210526316, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 31922 + }, + { + "epoch": 1.6194285714285714, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 31923 + }, + { + "epoch": 1.6194887218045113, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 31924 + }, + { + "epoch": 1.6195488721804512, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 31925 + }, + { + "epoch": 1.619609022556391, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 31926 + }, + { + "epoch": 1.6196691729323307, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 31927 + }, + { + "epoch": 1.6197293233082708, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 31928 + }, + { + "epoch": 1.6197894736842104, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 31929 + }, + { + "epoch": 1.6198496240601505, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 31930 + }, + { + "epoch": 1.6199097744360902, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 31931 + }, + { + "epoch": 1.61996992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 31932 + }, + { + "epoch": 1.62003007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9263, + "step": 31933 + }, + { + "epoch": 1.6200902255639098, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 31934 + }, + { + "epoch": 1.6201503759398497, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 31935 + }, + { + "epoch": 1.6202105263157893, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 31936 + }, + { + "epoch": 1.6202706766917294, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 31937 + }, + { + "epoch": 1.620330827067669, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 31938 + }, + { + "epoch": 1.6203909774436092, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 31939 + }, + { + "epoch": 1.6204511278195488, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 31940 + }, + { + "epoch": 1.6205112781954887, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 31941 + }, + { + "epoch": 1.6205714285714286, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 31942 + }, + { + "epoch": 1.6206315789473684, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 31943 + }, + { + "epoch": 1.6206917293233083, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 31944 + }, + { + "epoch": 1.6207518796992482, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 31945 + }, + { + "epoch": 1.620812030075188, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 31946 + }, + { + "epoch": 1.6208721804511277, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 31947 + }, + { + "epoch": 1.6209323308270678, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 31948 + }, + { + "epoch": 1.6209924812030074, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 31949 + }, + { + "epoch": 1.6210526315789475, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 31950 + }, + { + "epoch": 1.6211127819548872, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 31951 + }, + { + "epoch": 1.621172932330827, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 31952 + }, + { + "epoch": 1.621233082706767, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 31953 + }, + { + "epoch": 1.6212932330827068, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 31954 + }, + { + "epoch": 1.6213533834586467, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 31955 + }, + { + "epoch": 1.6214135338345863, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 31956 + }, + { + "epoch": 1.6214736842105264, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 31957 + }, + { + "epoch": 1.621533834586466, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1547, + "step": 31958 + }, + { + "epoch": 1.6215939849624061, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 31959 + }, + { + "epoch": 1.6216541353383458, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 31960 + }, + { + "epoch": 1.6217142857142857, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 31961 + }, + { + "epoch": 1.6217744360902255, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 31962 + }, + { + "epoch": 1.6218345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 31963 + }, + { + "epoch": 1.6218947368421053, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 31964 + }, + { + "epoch": 1.6219548872180451, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 31965 + }, + { + "epoch": 1.622015037593985, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 31966 + }, + { + "epoch": 1.6220751879699247, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 31967 + }, + { + "epoch": 1.6221353383458648, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 31968 + }, + { + "epoch": 1.6221954887218044, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 31969 + }, + { + "epoch": 1.6222556390977445, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 31970 + }, + { + "epoch": 1.6223157894736842, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 31971 + }, + { + "epoch": 1.622375939849624, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 31972 + }, + { + "epoch": 1.622436090225564, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 31973 + }, + { + "epoch": 1.6224962406015038, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 31974 + }, + { + "epoch": 1.6225563909774436, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 31975 + }, + { + "epoch": 1.6226165413533835, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 31976 + }, + { + "epoch": 1.6226766917293234, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 31977 + }, + { + "epoch": 1.622736842105263, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 31978 + }, + { + "epoch": 1.6227969924812031, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 31979 + }, + { + "epoch": 1.6228571428571428, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 31980 + }, + { + "epoch": 1.6229172932330829, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 31981 + }, + { + "epoch": 1.6229774436090225, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 31982 + }, + { + "epoch": 1.6230375939849624, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 31983 + }, + { + "epoch": 1.6230977443609023, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 31984 + }, + { + "epoch": 1.6231578947368421, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 31985 + }, + { + "epoch": 1.623218045112782, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 31986 + }, + { + "epoch": 1.6232781954887217, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 31987 + }, + { + "epoch": 1.6233383458646617, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 31988 + }, + { + "epoch": 1.6233984962406014, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.204, + "step": 31989 + }, + { + "epoch": 1.6234586466165415, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 31990 + }, + { + "epoch": 1.6235187969924811, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 31991 + }, + { + "epoch": 1.623578947368421, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 31992 + }, + { + "epoch": 1.6236390977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 31993 + }, + { + "epoch": 1.6236992481203008, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 31994 + }, + { + "epoch": 1.6237593984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 31995 + }, + { + "epoch": 1.6238195488721805, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 31996 + }, + { + "epoch": 1.6238796992481204, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 31997 + }, + { + "epoch": 1.62393984962406, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 31998 + }, + { + "epoch": 1.624, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 31999 + }, + { + "epoch": 1.6240601503759398, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 32000 + }, + { + "epoch": 1.6241203007518799, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 32001 + }, + { + "epoch": 1.6241804511278195, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 32002 + }, + { + "epoch": 1.6242406015037594, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 32003 + }, + { + "epoch": 1.6243007518796992, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 32004 + }, + { + "epoch": 1.6243609022556391, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 32005 + }, + { + "epoch": 1.624421052631579, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 32006 + }, + { + "epoch": 1.6244812030075186, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 32007 + }, + { + "epoch": 1.6245413533834587, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9533, + "step": 32008 + }, + { + "epoch": 1.6246015037593984, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 32009 + }, + { + "epoch": 1.6246616541353385, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2797, + "step": 32010 + }, + { + "epoch": 1.6247218045112781, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 32011 + }, + { + "epoch": 1.624781954887218, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9263, + "step": 32012 + }, + { + "epoch": 1.6248421052631579, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 32013 + }, + { + "epoch": 1.6249022556390977, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 32014 + }, + { + "epoch": 1.6249624060150376, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 32015 + }, + { + "epoch": 1.6250225563909775, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1615, + "step": 32016 + }, + { + "epoch": 1.6250827067669174, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 32017 + }, + { + "epoch": 1.625142857142857, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 32018 + }, + { + "epoch": 1.625203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 32019 + }, + { + "epoch": 1.6252631578947367, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2408, + "step": 32020 + }, + { + "epoch": 1.6253233082706768, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 32021 + }, + { + "epoch": 1.6253834586466165, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 32022 + }, + { + "epoch": 1.6254436090225564, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 32023 + }, + { + "epoch": 1.6255037593984962, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 32024 + }, + { + "epoch": 1.625563909774436, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 32025 + }, + { + "epoch": 1.625624060150376, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 32026 + }, + { + "epoch": 1.6256842105263158, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 32027 + }, + { + "epoch": 1.6257443609022557, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 32028 + }, + { + "epoch": 1.6258045112781954, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 32029 + }, + { + "epoch": 1.6258646616541355, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1734, + "step": 32030 + }, + { + "epoch": 1.625924812030075, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 32031 + }, + { + "epoch": 1.6259849624060152, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 32032 + }, + { + "epoch": 1.6260451127819548, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 32033 + }, + { + "epoch": 1.6261052631578947, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 32034 + }, + { + "epoch": 1.6261654135338346, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 32035 + }, + { + "epoch": 1.6262255639097745, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.2649, + "step": 32036 + }, + { + "epoch": 1.6262857142857143, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 32037 + }, + { + "epoch": 1.626345864661654, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 32038 + }, + { + "epoch": 1.626406015037594, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 32039 + }, + { + "epoch": 1.6264661654135337, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 32040 + }, + { + "epoch": 1.6265263157894738, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 32041 + }, + { + "epoch": 1.6265864661654135, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 32042 + }, + { + "epoch": 1.6266466165413533, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1636, + "step": 32043 + }, + { + "epoch": 1.6267067669172932, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 32044 + }, + { + "epoch": 1.626766917293233, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 32045 + }, + { + "epoch": 1.626827067669173, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 32046 + }, + { + "epoch": 1.6268872180451128, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 32047 + }, + { + "epoch": 1.6269473684210527, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 32048 + }, + { + "epoch": 1.6270075187969923, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 32049 + }, + { + "epoch": 1.6270676691729324, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 32050 + }, + { + "epoch": 1.627127819548872, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 32051 + }, + { + "epoch": 1.6271879699248122, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 32052 + }, + { + "epoch": 1.6272481203007518, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 32053 + }, + { + "epoch": 1.6273082706766917, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 32054 + }, + { + "epoch": 1.6273684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 32055 + }, + { + "epoch": 1.6274285714285714, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 32056 + }, + { + "epoch": 1.6274887218045113, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 32057 + }, + { + "epoch": 1.627548872180451, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 32058 + }, + { + "epoch": 1.627609022556391, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 32059 + }, + { + "epoch": 1.6276691729323307, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.211, + "step": 32060 + }, + { + "epoch": 1.6277293233082708, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 32061 + }, + { + "epoch": 1.6277894736842105, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 32062 + }, + { + "epoch": 1.6278496240601503, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 32063 + }, + { + "epoch": 1.6279097744360902, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 32064 + }, + { + "epoch": 1.62796992481203, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 32065 + }, + { + "epoch": 1.62803007518797, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 32066 + }, + { + "epoch": 1.6280902255639098, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 32067 + }, + { + "epoch": 1.6281503759398497, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 32068 + }, + { + "epoch": 1.6282105263157893, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 32069 + }, + { + "epoch": 1.6282706766917294, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 32070 + }, + { + "epoch": 1.628330827067669, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 32071 + }, + { + "epoch": 1.6283909774436092, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 32072 + }, + { + "epoch": 1.6284511278195488, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 32073 + }, + { + "epoch": 1.6285112781954887, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 32074 + }, + { + "epoch": 1.6285714285714286, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 32075 + }, + { + "epoch": 1.6286315789473684, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 32076 + }, + { + "epoch": 1.6286917293233083, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 32077 + }, + { + "epoch": 1.6287518796992482, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 32078 + }, + { + "epoch": 1.628812030075188, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 32079 + }, + { + "epoch": 1.6288721804511277, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 32080 + }, + { + "epoch": 1.6289323308270678, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 32081 + }, + { + "epoch": 1.6289924812030074, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 32082 + }, + { + "epoch": 1.6290526315789475, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 32083 + }, + { + "epoch": 1.6291127819548872, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 32084 + }, + { + "epoch": 1.629172932330827, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 32085 + }, + { + "epoch": 1.629233082706767, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 32086 + }, + { + "epoch": 1.6292932330827068, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 32087 + }, + { + "epoch": 1.6293533834586467, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 32088 + }, + { + "epoch": 1.6294135338345863, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 32089 + }, + { + "epoch": 1.6294736842105264, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 32090 + }, + { + "epoch": 1.629533834586466, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 32091 + }, + { + "epoch": 1.6295939849624062, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 32092 + }, + { + "epoch": 1.6296541353383458, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 32093 + }, + { + "epoch": 1.6297142857142857, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 32094 + }, + { + "epoch": 1.6297744360902255, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 32095 + }, + { + "epoch": 1.6298345864661654, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 32096 + }, + { + "epoch": 1.6298947368421053, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 32097 + }, + { + "epoch": 1.6299548872180452, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 32098 + }, + { + "epoch": 1.630015037593985, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 32099 + }, + { + "epoch": 1.6300751879699247, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 32100 + }, + { + "epoch": 1.6301353383458648, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 32101 + }, + { + "epoch": 1.6301954887218044, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 32102 + }, + { + "epoch": 1.6302556390977445, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 32103 + }, + { + "epoch": 1.6303157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 32104 + }, + { + "epoch": 1.630375939849624, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 32105 + }, + { + "epoch": 1.630436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 32106 + }, + { + "epoch": 1.6304962406015038, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 32107 + }, + { + "epoch": 1.6305563909774436, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 32108 + }, + { + "epoch": 1.6306165413533833, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 32109 + }, + { + "epoch": 1.6306766917293234, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 32110 + }, + { + "epoch": 1.630736842105263, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 32111 + }, + { + "epoch": 1.6307969924812031, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 32112 + }, + { + "epoch": 1.6308571428571428, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 32113 + }, + { + "epoch": 1.6309172932330827, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 32114 + }, + { + "epoch": 1.6309774436090225, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 32115 + }, + { + "epoch": 1.6310375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 32116 + }, + { + "epoch": 1.6310977443609023, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 32117 + }, + { + "epoch": 1.6311578947368421, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 32118 + }, + { + "epoch": 1.631218045112782, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 32119 + }, + { + "epoch": 1.6312781954887217, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 32120 + }, + { + "epoch": 1.6313383458646618, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 32121 + }, + { + "epoch": 1.6313984962406014, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 32122 + }, + { + "epoch": 1.6314586466165415, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 32123 + }, + { + "epoch": 1.6315187969924811, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 32124 + }, + { + "epoch": 1.631578947368421, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 32125 + }, + { + "epoch": 1.6316390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 32126 + }, + { + "epoch": 1.6316992481203008, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 32127 + }, + { + "epoch": 1.6317593984962406, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 32128 + }, + { + "epoch": 1.6318195488721805, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 32129 + }, + { + "epoch": 1.6318796992481204, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 32130 + }, + { + "epoch": 1.63193984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 32131 + }, + { + "epoch": 1.6320000000000001, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 32132 + }, + { + "epoch": 1.6320601503759398, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 32133 + }, + { + "epoch": 1.6321203007518799, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 32134 + }, + { + "epoch": 1.6321804511278195, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 32135 + }, + { + "epoch": 1.6322406015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 32136 + }, + { + "epoch": 1.6323007518796993, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 32137 + }, + { + "epoch": 1.6323609022556391, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 32138 + }, + { + "epoch": 1.632421052631579, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 32139 + }, + { + "epoch": 1.6324812030075186, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 32140 + }, + { + "epoch": 1.6325413533834587, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 32141 + }, + { + "epoch": 1.6326015037593984, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 32142 + }, + { + "epoch": 1.6326616541353385, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 32143 + }, + { + "epoch": 1.6327218045112781, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 32144 + }, + { + "epoch": 1.632781954887218, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 32145 + }, + { + "epoch": 1.6328421052631579, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 32146 + }, + { + "epoch": 1.6329022556390977, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 32147 + }, + { + "epoch": 1.6329624060150376, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 32148 + }, + { + "epoch": 1.6330225563909775, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 32149 + }, + { + "epoch": 1.6330827067669174, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.2237, + "step": 32150 + }, + { + "epoch": 1.633142857142857, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 32151 + }, + { + "epoch": 1.633203007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 32152 + }, + { + "epoch": 1.6332631578947368, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 32153 + }, + { + "epoch": 1.6333233082706768, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 32154 + }, + { + "epoch": 1.6333834586466165, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 32155 + }, + { + "epoch": 1.6334436090225564, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 32156 + }, + { + "epoch": 1.6335037593984962, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 32157 + }, + { + "epoch": 1.633563909774436, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 32158 + }, + { + "epoch": 1.633624060150376, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 32159 + }, + { + "epoch": 1.6336842105263156, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 32160 + }, + { + "epoch": 1.6337443609022557, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 32161 + }, + { + "epoch": 1.6338045112781954, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 32162 + }, + { + "epoch": 1.6338646616541355, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 32163 + }, + { + "epoch": 1.6339248120300751, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 32164 + }, + { + "epoch": 1.633984962406015, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 32165 + }, + { + "epoch": 1.6340451127819549, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 32166 + }, + { + "epoch": 1.6341052631578947, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 32167 + }, + { + "epoch": 1.6341654135338346, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 32168 + }, + { + "epoch": 1.6342255639097745, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 32169 + }, + { + "epoch": 1.6342857142857143, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 32170 + }, + { + "epoch": 1.634345864661654, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 32171 + }, + { + "epoch": 1.634406015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 32172 + }, + { + "epoch": 1.6344661654135337, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 32173 + }, + { + "epoch": 1.6345263157894738, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 32174 + }, + { + "epoch": 1.6345864661654135, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 32175 + }, + { + "epoch": 1.6346466165413533, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 32176 + }, + { + "epoch": 1.6347067669172932, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 32177 + }, + { + "epoch": 1.634766917293233, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 32178 + }, + { + "epoch": 1.634827067669173, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 32179 + }, + { + "epoch": 1.6348872180451128, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 32180 + }, + { + "epoch": 1.6349473684210527, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 32181 + }, + { + "epoch": 1.6350075187969924, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 32182 + }, + { + "epoch": 1.6350676691729324, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 32183 + }, + { + "epoch": 1.635127819548872, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 32184 + }, + { + "epoch": 1.6351879699248122, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 32185 + }, + { + "epoch": 1.6352481203007518, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 32186 + }, + { + "epoch": 1.6353082706766917, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1705, + "step": 32187 + }, + { + "epoch": 1.6353684210526316, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 32188 + }, + { + "epoch": 1.6354285714285715, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 32189 + }, + { + "epoch": 1.6354887218045113, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 32190 + }, + { + "epoch": 1.635548872180451, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1849, + "step": 32191 + }, + { + "epoch": 1.635609022556391, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 32192 + }, + { + "epoch": 1.6356691729323307, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 32193 + }, + { + "epoch": 1.6357293233082708, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 32194 + }, + { + "epoch": 1.6357894736842105, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1608, + "step": 32195 + }, + { + "epoch": 1.6358496240601503, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 32196 + }, + { + "epoch": 1.6359097744360902, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 32197 + }, + { + "epoch": 1.63596992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 32198 + }, + { + "epoch": 1.63603007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 32199 + }, + { + "epoch": 1.6360902255639098, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 32200 + }, + { + "epoch": 1.6361503759398497, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 32201 + }, + { + "epoch": 1.6362105263157893, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 32202 + }, + { + "epoch": 1.6362706766917294, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 32203 + }, + { + "epoch": 1.636330827067669, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 32204 + }, + { + "epoch": 1.6363909774436092, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 32205 + }, + { + "epoch": 1.6364511278195488, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 32206 + }, + { + "epoch": 1.6365112781954887, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 32207 + }, + { + "epoch": 1.6365714285714286, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 32208 + }, + { + "epoch": 1.6366315789473684, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 32209 + }, + { + "epoch": 1.6366917293233083, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 32210 + }, + { + "epoch": 1.636751879699248, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 32211 + }, + { + "epoch": 1.636812030075188, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 32212 + }, + { + "epoch": 1.6368721804511277, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 32213 + }, + { + "epoch": 1.6369323308270678, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 32214 + }, + { + "epoch": 1.6369924812030074, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 32215 + }, + { + "epoch": 1.6370526315789473, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 32216 + }, + { + "epoch": 1.6371127819548872, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 32217 + }, + { + "epoch": 1.637172932330827, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 32218 + }, + { + "epoch": 1.637233082706767, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 32219 + }, + { + "epoch": 1.6372932330827068, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 32220 + }, + { + "epoch": 1.6373533834586467, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 32221 + }, + { + "epoch": 1.6374135338345863, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 32222 + }, + { + "epoch": 1.6374736842105264, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9199, + "step": 32223 + }, + { + "epoch": 1.637533834586466, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 32224 + }, + { + "epoch": 1.6375939849624062, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 32225 + }, + { + "epoch": 1.6376541353383458, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 32226 + }, + { + "epoch": 1.6377142857142857, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 32227 + }, + { + "epoch": 1.6377744360902255, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 32228 + }, + { + "epoch": 1.6378345864661654, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 32229 + }, + { + "epoch": 1.6378947368421053, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 32230 + }, + { + "epoch": 1.6379548872180452, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 32231 + }, + { + "epoch": 1.638015037593985, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1357, + "step": 32232 + }, + { + "epoch": 1.6380751879699247, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 32233 + }, + { + "epoch": 1.6381353383458648, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 32234 + }, + { + "epoch": 1.6381954887218044, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 32235 + }, + { + "epoch": 1.6382556390977445, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 32236 + }, + { + "epoch": 1.6383157894736842, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 32237 + }, + { + "epoch": 1.638375939849624, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 32238 + }, + { + "epoch": 1.638436090225564, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 32239 + }, + { + "epoch": 1.6384962406015038, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 32240 + }, + { + "epoch": 1.6385563909774437, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 32241 + }, + { + "epoch": 1.6386165413533833, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1357, + "step": 32242 + }, + { + "epoch": 1.6386766917293234, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 32243 + }, + { + "epoch": 1.638736842105263, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 32244 + }, + { + "epoch": 1.6387969924812031, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 32245 + }, + { + "epoch": 1.6388571428571428, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 32246 + }, + { + "epoch": 1.6389172932330827, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 32247 + }, + { + "epoch": 1.6389774436090225, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9352, + "step": 32248 + }, + { + "epoch": 1.6390375939849624, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 32249 + }, + { + "epoch": 1.6390977443609023, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 32250 + }, + { + "epoch": 1.6391578947368421, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 32251 + }, + { + "epoch": 1.639218045112782, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 32252 + }, + { + "epoch": 1.6392781954887217, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 32253 + }, + { + "epoch": 1.6393383458646618, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 32254 + }, + { + "epoch": 1.6393984962406014, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 32255 + }, + { + "epoch": 1.6394586466165415, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 32256 + }, + { + "epoch": 1.6395187969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 32257 + }, + { + "epoch": 1.639578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 32258 + }, + { + "epoch": 1.639639097744361, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1481, + "step": 32259 + }, + { + "epoch": 1.6396992481203008, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1326, + "step": 32260 + }, + { + "epoch": 1.6397593984962406, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 32261 + }, + { + "epoch": 1.6398195488721803, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 32262 + }, + { + "epoch": 1.6398796992481204, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 32263 + }, + { + "epoch": 1.63993984962406, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 32264 + }, + { + "epoch": 1.6400000000000001, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 32265 + }, + { + "epoch": 1.6400601503759398, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 32266 + }, + { + "epoch": 1.6401203007518796, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 32267 + }, + { + "epoch": 1.6401804511278195, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 32268 + }, + { + "epoch": 1.6402406015037594, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 32269 + }, + { + "epoch": 1.6403007518796993, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 32270 + }, + { + "epoch": 1.6403609022556391, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 32271 + }, + { + "epoch": 1.640421052631579, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 32272 + }, + { + "epoch": 1.6404812030075187, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 32273 + }, + { + "epoch": 1.6405413533834587, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 32274 + }, + { + "epoch": 1.6406015037593984, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 32275 + }, + { + "epoch": 1.6406616541353385, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 32276 + }, + { + "epoch": 1.6407218045112781, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 32277 + }, + { + "epoch": 1.640781954887218, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 32278 + }, + { + "epoch": 1.6408421052631579, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 32279 + }, + { + "epoch": 1.6409022556390978, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 32280 + }, + { + "epoch": 1.6409624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 32281 + }, + { + "epoch": 1.6410225563909775, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 32282 + }, + { + "epoch": 1.6410827067669174, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 32283 + }, + { + "epoch": 1.641142857142857, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 32284 + }, + { + "epoch": 1.641203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 32285 + }, + { + "epoch": 1.6412631578947368, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 32286 + }, + { + "epoch": 1.6413233082706769, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 32287 + }, + { + "epoch": 1.6413834586466165, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 32288 + }, + { + "epoch": 1.6414436090225564, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 32289 + }, + { + "epoch": 1.6415037593984962, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 32290 + }, + { + "epoch": 1.6415639097744361, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9083, + "step": 32291 + }, + { + "epoch": 1.641624060150376, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 32292 + }, + { + "epoch": 1.6416842105263156, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 32293 + }, + { + "epoch": 1.6417443609022557, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 32294 + }, + { + "epoch": 1.6418045112781954, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 32295 + }, + { + "epoch": 1.6418646616541355, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 32296 + }, + { + "epoch": 1.6419248120300751, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 32297 + }, + { + "epoch": 1.641984962406015, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 32298 + }, + { + "epoch": 1.6420451127819549, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 32299 + }, + { + "epoch": 1.6421052631578947, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 32300 + }, + { + "epoch": 1.6421654135338346, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 32301 + }, + { + "epoch": 1.6422255639097745, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1869, + "step": 32302 + }, + { + "epoch": 1.6422857142857143, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 32303 + }, + { + "epoch": 1.642345864661654, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9447, + "step": 32304 + }, + { + "epoch": 1.642406015037594, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 32305 + }, + { + "epoch": 1.6424661654135337, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 32306 + }, + { + "epoch": 1.6425263157894738, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 32307 + }, + { + "epoch": 1.6425864661654135, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 32308 + }, + { + "epoch": 1.6426466165413534, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 32309 + }, + { + "epoch": 1.6427067669172932, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 32310 + }, + { + "epoch": 1.642766917293233, + "grad_norm": 2.234375, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 32311 + }, + { + "epoch": 1.642827067669173, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 32312 + }, + { + "epoch": 1.6428872180451126, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 32313 + }, + { + "epoch": 1.6429473684210527, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 32314 + }, + { + "epoch": 1.6430075187969924, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 32315 + }, + { + "epoch": 1.6430676691729325, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 32316 + }, + { + "epoch": 1.643127819548872, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.207, + "step": 32317 + }, + { + "epoch": 1.643187969924812, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 32318 + }, + { + "epoch": 1.6432481203007518, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 32319 + }, + { + "epoch": 1.6433082706766917, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 32320 + }, + { + "epoch": 1.6433684210526316, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 32321 + }, + { + "epoch": 1.6434285714285715, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 32322 + }, + { + "epoch": 1.6434887218045113, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 32323 + }, + { + "epoch": 1.643548872180451, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 32324 + }, + { + "epoch": 1.643609022556391, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 32325 + }, + { + "epoch": 1.6436691729323307, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 32326 + }, + { + "epoch": 1.6437293233082708, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 32327 + }, + { + "epoch": 1.6437894736842105, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 32328 + }, + { + "epoch": 1.6438496240601503, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 32329 + }, + { + "epoch": 1.6439097744360902, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 32330 + }, + { + "epoch": 1.64396992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9366, + "step": 32331 + }, + { + "epoch": 1.64403007518797, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 32332 + }, + { + "epoch": 1.6440902255639098, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9373, + "step": 32333 + }, + { + "epoch": 1.6441503759398497, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 32334 + }, + { + "epoch": 1.6442105263157893, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 32335 + }, + { + "epoch": 1.6442706766917294, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 32336 + }, + { + "epoch": 1.644330827067669, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 32337 + }, + { + "epoch": 1.6443909774436092, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 32338 + }, + { + "epoch": 1.6444511278195488, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 32339 + }, + { + "epoch": 1.6445112781954887, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 32340 + }, + { + "epoch": 1.6445714285714286, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 32341 + }, + { + "epoch": 1.6446315789473684, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 32342 + }, + { + "epoch": 1.6446917293233083, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 32343 + }, + { + "epoch": 1.644751879699248, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 32344 + }, + { + "epoch": 1.644812030075188, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 32345 + }, + { + "epoch": 1.6448721804511277, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 32346 + }, + { + "epoch": 1.6449323308270678, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 32347 + }, + { + "epoch": 1.6449924812030075, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 32348 + }, + { + "epoch": 1.6450526315789473, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 32349 + }, + { + "epoch": 1.6451127819548872, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 32350 + }, + { + "epoch": 1.645172932330827, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 32351 + }, + { + "epoch": 1.645233082706767, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 32352 + }, + { + "epoch": 1.6452932330827068, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 32353 + }, + { + "epoch": 1.6453533834586467, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 32354 + }, + { + "epoch": 1.6454135338345863, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 32355 + }, + { + "epoch": 1.6454736842105264, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1587, + "step": 32356 + }, + { + "epoch": 1.645533834586466, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 32357 + }, + { + "epoch": 1.6455939849624062, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 32358 + }, + { + "epoch": 1.6456541353383458, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 32359 + }, + { + "epoch": 1.6457142857142857, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 32360 + }, + { + "epoch": 1.6457744360902256, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 32361 + }, + { + "epoch": 1.6458345864661654, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 32362 + }, + { + "epoch": 1.6458947368421053, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 32363 + }, + { + "epoch": 1.645954887218045, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 32364 + }, + { + "epoch": 1.646015037593985, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 32365 + }, + { + "epoch": 1.6460751879699247, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 32366 + }, + { + "epoch": 1.6461353383458648, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 32367 + }, + { + "epoch": 1.6461954887218044, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 32368 + }, + { + "epoch": 1.6462556390977443, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 32369 + }, + { + "epoch": 1.6463157894736842, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 32370 + }, + { + "epoch": 1.646375939849624, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 32371 + }, + { + "epoch": 1.646436090225564, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 32372 + }, + { + "epoch": 1.6464962406015038, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 32373 + }, + { + "epoch": 1.6465563909774437, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 32374 + }, + { + "epoch": 1.6466165413533833, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 32375 + }, + { + "epoch": 1.6466766917293234, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 32376 + }, + { + "epoch": 1.646736842105263, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 32377 + }, + { + "epoch": 1.6467969924812031, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 32378 + }, + { + "epoch": 1.6468571428571428, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 32379 + }, + { + "epoch": 1.6469172932330827, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 32380 + }, + { + "epoch": 1.6469774436090225, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 32381 + }, + { + "epoch": 1.6470375939849624, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 32382 + }, + { + "epoch": 1.6470977443609023, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 32383 + }, + { + "epoch": 1.6471578947368422, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 32384 + }, + { + "epoch": 1.647218045112782, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 32385 + }, + { + "epoch": 1.6472781954887217, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 32386 + }, + { + "epoch": 1.6473383458646618, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 32387 + }, + { + "epoch": 1.6473984962406014, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 32388 + }, + { + "epoch": 1.6474586466165415, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 32389 + }, + { + "epoch": 1.6475187969924812, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 32390 + }, + { + "epoch": 1.647578947368421, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 32391 + }, + { + "epoch": 1.647639097744361, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2688, + "step": 32392 + }, + { + "epoch": 1.6476992481203008, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 32393 + }, + { + "epoch": 1.6477593984962406, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 32394 + }, + { + "epoch": 1.6478195488721803, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 32395 + }, + { + "epoch": 1.6478796992481204, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 32396 + }, + { + "epoch": 1.64793984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 32397 + }, + { + "epoch": 1.6480000000000001, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 32398 + }, + { + "epoch": 1.6480601503759398, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 32399 + }, + { + "epoch": 1.6481203007518797, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 32400 + }, + { + "epoch": 1.6481804511278195, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 32401 + }, + { + "epoch": 1.6482406015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 32402 + }, + { + "epoch": 1.6483007518796993, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 32403 + }, + { + "epoch": 1.6483609022556391, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 32404 + }, + { + "epoch": 1.648421052631579, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 32405 + }, + { + "epoch": 1.6484812030075187, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 32406 + }, + { + "epoch": 1.6485413533834588, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 32407 + }, + { + "epoch": 1.6486015037593984, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 32408 + }, + { + "epoch": 1.6486616541353385, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 32409 + }, + { + "epoch": 1.6487218045112781, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 32410 + }, + { + "epoch": 1.648781954887218, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 32411 + }, + { + "epoch": 1.6488421052631579, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 32412 + }, + { + "epoch": 1.6489022556390978, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 32413 + }, + { + "epoch": 1.6489624060150376, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 32414 + }, + { + "epoch": 1.6490225563909773, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 32415 + }, + { + "epoch": 1.6490827067669174, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 32416 + }, + { + "epoch": 1.649142857142857, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 32417 + }, + { + "epoch": 1.6492030075187971, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 32418 + }, + { + "epoch": 1.6492631578947368, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 32419 + }, + { + "epoch": 1.6493233082706769, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 32420 + }, + { + "epoch": 1.6493834586466165, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 32421 + }, + { + "epoch": 1.6494436090225564, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 32422 + }, + { + "epoch": 1.6495037593984962, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 32423 + }, + { + "epoch": 1.6495639097744361, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 32424 + }, + { + "epoch": 1.649624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 32425 + }, + { + "epoch": 1.6496842105263156, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 32426 + }, + { + "epoch": 1.6497443609022557, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 32427 + }, + { + "epoch": 1.6498045112781954, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 32428 + }, + { + "epoch": 1.6498646616541355, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 32429 + }, + { + "epoch": 1.6499248120300751, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 32430 + }, + { + "epoch": 1.649984962406015, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 32431 + }, + { + "epoch": 1.6500451127819549, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 32432 + }, + { + "epoch": 1.6501052631578947, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 32433 + }, + { + "epoch": 1.6501654135338346, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 32434 + }, + { + "epoch": 1.6502255639097745, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 32435 + }, + { + "epoch": 1.6502857142857144, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 32436 + }, + { + "epoch": 1.650345864661654, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 32437 + }, + { + "epoch": 1.650406015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 32438 + }, + { + "epoch": 1.6504661654135337, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 32439 + }, + { + "epoch": 1.6505263157894738, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 32440 + }, + { + "epoch": 1.6505864661654135, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 32441 + }, + { + "epoch": 1.6506466165413534, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 32442 + }, + { + "epoch": 1.6507067669172932, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 32443 + }, + { + "epoch": 1.650766917293233, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 32444 + }, + { + "epoch": 1.650827067669173, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 32445 + }, + { + "epoch": 1.6508872180451126, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 32446 + }, + { + "epoch": 1.6509473684210527, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 32447 + }, + { + "epoch": 1.6510075187969924, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 32448 + }, + { + "epoch": 1.6510676691729325, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 32449 + }, + { + "epoch": 1.651127819548872, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 32450 + }, + { + "epoch": 1.651187969924812, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 32451 + }, + { + "epoch": 1.6512481203007519, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 32452 + }, + { + "epoch": 1.6513082706766917, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 32453 + }, + { + "epoch": 1.6513684210526316, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 32454 + }, + { + "epoch": 1.6514285714285715, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 32455 + }, + { + "epoch": 1.6514887218045113, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 32456 + }, + { + "epoch": 1.651548872180451, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 32457 + }, + { + "epoch": 1.651609022556391, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 32458 + }, + { + "epoch": 1.6516691729323307, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 32459 + }, + { + "epoch": 1.6517293233082708, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1264, + "step": 32460 + }, + { + "epoch": 1.6517894736842105, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 32461 + }, + { + "epoch": 1.6518496240601503, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 32462 + }, + { + "epoch": 1.6519097744360902, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 32463 + }, + { + "epoch": 1.65196992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 32464 + }, + { + "epoch": 1.65203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 32465 + }, + { + "epoch": 1.6520902255639096, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 32466 + }, + { + "epoch": 1.6521503759398497, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 32467 + }, + { + "epoch": 1.6522105263157894, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 32468 + }, + { + "epoch": 1.6522706766917294, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 32469 + }, + { + "epoch": 1.652330827067669, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 32470 + }, + { + "epoch": 1.6523909774436092, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 32471 + }, + { + "epoch": 1.6524511278195488, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 32472 + }, + { + "epoch": 1.6525112781954887, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 32473 + }, + { + "epoch": 1.6525714285714286, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 32474 + }, + { + "epoch": 1.6526315789473685, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 32475 + }, + { + "epoch": 1.6526917293233083, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 32476 + }, + { + "epoch": 1.652751879699248, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 32477 + }, + { + "epoch": 1.652812030075188, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 32478 + }, + { + "epoch": 1.6528721804511277, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 32479 + }, + { + "epoch": 1.6529323308270678, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 32480 + }, + { + "epoch": 1.6529924812030075, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 32481 + }, + { + "epoch": 1.6530526315789473, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 32482 + }, + { + "epoch": 1.6531127819548872, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 32483 + }, + { + "epoch": 1.653172932330827, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 32484 + }, + { + "epoch": 1.653233082706767, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1949, + "step": 32485 + }, + { + "epoch": 1.6532932330827068, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 32486 + }, + { + "epoch": 1.6533533834586467, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 32487 + }, + { + "epoch": 1.6534135338345863, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 32488 + }, + { + "epoch": 1.6534736842105264, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 32489 + }, + { + "epoch": 1.653533834586466, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 32490 + }, + { + "epoch": 1.6535939849624062, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 32491 + }, + { + "epoch": 1.6536541353383458, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 32492 + }, + { + "epoch": 1.6537142857142857, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 32493 + }, + { + "epoch": 1.6537744360902256, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 32494 + }, + { + "epoch": 1.6538345864661654, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 32495 + }, + { + "epoch": 1.6538947368421053, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.3818, + "step": 32496 + }, + { + "epoch": 1.653954887218045, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 32497 + }, + { + "epoch": 1.654015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 32498 + }, + { + "epoch": 1.6540751879699247, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 32499 + }, + { + "epoch": 1.6541353383458648, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 32500 + }, + { + "epoch": 1.6541954887218044, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 32501 + }, + { + "epoch": 1.6542556390977443, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 32502 + }, + { + "epoch": 1.6543157894736842, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 32503 + }, + { + "epoch": 1.654375939849624, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 32504 + }, + { + "epoch": 1.654436090225564, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 32505 + }, + { + "epoch": 1.6544962406015038, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 32506 + }, + { + "epoch": 1.6545563909774437, + "grad_norm": 1.390625, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 32507 + }, + { + "epoch": 1.6546165413533833, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 32508 + }, + { + "epoch": 1.6546766917293234, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 32509 + }, + { + "epoch": 1.654736842105263, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 32510 + }, + { + "epoch": 1.6547969924812032, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 32511 + }, + { + "epoch": 1.6548571428571428, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 32512 + }, + { + "epoch": 1.6549172932330827, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 32513 + }, + { + "epoch": 1.6549774436090225, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 32514 + }, + { + "epoch": 1.6550375939849624, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 32515 + }, + { + "epoch": 1.6550977443609023, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 32516 + }, + { + "epoch": 1.655157894736842, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 32517 + }, + { + "epoch": 1.655218045112782, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 32518 + }, + { + "epoch": 1.6552781954887217, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 32519 + }, + { + "epoch": 1.6553383458646618, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 32520 + }, + { + "epoch": 1.6553984962406014, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 32521 + }, + { + "epoch": 1.6554586466165415, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 32522 + }, + { + "epoch": 1.6555187969924812, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 32523 + }, + { + "epoch": 1.655578947368421, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 32524 + }, + { + "epoch": 1.655639097744361, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 32525 + }, + { + "epoch": 1.6556992481203008, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 32526 + }, + { + "epoch": 1.6557593984962407, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 32527 + }, + { + "epoch": 1.6558195488721803, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 32528 + }, + { + "epoch": 1.6558796992481204, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 32529 + }, + { + "epoch": 1.65593984962406, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 32530 + }, + { + "epoch": 1.6560000000000001, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 32531 + }, + { + "epoch": 1.6560601503759398, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 32532 + }, + { + "epoch": 1.6561203007518797, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 32533 + }, + { + "epoch": 1.6561804511278195, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 32534 + }, + { + "epoch": 1.6562406015037594, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 32535 + }, + { + "epoch": 1.6563007518796993, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 32536 + }, + { + "epoch": 1.6563609022556391, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1498, + "step": 32537 + }, + { + "epoch": 1.656421052631579, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9206, + "step": 32538 + }, + { + "epoch": 1.6564812030075187, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 32539 + }, + { + "epoch": 1.6565413533834588, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 32540 + }, + { + "epoch": 1.6566015037593984, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 32541 + }, + { + "epoch": 1.6566616541353385, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 32542 + }, + { + "epoch": 1.6567218045112782, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 32543 + }, + { + "epoch": 1.656781954887218, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 32544 + }, + { + "epoch": 1.656842105263158, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1418, + "step": 32545 + }, + { + "epoch": 1.6569022556390978, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 32546 + }, + { + "epoch": 1.6569624060150376, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 32547 + }, + { + "epoch": 1.6570225563909773, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 32548 + }, + { + "epoch": 1.6570827067669174, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 32549 + }, + { + "epoch": 1.657142857142857, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 32550 + }, + { + "epoch": 1.6572030075187971, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 32551 + }, + { + "epoch": 1.6572631578947368, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 32552 + }, + { + "epoch": 1.6573233082706766, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 32553 + }, + { + "epoch": 1.6573834586466165, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 32554 + }, + { + "epoch": 1.6574436090225564, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 32555 + }, + { + "epoch": 1.6575037593984963, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 32556 + }, + { + "epoch": 1.6575639097744361, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 32557 + }, + { + "epoch": 1.657624060150376, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 32558 + }, + { + "epoch": 1.6576842105263156, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 32559 + }, + { + "epoch": 1.6577443609022557, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 32560 + }, + { + "epoch": 1.6578045112781954, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 32561 + }, + { + "epoch": 1.6578646616541355, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 32562 + }, + { + "epoch": 1.6579248120300751, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 32563 + }, + { + "epoch": 1.657984962406015, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9281, + "step": 32564 + }, + { + "epoch": 1.6580451127819549, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 32565 + }, + { + "epoch": 1.6581052631578947, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 32566 + }, + { + "epoch": 1.6581654135338346, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 32567 + }, + { + "epoch": 1.6582255639097745, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 32568 + }, + { + "epoch": 1.6582857142857144, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 32569 + }, + { + "epoch": 1.658345864661654, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 32570 + }, + { + "epoch": 1.658406015037594, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 32571 + }, + { + "epoch": 1.6584661654135338, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 32572 + }, + { + "epoch": 1.6585263157894738, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 32573 + }, + { + "epoch": 1.6585864661654135, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 32574 + }, + { + "epoch": 1.6586466165413534, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 32575 + }, + { + "epoch": 1.6587067669172932, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 32576 + }, + { + "epoch": 1.6587669172932331, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 32577 + }, + { + "epoch": 1.658827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 32578 + }, + { + "epoch": 1.6588872180451126, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 32579 + }, + { + "epoch": 1.6589473684210527, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 32580 + }, + { + "epoch": 1.6590075187969924, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2027, + "step": 32581 + }, + { + "epoch": 1.6590676691729325, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 32582 + }, + { + "epoch": 1.6591278195488721, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 32583 + }, + { + "epoch": 1.659187969924812, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 32584 + }, + { + "epoch": 1.6592481203007519, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 32585 + }, + { + "epoch": 1.6593082706766917, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1628, + "step": 32586 + }, + { + "epoch": 1.6593684210526316, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 32587 + }, + { + "epoch": 1.6594285714285715, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 32588 + }, + { + "epoch": 1.6594887218045113, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 32589 + }, + { + "epoch": 1.659548872180451, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 32590 + }, + { + "epoch": 1.659609022556391, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.146, + "step": 32591 + }, + { + "epoch": 1.6596691729323307, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 32592 + }, + { + "epoch": 1.6597293233082708, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 32593 + }, + { + "epoch": 1.6597894736842105, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 32594 + }, + { + "epoch": 1.6598496240601504, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 32595 + }, + { + "epoch": 1.6599097744360902, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 32596 + }, + { + "epoch": 1.65996992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 32597 + }, + { + "epoch": 1.66003007518797, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 32598 + }, + { + "epoch": 1.6600902255639096, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 32599 + }, + { + "epoch": 1.6601503759398497, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 32600 + }, + { + "epoch": 1.6602105263157894, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 32601 + }, + { + "epoch": 1.6602706766917295, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 32602 + }, + { + "epoch": 1.660330827067669, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 32603 + }, + { + "epoch": 1.660390977443609, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 32604 + }, + { + "epoch": 1.6604511278195488, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 32605 + }, + { + "epoch": 1.6605112781954887, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 32606 + }, + { + "epoch": 1.6605714285714286, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 32607 + }, + { + "epoch": 1.6606315789473685, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 32608 + }, + { + "epoch": 1.6606917293233083, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.164, + "step": 32609 + }, + { + "epoch": 1.660751879699248, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 32610 + }, + { + "epoch": 1.660812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 32611 + }, + { + "epoch": 1.6608721804511277, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 32612 + }, + { + "epoch": 1.6609323308270678, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 32613 + }, + { + "epoch": 1.6609924812030075, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 32614 + }, + { + "epoch": 1.6610526315789473, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1436, + "step": 32615 + }, + { + "epoch": 1.6611127819548872, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 32616 + }, + { + "epoch": 1.661172932330827, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 32617 + }, + { + "epoch": 1.661233082706767, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 32618 + }, + { + "epoch": 1.6612932330827068, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 32619 + }, + { + "epoch": 1.6613533834586467, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 32620 + }, + { + "epoch": 1.6614135338345863, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 32621 + }, + { + "epoch": 1.6614736842105264, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 32622 + }, + { + "epoch": 1.661533834586466, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 32623 + }, + { + "epoch": 1.6615939849624062, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 32624 + }, + { + "epoch": 1.6616541353383458, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 32625 + }, + { + "epoch": 1.6617142857142857, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1891, + "step": 32626 + }, + { + "epoch": 1.6617744360902256, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 32627 + }, + { + "epoch": 1.6618345864661654, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.8987, + "step": 32628 + }, + { + "epoch": 1.6618947368421053, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 32629 + }, + { + "epoch": 1.661954887218045, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 32630 + }, + { + "epoch": 1.662015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 32631 + }, + { + "epoch": 1.6620751879699247, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 32632 + }, + { + "epoch": 1.6621353383458648, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 32633 + }, + { + "epoch": 1.6621954887218044, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 32634 + }, + { + "epoch": 1.6622556390977443, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 32635 + }, + { + "epoch": 1.6623157894736842, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 32636 + }, + { + "epoch": 1.662375939849624, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 32637 + }, + { + "epoch": 1.662436090225564, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 32638 + }, + { + "epoch": 1.6624962406015038, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 32639 + }, + { + "epoch": 1.6625563909774437, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 32640 + }, + { + "epoch": 1.6626165413533833, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 32641 + }, + { + "epoch": 1.6626766917293234, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 32642 + }, + { + "epoch": 1.662736842105263, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 32643 + }, + { + "epoch": 1.6627969924812032, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 32644 + }, + { + "epoch": 1.6628571428571428, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 32645 + }, + { + "epoch": 1.6629172932330827, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 32646 + }, + { + "epoch": 1.6629774436090226, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 32647 + }, + { + "epoch": 1.6630375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 32648 + }, + { + "epoch": 1.6630977443609023, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 32649 + }, + { + "epoch": 1.663157894736842, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 32650 + }, + { + "epoch": 1.663218045112782, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 32651 + }, + { + "epoch": 1.6632781954887217, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 32652 + }, + { + "epoch": 1.6633383458646618, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 32653 + }, + { + "epoch": 1.6633984962406014, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 32654 + }, + { + "epoch": 1.6634586466165413, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 32655 + }, + { + "epoch": 1.6635187969924812, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 32656 + }, + { + "epoch": 1.663578947368421, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 32657 + }, + { + "epoch": 1.663639097744361, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 32658 + }, + { + "epoch": 1.6636992481203008, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 32659 + }, + { + "epoch": 1.6637593984962407, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 32660 + }, + { + "epoch": 1.6638195488721803, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 32661 + }, + { + "epoch": 1.6638796992481204, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 32662 + }, + { + "epoch": 1.66393984962406, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 32663 + }, + { + "epoch": 1.6640000000000001, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 32664 + }, + { + "epoch": 1.6640601503759398, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 32665 + }, + { + "epoch": 1.6641203007518797, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 32666 + }, + { + "epoch": 1.6641804511278195, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 32667 + }, + { + "epoch": 1.6642406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 32668 + }, + { + "epoch": 1.6643007518796993, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 32669 + }, + { + "epoch": 1.6643609022556392, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 32670 + }, + { + "epoch": 1.664421052631579, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 32671 + }, + { + "epoch": 1.6644812030075187, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 32672 + }, + { + "epoch": 1.6645413533834588, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 32673 + }, + { + "epoch": 1.6646015037593984, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 32674 + }, + { + "epoch": 1.6646616541353385, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 32675 + }, + { + "epoch": 1.6647218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 32676 + }, + { + "epoch": 1.664781954887218, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 32677 + }, + { + "epoch": 1.664842105263158, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 32678 + }, + { + "epoch": 1.6649022556390978, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 32679 + }, + { + "epoch": 1.6649624060150376, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 32680 + }, + { + "epoch": 1.6650225563909773, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 32681 + }, + { + "epoch": 1.6650827067669174, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 32682 + }, + { + "epoch": 1.665142857142857, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 32683 + }, + { + "epoch": 1.6652030075187971, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 32684 + }, + { + "epoch": 1.6652631578947368, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 32685 + }, + { + "epoch": 1.6653233082706766, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 32686 + }, + { + "epoch": 1.6653834586466165, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 32687 + }, + { + "epoch": 1.6654436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 32688 + }, + { + "epoch": 1.6655037593984963, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 32689 + }, + { + "epoch": 1.6655639097744361, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 32690 + }, + { + "epoch": 1.665624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 32691 + }, + { + "epoch": 1.6656842105263157, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 32692 + }, + { + "epoch": 1.6657443609022557, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 32693 + }, + { + "epoch": 1.6658045112781954, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 32694 + }, + { + "epoch": 1.6658646616541355, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9316, + "step": 32695 + }, + { + "epoch": 1.6659248120300751, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 32696 + }, + { + "epoch": 1.665984962406015, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 32697 + }, + { + "epoch": 1.6660451127819549, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 32698 + }, + { + "epoch": 1.6661052631578948, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1779, + "step": 32699 + }, + { + "epoch": 1.6661654135338346, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 32700 + }, + { + "epoch": 1.6662255639097743, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 32701 + }, + { + "epoch": 1.6662857142857144, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1602, + "step": 32702 + }, + { + "epoch": 1.666345864661654, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 32703 + }, + { + "epoch": 1.6664060150375941, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 32704 + }, + { + "epoch": 1.6664661654135338, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 32705 + }, + { + "epoch": 1.6665263157894736, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 32706 + }, + { + "epoch": 1.6665864661654135, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2092, + "step": 32707 + }, + { + "epoch": 1.6666466165413534, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1876, + "step": 32708 + }, + { + "epoch": 1.6667067669172932, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 32709 + }, + { + "epoch": 1.6667669172932331, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 32710 + }, + { + "epoch": 1.666827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 32711 + }, + { + "epoch": 1.6668872180451126, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 32712 + }, + { + "epoch": 1.6669473684210527, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 32713 + }, + { + "epoch": 1.6670075187969924, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 32714 + }, + { + "epoch": 1.6670676691729325, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 32715 + }, + { + "epoch": 1.6671278195488721, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 32716 + }, + { + "epoch": 1.667187969924812, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 32717 + }, + { + "epoch": 1.6672481203007519, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 32718 + }, + { + "epoch": 1.6673082706766917, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 32719 + }, + { + "epoch": 1.6673684210526316, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 32720 + }, + { + "epoch": 1.6674285714285715, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 32721 + }, + { + "epoch": 1.6674887218045114, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 32722 + }, + { + "epoch": 1.667548872180451, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 32723 + }, + { + "epoch": 1.667609022556391, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 32724 + }, + { + "epoch": 1.6676691729323307, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 32725 + }, + { + "epoch": 1.6677293233082708, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2657, + "step": 32726 + }, + { + "epoch": 1.6677894736842105, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 32727 + }, + { + "epoch": 1.6678496240601504, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 32728 + }, + { + "epoch": 1.6679097744360902, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 32729 + }, + { + "epoch": 1.66796992481203, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 32730 + }, + { + "epoch": 1.66803007518797, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 32731 + }, + { + "epoch": 1.6680902255639096, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 32732 + }, + { + "epoch": 1.6681503759398497, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2307, + "step": 32733 + }, + { + "epoch": 1.6682105263157894, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 32734 + }, + { + "epoch": 1.6682706766917295, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 32735 + }, + { + "epoch": 1.668330827067669, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.3837, + "step": 32736 + }, + { + "epoch": 1.668390977443609, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 32737 + }, + { + "epoch": 1.6684511278195489, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 32738 + }, + { + "epoch": 1.6685112781954887, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 32739 + }, + { + "epoch": 1.6685714285714286, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 32740 + }, + { + "epoch": 1.6686315789473685, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 32741 + }, + { + "epoch": 1.6686917293233083, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 32742 + }, + { + "epoch": 1.668751879699248, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1846, + "step": 32743 + }, + { + "epoch": 1.668812030075188, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 32744 + }, + { + "epoch": 1.6688721804511277, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 32745 + }, + { + "epoch": 1.6689323308270678, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 32746 + }, + { + "epoch": 1.6689924812030075, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 32747 + }, + { + "epoch": 1.6690526315789473, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 32748 + }, + { + "epoch": 1.6691127819548872, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 32749 + }, + { + "epoch": 1.669172932330827, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 32750 + }, + { + "epoch": 1.669233082706767, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 32751 + }, + { + "epoch": 1.6692932330827066, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 32752 + }, + { + "epoch": 1.6693533834586467, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 32753 + }, + { + "epoch": 1.6694135338345863, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 32754 + }, + { + "epoch": 1.6694736842105264, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2185, + "step": 32755 + }, + { + "epoch": 1.669533834586466, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 32756 + }, + { + "epoch": 1.669593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 32757 + }, + { + "epoch": 1.6696541353383458, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 32758 + }, + { + "epoch": 1.6697142857142857, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 32759 + }, + { + "epoch": 1.6697744360902256, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 32760 + }, + { + "epoch": 1.6698345864661654, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 32761 + }, + { + "epoch": 1.6698947368421053, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 32762 + }, + { + "epoch": 1.669954887218045, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 32763 + }, + { + "epoch": 1.670015037593985, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 32764 + }, + { + "epoch": 1.6700751879699247, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 32765 + }, + { + "epoch": 1.6701353383458648, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 32766 + }, + { + "epoch": 1.6701954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 32767 + }, + { + "epoch": 1.6702556390977443, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 32768 + }, + { + "epoch": 1.6703157894736842, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2256, + "step": 32769 + }, + { + "epoch": 1.670375939849624, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 32770 + }, + { + "epoch": 1.670436090225564, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 32771 + }, + { + "epoch": 1.6704962406015038, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 32772 + }, + { + "epoch": 1.6705563909774437, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 32773 + }, + { + "epoch": 1.6706165413533833, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 32774 + }, + { + "epoch": 1.6706766917293234, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 32775 + }, + { + "epoch": 1.670736842105263, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 32776 + }, + { + "epoch": 1.6707969924812032, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 32777 + }, + { + "epoch": 1.6708571428571428, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 32778 + }, + { + "epoch": 1.6709172932330827, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 32779 + }, + { + "epoch": 1.6709774436090226, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 32780 + }, + { + "epoch": 1.6710375939849624, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 32781 + }, + { + "epoch": 1.6710977443609023, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 32782 + }, + { + "epoch": 1.671157894736842, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 32783 + }, + { + "epoch": 1.671218045112782, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 32784 + }, + { + "epoch": 1.6712781954887217, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 32785 + }, + { + "epoch": 1.6713383458646618, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 32786 + }, + { + "epoch": 1.6713984962406014, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9007, + "step": 32787 + }, + { + "epoch": 1.6714586466165413, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 32788 + }, + { + "epoch": 1.6715187969924812, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 32789 + }, + { + "epoch": 1.671578947368421, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 32790 + }, + { + "epoch": 1.671639097744361, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 32791 + }, + { + "epoch": 1.6716992481203008, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 32792 + }, + { + "epoch": 1.6717593984962407, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 32793 + }, + { + "epoch": 1.6718195488721803, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 32794 + }, + { + "epoch": 1.6718796992481204, + "grad_norm": 2.078125, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 32795 + }, + { + "epoch": 1.67193984962406, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 32796 + }, + { + "epoch": 1.6720000000000002, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 32797 + }, + { + "epoch": 1.6720601503759398, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 32798 + }, + { + "epoch": 1.6721203007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 32799 + }, + { + "epoch": 1.6721804511278195, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 32800 + }, + { + "epoch": 1.6722406015037594, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 32801 + }, + { + "epoch": 1.6723007518796993, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 32802 + }, + { + "epoch": 1.672360902255639, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 32803 + }, + { + "epoch": 1.672421052631579, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 32804 + }, + { + "epoch": 1.6724812030075187, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 32805 + }, + { + "epoch": 1.6725413533834588, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 32806 + }, + { + "epoch": 1.6726015037593984, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 32807 + }, + { + "epoch": 1.6726616541353383, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 32808 + }, + { + "epoch": 1.6727218045112782, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 32809 + }, + { + "epoch": 1.672781954887218, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 32810 + }, + { + "epoch": 1.672842105263158, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 32811 + }, + { + "epoch": 1.6729022556390978, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 32812 + }, + { + "epoch": 1.6729624060150377, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 32813 + }, + { + "epoch": 1.6730225563909773, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 32814 + }, + { + "epoch": 1.6730827067669174, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 32815 + }, + { + "epoch": 1.673142857142857, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 32816 + }, + { + "epoch": 1.6732030075187971, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 32817 + }, + { + "epoch": 1.6732631578947368, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 32818 + }, + { + "epoch": 1.6733233082706767, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 32819 + }, + { + "epoch": 1.6733834586466165, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 32820 + }, + { + "epoch": 1.6734436090225564, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 32821 + }, + { + "epoch": 1.6735037593984963, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 32822 + }, + { + "epoch": 1.6735639097744361, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 32823 + }, + { + "epoch": 1.673624060150376, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 32824 + }, + { + "epoch": 1.6736842105263157, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 32825 + }, + { + "epoch": 1.6737443609022558, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 32826 + }, + { + "epoch": 1.6738045112781954, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 32827 + }, + { + "epoch": 1.6738646616541355, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 32828 + }, + { + "epoch": 1.6739248120300751, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 32829 + }, + { + "epoch": 1.673984962406015, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 32830 + }, + { + "epoch": 1.674045112781955, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 32831 + }, + { + "epoch": 1.6741052631578948, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 32832 + }, + { + "epoch": 1.6741654135338346, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 32833 + }, + { + "epoch": 1.6742255639097743, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 32834 + }, + { + "epoch": 1.6742857142857144, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1599, + "step": 32835 + }, + { + "epoch": 1.674345864661654, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 32836 + }, + { + "epoch": 1.6744060150375941, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 32837 + }, + { + "epoch": 1.6744661654135338, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 32838 + }, + { + "epoch": 1.6745263157894736, + "grad_norm": 1.3984375, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 32839 + }, + { + "epoch": 1.6745864661654135, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 32840 + }, + { + "epoch": 1.6746466165413534, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 32841 + }, + { + "epoch": 1.6747067669172933, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1912, + "step": 32842 + }, + { + "epoch": 1.6747669172932331, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 32843 + }, + { + "epoch": 1.674827067669173, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 32844 + }, + { + "epoch": 1.6748872180451126, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 32845 + }, + { + "epoch": 1.6749473684210527, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 32846 + }, + { + "epoch": 1.6750075187969924, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 32847 + }, + { + "epoch": 1.6750676691729325, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 32848 + }, + { + "epoch": 1.6751278195488721, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 32849 + }, + { + "epoch": 1.675187969924812, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 32850 + }, + { + "epoch": 1.6752481203007519, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 32851 + }, + { + "epoch": 1.6753082706766917, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 32852 + }, + { + "epoch": 1.6753684210526316, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 32853 + }, + { + "epoch": 1.6754285714285713, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 32854 + }, + { + "epoch": 1.6754887218045114, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 32855 + }, + { + "epoch": 1.675548872180451, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 32856 + }, + { + "epoch": 1.675609022556391, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 32857 + }, + { + "epoch": 1.6756691729323308, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 32858 + }, + { + "epoch": 1.6757293233082706, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 32859 + }, + { + "epoch": 1.6757894736842105, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 32860 + }, + { + "epoch": 1.6758496240601504, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 32861 + }, + { + "epoch": 1.6759097744360902, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 32862 + }, + { + "epoch": 1.67596992481203, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 32863 + }, + { + "epoch": 1.67603007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9315, + "step": 32864 + }, + { + "epoch": 1.6760902255639096, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 32865 + }, + { + "epoch": 1.6761503759398497, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 32866 + }, + { + "epoch": 1.6762105263157894, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 32867 + }, + { + "epoch": 1.6762706766917295, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 32868 + }, + { + "epoch": 1.6763308270676691, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 32869 + }, + { + "epoch": 1.676390977443609, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 32870 + }, + { + "epoch": 1.6764511278195489, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 32871 + }, + { + "epoch": 1.6765112781954887, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 32872 + }, + { + "epoch": 1.6765714285714286, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 32873 + }, + { + "epoch": 1.6766315789473685, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 32874 + }, + { + "epoch": 1.6766917293233083, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 32875 + }, + { + "epoch": 1.676751879699248, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 32876 + }, + { + "epoch": 1.676812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 32877 + }, + { + "epoch": 1.6768721804511277, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 32878 + }, + { + "epoch": 1.6769323308270678, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 32879 + }, + { + "epoch": 1.6769924812030075, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 32880 + }, + { + "epoch": 1.6770526315789474, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 32881 + }, + { + "epoch": 1.6771127819548872, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 32882 + }, + { + "epoch": 1.677172932330827, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 32883 + }, + { + "epoch": 1.677233082706767, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 32884 + }, + { + "epoch": 1.6772932330827066, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 32885 + }, + { + "epoch": 1.6773533834586467, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 32886 + }, + { + "epoch": 1.6774135338345864, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 32887 + }, + { + "epoch": 1.6774736842105265, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 32888 + }, + { + "epoch": 1.677533834586466, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 32889 + }, + { + "epoch": 1.677593984962406, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 32890 + }, + { + "epoch": 1.6776541353383458, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 32891 + }, + { + "epoch": 1.6777142857142857, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 32892 + }, + { + "epoch": 1.6777744360902256, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 32893 + }, + { + "epoch": 1.6778345864661655, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 32894 + }, + { + "epoch": 1.6778947368421053, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 32895 + }, + { + "epoch": 1.677954887218045, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 32896 + }, + { + "epoch": 1.678015037593985, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 32897 + }, + { + "epoch": 1.6780751879699247, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 32898 + }, + { + "epoch": 1.6781353383458648, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 32899 + }, + { + "epoch": 1.6781954887218045, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 32900 + }, + { + "epoch": 1.6782556390977443, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 32901 + }, + { + "epoch": 1.6783157894736842, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 32902 + }, + { + "epoch": 1.678375939849624, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 32903 + }, + { + "epoch": 1.678436090225564, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 32904 + }, + { + "epoch": 1.6784962406015036, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2936, + "step": 32905 + }, + { + "epoch": 1.6785563909774437, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 32906 + }, + { + "epoch": 1.6786165413533833, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 32907 + }, + { + "epoch": 1.6786766917293234, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 32908 + }, + { + "epoch": 1.678736842105263, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 32909 + }, + { + "epoch": 1.678796992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 32910 + }, + { + "epoch": 1.6788571428571428, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 32911 + }, + { + "epoch": 1.6789172932330827, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 32912 + }, + { + "epoch": 1.6789774436090226, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 32913 + }, + { + "epoch": 1.6790375939849624, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 32914 + }, + { + "epoch": 1.6790977443609023, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 32915 + }, + { + "epoch": 1.679157894736842, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 32916 + }, + { + "epoch": 1.679218045112782, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 32917 + }, + { + "epoch": 1.6792781954887217, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 32918 + }, + { + "epoch": 1.6793383458646618, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 32919 + }, + { + "epoch": 1.6793984962406014, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 32920 + }, + { + "epoch": 1.6794586466165413, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1652, + "step": 32921 + }, + { + "epoch": 1.6795187969924812, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 32922 + }, + { + "epoch": 1.679578947368421, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 32923 + }, + { + "epoch": 1.679639097744361, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2663, + "step": 32924 + }, + { + "epoch": 1.6796992481203008, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 32925 + }, + { + "epoch": 1.6797593984962407, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 32926 + }, + { + "epoch": 1.6798195488721803, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 32927 + }, + { + "epoch": 1.6798796992481204, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 32928 + }, + { + "epoch": 1.67993984962406, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 32929 + }, + { + "epoch": 1.6800000000000002, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 32930 + }, + { + "epoch": 1.6800601503759398, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2287, + "step": 32931 + }, + { + "epoch": 1.6801203007518797, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 32932 + }, + { + "epoch": 1.6801804511278196, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 32933 + }, + { + "epoch": 1.6802406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 32934 + }, + { + "epoch": 1.6803007518796993, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 32935 + }, + { + "epoch": 1.680360902255639, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 32936 + }, + { + "epoch": 1.680421052631579, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 32937 + }, + { + "epoch": 1.6804812030075187, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 32938 + }, + { + "epoch": 1.6805413533834588, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 32939 + }, + { + "epoch": 1.6806015037593984, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 32940 + }, + { + "epoch": 1.6806616541353383, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 32941 + }, + { + "epoch": 1.6807218045112782, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 32942 + }, + { + "epoch": 1.680781954887218, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 32943 + }, + { + "epoch": 1.680842105263158, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 32944 + }, + { + "epoch": 1.6809022556390978, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 32945 + }, + { + "epoch": 1.6809624060150377, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1764, + "step": 32946 + }, + { + "epoch": 1.6810225563909773, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 32947 + }, + { + "epoch": 1.6810827067669174, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 32948 + }, + { + "epoch": 1.681142857142857, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 32949 + }, + { + "epoch": 1.6812030075187971, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0994, + "step": 32950 + }, + { + "epoch": 1.6812631578947368, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 32951 + }, + { + "epoch": 1.6813233082706767, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 32952 + }, + { + "epoch": 1.6813834586466165, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 32953 + }, + { + "epoch": 1.6814436090225564, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 32954 + }, + { + "epoch": 1.6815037593984963, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 32955 + }, + { + "epoch": 1.681563909774436, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 32956 + }, + { + "epoch": 1.681624060150376, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 32957 + }, + { + "epoch": 1.6816842105263157, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 32958 + }, + { + "epoch": 1.6817443609022558, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 32959 + }, + { + "epoch": 1.6818045112781954, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 32960 + }, + { + "epoch": 1.6818646616541353, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 32961 + }, + { + "epoch": 1.6819248120300752, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 32962 + }, + { + "epoch": 1.681984962406015, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1873, + "step": 32963 + }, + { + "epoch": 1.682045112781955, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 32964 + }, + { + "epoch": 1.6821052631578948, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 32965 + }, + { + "epoch": 1.6821654135338346, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 32966 + }, + { + "epoch": 1.6822255639097743, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 32967 + }, + { + "epoch": 1.6822857142857144, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 32968 + }, + { + "epoch": 1.682345864661654, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 32969 + }, + { + "epoch": 1.6824060150375941, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 32970 + }, + { + "epoch": 1.6824661654135338, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 32971 + }, + { + "epoch": 1.6825263157894736, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 32972 + }, + { + "epoch": 1.6825864661654135, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 32973 + }, + { + "epoch": 1.6826466165413534, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 32974 + }, + { + "epoch": 1.6827067669172933, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 32975 + }, + { + "epoch": 1.6827669172932331, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 32976 + }, + { + "epoch": 1.682827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 32977 + }, + { + "epoch": 1.6828872180451127, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 32978 + }, + { + "epoch": 1.6829473684210527, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1537, + "step": 32979 + }, + { + "epoch": 1.6830075187969924, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 32980 + }, + { + "epoch": 1.6830676691729325, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 32981 + }, + { + "epoch": 1.6831278195488721, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 32982 + }, + { + "epoch": 1.683187969924812, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 32983 + }, + { + "epoch": 1.6832481203007519, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 32984 + }, + { + "epoch": 1.6833082706766918, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 32985 + }, + { + "epoch": 1.6833684210526316, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 32986 + }, + { + "epoch": 1.6834285714285713, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 32987 + }, + { + "epoch": 1.6834887218045114, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 32988 + }, + { + "epoch": 1.683548872180451, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1666, + "step": 32989 + }, + { + "epoch": 1.683609022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 32990 + }, + { + "epoch": 1.6836691729323308, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 32991 + }, + { + "epoch": 1.6837293233082706, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 32992 + }, + { + "epoch": 1.6837894736842105, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 32993 + }, + { + "epoch": 1.6838496240601504, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 32994 + }, + { + "epoch": 1.6839097744360902, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 32995 + }, + { + "epoch": 1.6839699248120301, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 32996 + }, + { + "epoch": 1.68403007518797, + "grad_norm": 2.328125, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 32997 + }, + { + "epoch": 1.6840902255639096, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 32998 + }, + { + "epoch": 1.6841503759398497, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 32999 + }, + { + "epoch": 1.6842105263157894, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 33000 + }, + { + "epoch": 1.6842706766917295, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 33001 + }, + { + "epoch": 1.6843308270676691, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 33002 + }, + { + "epoch": 1.684390977443609, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 33003 + }, + { + "epoch": 1.6844511278195489, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 33004 + }, + { + "epoch": 1.6845112781954887, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1509, + "step": 33005 + }, + { + "epoch": 1.6845714285714286, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 33006 + }, + { + "epoch": 1.6846315789473683, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 33007 + }, + { + "epoch": 1.6846917293233084, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 33008 + }, + { + "epoch": 1.684751879699248, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 33009 + }, + { + "epoch": 1.684812030075188, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 33010 + }, + { + "epoch": 1.6848721804511277, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 33011 + }, + { + "epoch": 1.6849323308270676, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 33012 + }, + { + "epoch": 1.6849924812030075, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 33013 + }, + { + "epoch": 1.6850526315789474, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 33014 + }, + { + "epoch": 1.6851127819548872, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 33015 + }, + { + "epoch": 1.685172932330827, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 33016 + }, + { + "epoch": 1.685233082706767, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 33017 + }, + { + "epoch": 1.6852932330827066, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2335, + "step": 33018 + }, + { + "epoch": 1.6853533834586467, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 33019 + }, + { + "epoch": 1.6854135338345864, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 33020 + }, + { + "epoch": 1.6854736842105265, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 33021 + }, + { + "epoch": 1.685533834586466, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 33022 + }, + { + "epoch": 1.685593984962406, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 33023 + }, + { + "epoch": 1.6856541353383458, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 33024 + }, + { + "epoch": 1.6857142857142857, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 33025 + }, + { + "epoch": 1.6857744360902256, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 33026 + }, + { + "epoch": 1.6858345864661655, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1687, + "step": 33027 + }, + { + "epoch": 1.6858947368421053, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 33028 + }, + { + "epoch": 1.685954887218045, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 33029 + }, + { + "epoch": 1.686015037593985, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 33030 + }, + { + "epoch": 1.6860751879699247, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 33031 + }, + { + "epoch": 1.6861353383458648, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 33032 + }, + { + "epoch": 1.6861954887218045, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 33033 + }, + { + "epoch": 1.6862556390977443, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 33034 + }, + { + "epoch": 1.6863157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 33035 + }, + { + "epoch": 1.686375939849624, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 33036 + }, + { + "epoch": 1.686436090225564, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1493, + "step": 33037 + }, + { + "epoch": 1.6864962406015036, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 33038 + }, + { + "epoch": 1.6865563909774437, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 33039 + }, + { + "epoch": 1.6866165413533833, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 33040 + }, + { + "epoch": 1.6866766917293234, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 33041 + }, + { + "epoch": 1.686736842105263, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 33042 + }, + { + "epoch": 1.686796992481203, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 33043 + }, + { + "epoch": 1.6868571428571428, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 33044 + }, + { + "epoch": 1.6869172932330827, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 33045 + }, + { + "epoch": 1.6869774436090226, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 33046 + }, + { + "epoch": 1.6870375939849624, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 33047 + }, + { + "epoch": 1.6870977443609023, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 33048 + }, + { + "epoch": 1.687157894736842, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 33049 + }, + { + "epoch": 1.687218045112782, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 33050 + }, + { + "epoch": 1.6872781954887217, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 33051 + }, + { + "epoch": 1.6873383458646618, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 33052 + }, + { + "epoch": 1.6873984962406015, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 33053 + }, + { + "epoch": 1.6874586466165413, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 33054 + }, + { + "epoch": 1.6875187969924812, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 33055 + }, + { + "epoch": 1.687578947368421, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 33056 + }, + { + "epoch": 1.687639097744361, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 33057 + }, + { + "epoch": 1.6876992481203006, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 33058 + }, + { + "epoch": 1.6877593984962407, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 33059 + }, + { + "epoch": 1.6878195488721803, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 33060 + }, + { + "epoch": 1.6878796992481204, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 33061 + }, + { + "epoch": 1.68793984962406, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 33062 + }, + { + "epoch": 1.688, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 33063 + }, + { + "epoch": 1.6880601503759398, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 33064 + }, + { + "epoch": 1.6881203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 33065 + }, + { + "epoch": 1.6881804511278196, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 33066 + }, + { + "epoch": 1.6882406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 33067 + }, + { + "epoch": 1.6883007518796993, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 33068 + }, + { + "epoch": 1.688360902255639, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 33069 + }, + { + "epoch": 1.688421052631579, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 33070 + }, + { + "epoch": 1.6884812030075187, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 33071 + }, + { + "epoch": 1.6885413533834588, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 33072 + }, + { + "epoch": 1.6886015037593984, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 33073 + }, + { + "epoch": 1.6886616541353383, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 33074 + }, + { + "epoch": 1.6887218045112782, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 33075 + }, + { + "epoch": 1.688781954887218, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 33076 + }, + { + "epoch": 1.688842105263158, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 33077 + }, + { + "epoch": 1.6889022556390978, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 33078 + }, + { + "epoch": 1.6889624060150377, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 33079 + }, + { + "epoch": 1.6890225563909773, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 33080 + }, + { + "epoch": 1.6890827067669174, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 33081 + }, + { + "epoch": 1.689142857142857, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 33082 + }, + { + "epoch": 1.6892030075187972, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 33083 + }, + { + "epoch": 1.6892631578947368, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 33084 + }, + { + "epoch": 1.6893233082706767, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 33085 + }, + { + "epoch": 1.6893834586466165, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.8665, + "step": 33086 + }, + { + "epoch": 1.6894436090225564, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 33087 + }, + { + "epoch": 1.6895037593984963, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 33088 + }, + { + "epoch": 1.689563909774436, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 33089 + }, + { + "epoch": 1.689624060150376, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 33090 + }, + { + "epoch": 1.6896842105263157, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 33091 + }, + { + "epoch": 1.6897443609022558, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 33092 + }, + { + "epoch": 1.6898045112781954, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.2333, + "step": 33093 + }, + { + "epoch": 1.6898646616541353, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.4148, + "step": 33094 + }, + { + "epoch": 1.6899248120300752, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 33095 + }, + { + "epoch": 1.689984962406015, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 33096 + }, + { + "epoch": 1.690045112781955, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 33097 + }, + { + "epoch": 1.6901052631578948, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 33098 + }, + { + "epoch": 1.6901654135338346, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1124, + "step": 33099 + }, + { + "epoch": 1.6902255639097743, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 33100 + }, + { + "epoch": 1.6902857142857144, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 33101 + }, + { + "epoch": 1.690345864661654, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 33102 + }, + { + "epoch": 1.6904060150375941, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9357, + "step": 33103 + }, + { + "epoch": 1.6904661654135338, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 33104 + }, + { + "epoch": 1.6905263157894737, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 33105 + }, + { + "epoch": 1.6905864661654135, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 33106 + }, + { + "epoch": 1.6906466165413534, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 33107 + }, + { + "epoch": 1.6907067669172933, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 33108 + }, + { + "epoch": 1.690766917293233, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 33109 + }, + { + "epoch": 1.690827067669173, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 33110 + }, + { + "epoch": 1.6908872180451127, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 33111 + }, + { + "epoch": 1.6909473684210528, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 33112 + }, + { + "epoch": 1.6910075187969924, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 33113 + }, + { + "epoch": 1.6910676691729325, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 33114 + }, + { + "epoch": 1.6911278195488721, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 33115 + }, + { + "epoch": 1.691187969924812, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 33116 + }, + { + "epoch": 1.691248120300752, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 33117 + }, + { + "epoch": 1.6913082706766918, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 33118 + }, + { + "epoch": 1.6913684210526316, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 33119 + }, + { + "epoch": 1.6914285714285713, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 33120 + }, + { + "epoch": 1.6914887218045114, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 33121 + }, + { + "epoch": 1.691548872180451, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 33122 + }, + { + "epoch": 1.6916090225563911, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1624, + "step": 33123 + }, + { + "epoch": 1.6916691729323308, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 33124 + }, + { + "epoch": 1.6917293233082706, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 33125 + }, + { + "epoch": 1.6917894736842105, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.165, + "step": 33126 + }, + { + "epoch": 1.6918496240601504, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 33127 + }, + { + "epoch": 1.6919097744360903, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 33128 + }, + { + "epoch": 1.6919699248120301, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 33129 + }, + { + "epoch": 1.69203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 33130 + }, + { + "epoch": 1.6920902255639096, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 33131 + }, + { + "epoch": 1.6921503759398497, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 33132 + }, + { + "epoch": 1.6922105263157894, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 33133 + }, + { + "epoch": 1.6922706766917295, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 33134 + }, + { + "epoch": 1.6923308270676691, + "grad_norm": 1.5078125, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 33135 + }, + { + "epoch": 1.692390977443609, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 33136 + }, + { + "epoch": 1.6924511278195489, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 33137 + }, + { + "epoch": 1.6925112781954887, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 33138 + }, + { + "epoch": 1.6925714285714286, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 33139 + }, + { + "epoch": 1.6926315789473683, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 33140 + }, + { + "epoch": 1.6926917293233084, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 33141 + }, + { + "epoch": 1.692751879699248, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 33142 + }, + { + "epoch": 1.692812030075188, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9104, + "step": 33143 + }, + { + "epoch": 1.6928721804511278, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2232, + "step": 33144 + }, + { + "epoch": 1.6929323308270676, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 33145 + }, + { + "epoch": 1.6929924812030075, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 33146 + }, + { + "epoch": 1.6930526315789474, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 33147 + }, + { + "epoch": 1.6931127819548872, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 33148 + }, + { + "epoch": 1.693172932330827, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 33149 + }, + { + "epoch": 1.693233082706767, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 33150 + }, + { + "epoch": 1.6932932330827066, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 33151 + }, + { + "epoch": 1.6933533834586467, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 33152 + }, + { + "epoch": 1.6934135338345864, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 33153 + }, + { + "epoch": 1.6934736842105265, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 33154 + }, + { + "epoch": 1.6935338345864661, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 33155 + }, + { + "epoch": 1.693593984962406, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 33156 + }, + { + "epoch": 1.6936541353383459, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 33157 + }, + { + "epoch": 1.6937142857142857, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 33158 + }, + { + "epoch": 1.6937744360902256, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 33159 + }, + { + "epoch": 1.6938345864661652, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 33160 + }, + { + "epoch": 1.6938947368421053, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 33161 + }, + { + "epoch": 1.693954887218045, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 33162 + }, + { + "epoch": 1.694015037593985, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 33163 + }, + { + "epoch": 1.6940751879699247, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 33164 + }, + { + "epoch": 1.6941353383458648, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 33165 + }, + { + "epoch": 1.6941954887218045, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 33166 + }, + { + "epoch": 1.6942556390977443, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 33167 + }, + { + "epoch": 1.6943157894736842, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 33168 + }, + { + "epoch": 1.694375939849624, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 33169 + }, + { + "epoch": 1.694436090225564, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 33170 + }, + { + "epoch": 1.6944962406015036, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 33171 + }, + { + "epoch": 1.6945563909774437, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 33172 + }, + { + "epoch": 1.6946165413533834, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 33173 + }, + { + "epoch": 1.6946766917293234, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 33174 + }, + { + "epoch": 1.694736842105263, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 33175 + }, + { + "epoch": 1.694796992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 33176 + }, + { + "epoch": 1.6948571428571428, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 33177 + }, + { + "epoch": 1.6949172932330827, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 33178 + }, + { + "epoch": 1.6949774436090226, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 33179 + }, + { + "epoch": 1.6950375939849625, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 33180 + }, + { + "epoch": 1.6950977443609023, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 33181 + }, + { + "epoch": 1.695157894736842, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 33182 + }, + { + "epoch": 1.695218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 33183 + }, + { + "epoch": 1.6952781954887217, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 33184 + }, + { + "epoch": 1.6953383458646618, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 33185 + }, + { + "epoch": 1.6953984962406015, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 33186 + }, + { + "epoch": 1.6954586466165413, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 33187 + }, + { + "epoch": 1.6955187969924812, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 33188 + }, + { + "epoch": 1.695578947368421, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 33189 + }, + { + "epoch": 1.695639097744361, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 33190 + }, + { + "epoch": 1.6956992481203006, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 33191 + }, + { + "epoch": 1.6957593984962407, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 33192 + }, + { + "epoch": 1.6958195488721803, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 33193 + }, + { + "epoch": 1.6958796992481204, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 33194 + }, + { + "epoch": 1.69593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 33195 + }, + { + "epoch": 1.696, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 33196 + }, + { + "epoch": 1.6960601503759398, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 33197 + }, + { + "epoch": 1.6961203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 33198 + }, + { + "epoch": 1.6961804511278196, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2602, + "step": 33199 + }, + { + "epoch": 1.6962406015037594, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 33200 + }, + { + "epoch": 1.6963007518796993, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 33201 + }, + { + "epoch": 1.696360902255639, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 33202 + }, + { + "epoch": 1.696421052631579, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 33203 + }, + { + "epoch": 1.6964812030075187, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 33204 + }, + { + "epoch": 1.6965413533834588, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 33205 + }, + { + "epoch": 1.6966015037593984, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9519, + "step": 33206 + }, + { + "epoch": 1.6966616541353383, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 33207 + }, + { + "epoch": 1.6967218045112782, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 33208 + }, + { + "epoch": 1.696781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 33209 + }, + { + "epoch": 1.696842105263158, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 33210 + }, + { + "epoch": 1.6969022556390976, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 33211 + }, + { + "epoch": 1.6969624060150377, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 33212 + }, + { + "epoch": 1.6970225563909773, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 33213 + }, + { + "epoch": 1.6970827067669174, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 33214 + }, + { + "epoch": 1.697142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 33215 + }, + { + "epoch": 1.6972030075187972, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 33216 + }, + { + "epoch": 1.6972631578947368, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 33217 + }, + { + "epoch": 1.6973233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1446, + "step": 33218 + }, + { + "epoch": 1.6973834586466165, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 33219 + }, + { + "epoch": 1.6974436090225564, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 33220 + }, + { + "epoch": 1.6975037593984963, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 33221 + }, + { + "epoch": 1.697563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 33222 + }, + { + "epoch": 1.697624060150376, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 33223 + }, + { + "epoch": 1.6976842105263157, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 33224 + }, + { + "epoch": 1.6977443609022558, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 33225 + }, + { + "epoch": 1.6978045112781954, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 33226 + }, + { + "epoch": 1.6978646616541353, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 33227 + }, + { + "epoch": 1.6979248120300752, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 33228 + }, + { + "epoch": 1.697984962406015, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 33229 + }, + { + "epoch": 1.698045112781955, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 33230 + }, + { + "epoch": 1.6981052631578948, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 33231 + }, + { + "epoch": 1.6981654135338347, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 33232 + }, + { + "epoch": 1.6982255639097743, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 33233 + }, + { + "epoch": 1.6982857142857144, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 33234 + }, + { + "epoch": 1.698345864661654, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 33235 + }, + { + "epoch": 1.6984060150375941, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 33236 + }, + { + "epoch": 1.6984661654135338, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 33237 + }, + { + "epoch": 1.6985263157894737, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.3557, + "step": 33238 + }, + { + "epoch": 1.6985864661654135, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 33239 + }, + { + "epoch": 1.6986466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 33240 + }, + { + "epoch": 1.6987067669172933, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 33241 + }, + { + "epoch": 1.698766917293233, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 33242 + }, + { + "epoch": 1.698827067669173, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 33243 + }, + { + "epoch": 1.6988872180451127, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 33244 + }, + { + "epoch": 1.6989473684210528, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 33245 + }, + { + "epoch": 1.6990075187969924, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 33246 + }, + { + "epoch": 1.6990676691729323, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 33247 + }, + { + "epoch": 1.6991278195488722, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 33248 + }, + { + "epoch": 1.699187969924812, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 33249 + }, + { + "epoch": 1.699248120300752, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 33250 + }, + { + "epoch": 1.6993082706766918, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 33251 + }, + { + "epoch": 1.6993684210526316, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 33252 + }, + { + "epoch": 1.6994285714285713, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 33253 + }, + { + "epoch": 1.6994887218045114, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 33254 + }, + { + "epoch": 1.699548872180451, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 33255 + }, + { + "epoch": 1.6996090225563911, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 33256 + }, + { + "epoch": 1.6996691729323308, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 33257 + }, + { + "epoch": 1.6997293233082706, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 33258 + }, + { + "epoch": 1.6997894736842105, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1839, + "step": 33259 + }, + { + "epoch": 1.6998496240601504, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 33260 + }, + { + "epoch": 1.6999097744360903, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 33261 + }, + { + "epoch": 1.69996992481203, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 33262 + }, + { + "epoch": 1.70003007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 33263 + }, + { + "epoch": 1.7000902255639097, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 33264 + }, + { + "epoch": 1.7001503759398497, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 33265 + }, + { + "epoch": 1.7002105263157894, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 33266 + }, + { + "epoch": 1.7002706766917295, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 33267 + }, + { + "epoch": 1.7003308270676691, + "grad_norm": 1.46875, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 33268 + }, + { + "epoch": 1.700390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 33269 + }, + { + "epoch": 1.7004511278195489, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 33270 + }, + { + "epoch": 1.7005112781954888, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 33271 + }, + { + "epoch": 1.7005714285714286, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 33272 + }, + { + "epoch": 1.7006315789473683, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 33273 + }, + { + "epoch": 1.7006917293233084, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 33274 + }, + { + "epoch": 1.700751879699248, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9459, + "step": 33275 + }, + { + "epoch": 1.700812030075188, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9296, + "step": 33276 + }, + { + "epoch": 1.7008721804511278, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 33277 + }, + { + "epoch": 1.7009323308270676, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 33278 + }, + { + "epoch": 1.7009924812030075, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 33279 + }, + { + "epoch": 1.7010526315789474, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 33280 + }, + { + "epoch": 1.7011127819548872, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 33281 + }, + { + "epoch": 1.7011729323308271, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 33282 + }, + { + "epoch": 1.701233082706767, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 33283 + }, + { + "epoch": 1.7012932330827066, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 33284 + }, + { + "epoch": 1.7013533834586467, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 33285 + }, + { + "epoch": 1.7014135338345864, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 33286 + }, + { + "epoch": 1.7014736842105265, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 33287 + }, + { + "epoch": 1.7015338345864661, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2297, + "step": 33288 + }, + { + "epoch": 1.701593984962406, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 33289 + }, + { + "epoch": 1.7016541353383459, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 33290 + }, + { + "epoch": 1.7017142857142857, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 33291 + }, + { + "epoch": 1.7017744360902256, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 33292 + }, + { + "epoch": 1.7018345864661653, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 33293 + }, + { + "epoch": 1.7018947368421053, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 33294 + }, + { + "epoch": 1.701954887218045, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 33295 + }, + { + "epoch": 1.702015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 33296 + }, + { + "epoch": 1.7020751879699247, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 33297 + }, + { + "epoch": 1.7021353383458646, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 33298 + }, + { + "epoch": 1.7021954887218045, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 33299 + }, + { + "epoch": 1.7022556390977444, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 33300 + }, + { + "epoch": 1.7023157894736842, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 33301 + }, + { + "epoch": 1.702375939849624, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1402, + "step": 33302 + }, + { + "epoch": 1.702436090225564, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 33303 + }, + { + "epoch": 1.7024962406015036, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 33304 + }, + { + "epoch": 1.7025563909774437, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 33305 + }, + { + "epoch": 1.7026165413533834, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 33306 + }, + { + "epoch": 1.7026766917293235, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 33307 + }, + { + "epoch": 1.702736842105263, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 33308 + }, + { + "epoch": 1.702796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 33309 + }, + { + "epoch": 1.7028571428571428, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 33310 + }, + { + "epoch": 1.7029172932330827, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 33311 + }, + { + "epoch": 1.7029774436090226, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 33312 + }, + { + "epoch": 1.7030375939849625, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 33313 + }, + { + "epoch": 1.7030977443609023, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 33314 + }, + { + "epoch": 1.703157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 33315 + }, + { + "epoch": 1.703218045112782, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 33316 + }, + { + "epoch": 1.7032781954887217, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 33317 + }, + { + "epoch": 1.7033383458646618, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 33318 + }, + { + "epoch": 1.7033984962406015, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 33319 + }, + { + "epoch": 1.7034586466165413, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 33320 + }, + { + "epoch": 1.7035187969924812, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 33321 + }, + { + "epoch": 1.703578947368421, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 33322 + }, + { + "epoch": 1.703639097744361, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 33323 + }, + { + "epoch": 1.7036992481203006, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 33324 + }, + { + "epoch": 1.7037593984962407, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 33325 + }, + { + "epoch": 1.7038195488721803, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 33326 + }, + { + "epoch": 1.7038796992481204, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 33327 + }, + { + "epoch": 1.70393984962406, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 33328 + }, + { + "epoch": 1.704, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 33329 + }, + { + "epoch": 1.7040601503759398, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 33330 + }, + { + "epoch": 1.7041203007518797, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.2744, + "step": 33331 + }, + { + "epoch": 1.7041804511278196, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 33332 + }, + { + "epoch": 1.7042406015037594, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 33333 + }, + { + "epoch": 1.7043007518796993, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 33334 + }, + { + "epoch": 1.704360902255639, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 33335 + }, + { + "epoch": 1.704421052631579, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 33336 + }, + { + "epoch": 1.7044812030075187, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 33337 + }, + { + "epoch": 1.7045413533834588, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 33338 + }, + { + "epoch": 1.7046015037593985, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 33339 + }, + { + "epoch": 1.7046616541353383, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 33340 + }, + { + "epoch": 1.7047218045112782, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 33341 + }, + { + "epoch": 1.704781954887218, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 33342 + }, + { + "epoch": 1.704842105263158, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 33343 + }, + { + "epoch": 1.7049022556390976, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 33344 + }, + { + "epoch": 1.7049624060150377, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 33345 + }, + { + "epoch": 1.7050225563909773, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 33346 + }, + { + "epoch": 1.7050827067669174, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 33347 + }, + { + "epoch": 1.705142857142857, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 33348 + }, + { + "epoch": 1.705203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 33349 + }, + { + "epoch": 1.7052631578947368, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 33350 + }, + { + "epoch": 1.7053233082706767, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 33351 + }, + { + "epoch": 1.7053834586466166, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 33352 + }, + { + "epoch": 1.7054436090225564, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 33353 + }, + { + "epoch": 1.7055037593984963, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 33354 + }, + { + "epoch": 1.705563909774436, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 33355 + }, + { + "epoch": 1.705624060150376, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 33356 + }, + { + "epoch": 1.7056842105263157, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 33357 + }, + { + "epoch": 1.7057443609022558, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 33358 + }, + { + "epoch": 1.7058045112781954, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 33359 + }, + { + "epoch": 1.7058646616541353, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 33360 + }, + { + "epoch": 1.7059248120300752, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 33361 + }, + { + "epoch": 1.705984962406015, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 33362 + }, + { + "epoch": 1.706045112781955, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 33363 + }, + { + "epoch": 1.7061052631578948, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 33364 + }, + { + "epoch": 1.7061654135338347, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 33365 + }, + { + "epoch": 1.7062255639097743, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2056, + "step": 33366 + }, + { + "epoch": 1.7062857142857144, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 33367 + }, + { + "epoch": 1.706345864661654, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 33368 + }, + { + "epoch": 1.7064060150375941, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 33369 + }, + { + "epoch": 1.7064661654135338, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 33370 + }, + { + "epoch": 1.7065263157894737, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 33371 + }, + { + "epoch": 1.7065864661654135, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 33372 + }, + { + "epoch": 1.7066466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 33373 + }, + { + "epoch": 1.7067067669172933, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 33374 + }, + { + "epoch": 1.706766917293233, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 33375 + }, + { + "epoch": 1.706827067669173, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 33376 + }, + { + "epoch": 1.7068872180451127, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 33377 + }, + { + "epoch": 1.7069473684210528, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 33378 + }, + { + "epoch": 1.7070075187969924, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 33379 + }, + { + "epoch": 1.7070676691729323, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 33380 + }, + { + "epoch": 1.7071278195488722, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 33381 + }, + { + "epoch": 1.707187969924812, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 33382 + }, + { + "epoch": 1.707248120300752, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 33383 + }, + { + "epoch": 1.7073082706766918, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 33384 + }, + { + "epoch": 1.7073684210526316, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 33385 + }, + { + "epoch": 1.7074285714285713, + "grad_norm": 3.5, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 33386 + }, + { + "epoch": 1.7074887218045114, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 33387 + }, + { + "epoch": 1.707548872180451, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 33388 + }, + { + "epoch": 1.7076090225563911, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 33389 + }, + { + "epoch": 1.7076691729323308, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 33390 + }, + { + "epoch": 1.7077293233082707, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 33391 + }, + { + "epoch": 1.7077894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 33392 + }, + { + "epoch": 1.7078496240601504, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 33393 + }, + { + "epoch": 1.7079097744360903, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 33394 + }, + { + "epoch": 1.70796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 33395 + }, + { + "epoch": 1.70803007518797, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 33396 + }, + { + "epoch": 1.7080902255639097, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 33397 + }, + { + "epoch": 1.7081503759398498, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 33398 + }, + { + "epoch": 1.7082105263157894, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 33399 + }, + { + "epoch": 1.7082706766917293, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 33400 + }, + { + "epoch": 1.7083308270676691, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 33401 + }, + { + "epoch": 1.708390977443609, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 33402 + }, + { + "epoch": 1.7084511278195489, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 33403 + }, + { + "epoch": 1.7085112781954888, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 33404 + }, + { + "epoch": 1.7085714285714286, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 33405 + }, + { + "epoch": 1.7086315789473683, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 33406 + }, + { + "epoch": 1.7086917293233084, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 33407 + }, + { + "epoch": 1.708751879699248, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 33408 + }, + { + "epoch": 1.7088120300751881, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 33409 + }, + { + "epoch": 1.7088721804511278, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 33410 + }, + { + "epoch": 1.7089323308270676, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 33411 + }, + { + "epoch": 1.7089924812030075, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 33412 + }, + { + "epoch": 1.7090526315789474, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 33413 + }, + { + "epoch": 1.7091127819548873, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 33414 + }, + { + "epoch": 1.7091729323308271, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 33415 + }, + { + "epoch": 1.709233082706767, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 33416 + }, + { + "epoch": 1.7092932330827066, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 33417 + }, + { + "epoch": 1.7093533834586467, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 33418 + }, + { + "epoch": 1.7094135338345864, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 33419 + }, + { + "epoch": 1.7094736842105265, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 33420 + }, + { + "epoch": 1.7095338345864661, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 33421 + }, + { + "epoch": 1.709593984962406, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 33422 + }, + { + "epoch": 1.7096541353383459, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 33423 + }, + { + "epoch": 1.7097142857142857, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 33424 + }, + { + "epoch": 1.7097744360902256, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1573, + "step": 33425 + }, + { + "epoch": 1.7098345864661653, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 33426 + }, + { + "epoch": 1.7098947368421054, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 33427 + }, + { + "epoch": 1.709954887218045, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 33428 + }, + { + "epoch": 1.710015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 33429 + }, + { + "epoch": 1.7100751879699247, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 33430 + }, + { + "epoch": 1.7101353383458646, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 33431 + }, + { + "epoch": 1.7101954887218045, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 33432 + }, + { + "epoch": 1.7102556390977444, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 33433 + }, + { + "epoch": 1.7103157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 33434 + }, + { + "epoch": 1.710375939849624, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 33435 + }, + { + "epoch": 1.710436090225564, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 33436 + }, + { + "epoch": 1.7104962406015036, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 33437 + }, + { + "epoch": 1.7105563909774437, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 33438 + }, + { + "epoch": 1.7106165413533834, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 33439 + }, + { + "epoch": 1.7106766917293235, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 33440 + }, + { + "epoch": 1.7107368421052631, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 33441 + }, + { + "epoch": 1.710796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 33442 + }, + { + "epoch": 1.7108571428571429, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 33443 + }, + { + "epoch": 1.7109172932330827, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 33444 + }, + { + "epoch": 1.7109774436090226, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 33445 + }, + { + "epoch": 1.7110375939849622, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 33446 + }, + { + "epoch": 1.7110977443609023, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 33447 + }, + { + "epoch": 1.711157894736842, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 33448 + }, + { + "epoch": 1.711218045112782, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 33449 + }, + { + "epoch": 1.7112781954887217, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1537, + "step": 33450 + }, + { + "epoch": 1.7113383458646616, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 33451 + }, + { + "epoch": 1.7113984962406015, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 33452 + }, + { + "epoch": 1.7114586466165413, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 33453 + }, + { + "epoch": 1.7115187969924812, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 33454 + }, + { + "epoch": 1.711578947368421, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 33455 + }, + { + "epoch": 1.711639097744361, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 33456 + }, + { + "epoch": 1.7116992481203006, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 33457 + }, + { + "epoch": 1.7117593984962407, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 33458 + }, + { + "epoch": 1.7118195488721804, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 33459 + }, + { + "epoch": 1.7118796992481204, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9429, + "step": 33460 + }, + { + "epoch": 1.71193984962406, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 33461 + }, + { + "epoch": 1.712, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 33462 + }, + { + "epoch": 1.7120601503759398, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 33463 + }, + { + "epoch": 1.7121203007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 33464 + }, + { + "epoch": 1.7121804511278196, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 33465 + }, + { + "epoch": 1.7122406015037595, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 33466 + }, + { + "epoch": 1.7123007518796993, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 33467 + }, + { + "epoch": 1.712360902255639, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 33468 + }, + { + "epoch": 1.712421052631579, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 33469 + }, + { + "epoch": 1.7124812030075187, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 33470 + }, + { + "epoch": 1.7125413533834588, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 33471 + }, + { + "epoch": 1.7126015037593985, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 33472 + }, + { + "epoch": 1.7126616541353383, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 33473 + }, + { + "epoch": 1.7127218045112782, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 33474 + }, + { + "epoch": 1.712781954887218, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 33475 + }, + { + "epoch": 1.712842105263158, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 33476 + }, + { + "epoch": 1.7129022556390976, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 33477 + }, + { + "epoch": 1.7129624060150377, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 33478 + }, + { + "epoch": 1.7130225563909773, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 33479 + }, + { + "epoch": 1.7130827067669174, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 33480 + }, + { + "epoch": 1.713142857142857, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 33481 + }, + { + "epoch": 1.713203007518797, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 33482 + }, + { + "epoch": 1.7132631578947368, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 33483 + }, + { + "epoch": 1.7133233082706767, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 33484 + }, + { + "epoch": 1.7133834586466166, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 33485 + }, + { + "epoch": 1.7134436090225564, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 33486 + }, + { + "epoch": 1.7135037593984963, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 33487 + }, + { + "epoch": 1.713563909774436, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 33488 + }, + { + "epoch": 1.713624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 33489 + }, + { + "epoch": 1.7136842105263157, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 33490 + }, + { + "epoch": 1.7137443609022558, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 33491 + }, + { + "epoch": 1.7138045112781954, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 33492 + }, + { + "epoch": 1.7138646616541353, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 33493 + }, + { + "epoch": 1.7139248120300752, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 33494 + }, + { + "epoch": 1.713984962406015, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 33495 + }, + { + "epoch": 1.714045112781955, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 33496 + }, + { + "epoch": 1.7141052631578946, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1073, + "step": 33497 + }, + { + "epoch": 1.7141654135338347, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 33498 + }, + { + "epoch": 1.7142255639097743, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.8791, + "step": 33499 + }, + { + "epoch": 1.7142857142857144, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 33500 + }, + { + "epoch": 1.714345864661654, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 33501 + }, + { + "epoch": 1.714406015037594, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 33502 + }, + { + "epoch": 1.7144661654135338, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 33503 + }, + { + "epoch": 1.7145263157894737, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 33504 + }, + { + "epoch": 1.7145864661654135, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 33505 + }, + { + "epoch": 1.7146466165413534, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 33506 + }, + { + "epoch": 1.7147067669172933, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 33507 + }, + { + "epoch": 1.714766917293233, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 33508 + }, + { + "epoch": 1.714827067669173, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 33509 + }, + { + "epoch": 1.7148872180451127, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 33510 + }, + { + "epoch": 1.7149473684210528, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 33511 + }, + { + "epoch": 1.7150075187969924, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 33512 + }, + { + "epoch": 1.7150676691729323, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 33513 + }, + { + "epoch": 1.7151278195488722, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9389, + "step": 33514 + }, + { + "epoch": 1.715187969924812, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 33515 + }, + { + "epoch": 1.715248120300752, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 33516 + }, + { + "epoch": 1.7153082706766918, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 33517 + }, + { + "epoch": 1.7153684210526317, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 33518 + }, + { + "epoch": 1.7154285714285713, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 33519 + }, + { + "epoch": 1.7154887218045114, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 33520 + }, + { + "epoch": 1.715548872180451, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 33521 + }, + { + "epoch": 1.7156090225563911, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 33522 + }, + { + "epoch": 1.7156691729323308, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 33523 + }, + { + "epoch": 1.7157293233082707, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 33524 + }, + { + "epoch": 1.7157894736842105, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9344, + "step": 33525 + }, + { + "epoch": 1.7158496240601504, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 33526 + }, + { + "epoch": 1.7159097744360903, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 33527 + }, + { + "epoch": 1.71596992481203, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 33528 + }, + { + "epoch": 1.71603007518797, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 33529 + }, + { + "epoch": 1.7160902255639097, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 33530 + }, + { + "epoch": 1.7161503759398498, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1636, + "step": 33531 + }, + { + "epoch": 1.7162105263157894, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 33532 + }, + { + "epoch": 1.7162706766917293, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 33533 + }, + { + "epoch": 1.7163308270676692, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 33534 + }, + { + "epoch": 1.716390977443609, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 33535 + }, + { + "epoch": 1.716451127819549, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 33536 + }, + { + "epoch": 1.7165112781954888, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 33537 + }, + { + "epoch": 1.7165714285714286, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 33538 + }, + { + "epoch": 1.7166315789473683, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 33539 + }, + { + "epoch": 1.7166917293233084, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.148, + "step": 33540 + }, + { + "epoch": 1.716751879699248, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 33541 + }, + { + "epoch": 1.7168120300751881, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 33542 + }, + { + "epoch": 1.7168721804511278, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 33543 + }, + { + "epoch": 1.7169323308270676, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 33544 + }, + { + "epoch": 1.7169924812030075, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 33545 + }, + { + "epoch": 1.7170526315789474, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 33546 + }, + { + "epoch": 1.7171127819548873, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 33547 + }, + { + "epoch": 1.717172932330827, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 33548 + }, + { + "epoch": 1.717233082706767, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 33549 + }, + { + "epoch": 1.7172932330827066, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 33550 + }, + { + "epoch": 1.7173533834586467, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 33551 + }, + { + "epoch": 1.7174135338345864, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 33552 + }, + { + "epoch": 1.7174736842105263, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 33553 + }, + { + "epoch": 1.7175338345864661, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 33554 + }, + { + "epoch": 1.717593984962406, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 33555 + }, + { + "epoch": 1.7176541353383459, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 33556 + }, + { + "epoch": 1.7177142857142857, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 33557 + }, + { + "epoch": 1.7177744360902256, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 33558 + }, + { + "epoch": 1.7178345864661653, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 33559 + }, + { + "epoch": 1.7178947368421054, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 33560 + }, + { + "epoch": 1.717954887218045, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 33561 + }, + { + "epoch": 1.718015037593985, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1465, + "step": 33562 + }, + { + "epoch": 1.7180751879699248, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 33563 + }, + { + "epoch": 1.7181353383458646, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 33564 + }, + { + "epoch": 1.7181954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 33565 + }, + { + "epoch": 1.7182556390977444, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 33566 + }, + { + "epoch": 1.7183157894736842, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 33567 + }, + { + "epoch": 1.7183759398496241, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 33568 + }, + { + "epoch": 1.718436090225564, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 33569 + }, + { + "epoch": 1.7184962406015036, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 33570 + }, + { + "epoch": 1.7185563909774437, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9291, + "step": 33571 + }, + { + "epoch": 1.7186165413533834, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 33572 + }, + { + "epoch": 1.7186766917293235, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 33573 + }, + { + "epoch": 1.7187368421052631, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 33574 + }, + { + "epoch": 1.718796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 33575 + }, + { + "epoch": 1.7188571428571429, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 33576 + }, + { + "epoch": 1.7189172932330827, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 33577 + }, + { + "epoch": 1.7189774436090226, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 33578 + }, + { + "epoch": 1.7190375939849623, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 33579 + }, + { + "epoch": 1.7190977443609023, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 33580 + }, + { + "epoch": 1.719157894736842, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 33581 + }, + { + "epoch": 1.719218045112782, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 33582 + }, + { + "epoch": 1.7192781954887217, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 33583 + }, + { + "epoch": 1.7193383458646616, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 33584 + }, + { + "epoch": 1.7193984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 33585 + }, + { + "epoch": 1.7194586466165414, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 33586 + }, + { + "epoch": 1.7195187969924812, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 33587 + }, + { + "epoch": 1.719578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 33588 + }, + { + "epoch": 1.719639097744361, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 33589 + }, + { + "epoch": 1.7196992481203006, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 33590 + }, + { + "epoch": 1.7197593984962407, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 33591 + }, + { + "epoch": 1.7198195488721804, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 33592 + }, + { + "epoch": 1.7198796992481205, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 33593 + }, + { + "epoch": 1.71993984962406, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 33594 + }, + { + "epoch": 1.72, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 33595 + }, + { + "epoch": 1.7200601503759398, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 33596 + }, + { + "epoch": 1.7201203007518797, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 33597 + }, + { + "epoch": 1.7201804511278196, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 33598 + }, + { + "epoch": 1.7202406015037592, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 33599 + }, + { + "epoch": 1.7203007518796993, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 33600 + }, + { + "epoch": 1.720360902255639, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 33601 + }, + { + "epoch": 1.720421052631579, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 33602 + }, + { + "epoch": 1.7204812030075187, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 33603 + }, + { + "epoch": 1.7205413533834586, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 33604 + }, + { + "epoch": 1.7206015037593985, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 33605 + }, + { + "epoch": 1.7206616541353383, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 33606 + }, + { + "epoch": 1.7207218045112782, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 33607 + }, + { + "epoch": 1.720781954887218, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 33608 + }, + { + "epoch": 1.720842105263158, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 33609 + }, + { + "epoch": 1.7209022556390976, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 33610 + }, + { + "epoch": 1.7209624060150377, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 33611 + }, + { + "epoch": 1.7210225563909773, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 33612 + }, + { + "epoch": 1.7210827067669174, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 33613 + }, + { + "epoch": 1.721142857142857, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1546, + "step": 33614 + }, + { + "epoch": 1.721203007518797, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 33615 + }, + { + "epoch": 1.7212631578947368, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 33616 + }, + { + "epoch": 1.7213233082706767, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 33617 + }, + { + "epoch": 1.7213834586466166, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 33618 + }, + { + "epoch": 1.7214436090225564, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 33619 + }, + { + "epoch": 1.7215037593984963, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 33620 + }, + { + "epoch": 1.721563909774436, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 33621 + }, + { + "epoch": 1.721624060150376, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 33622 + }, + { + "epoch": 1.7216842105263157, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 33623 + }, + { + "epoch": 1.7217443609022558, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 33624 + }, + { + "epoch": 1.7218045112781954, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9222, + "step": 33625 + }, + { + "epoch": 1.7218646616541353, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 33626 + }, + { + "epoch": 1.7219248120300752, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 33627 + }, + { + "epoch": 1.721984962406015, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.141, + "step": 33628 + }, + { + "epoch": 1.722045112781955, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 33629 + }, + { + "epoch": 1.7221052631578946, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 33630 + }, + { + "epoch": 1.7221654135338347, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 33631 + }, + { + "epoch": 1.7222255639097743, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 33632 + }, + { + "epoch": 1.7222857142857144, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 33633 + }, + { + "epoch": 1.722345864661654, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 33634 + }, + { + "epoch": 1.722406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 33635 + }, + { + "epoch": 1.7224661654135338, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 33636 + }, + { + "epoch": 1.7225263157894737, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 33637 + }, + { + "epoch": 1.7225864661654136, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 33638 + }, + { + "epoch": 1.7226466165413534, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 33639 + }, + { + "epoch": 1.7227067669172933, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1633, + "step": 33640 + }, + { + "epoch": 1.722766917293233, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 33641 + }, + { + "epoch": 1.722827067669173, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 33642 + }, + { + "epoch": 1.7228872180451127, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 33643 + }, + { + "epoch": 1.7229473684210528, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 33644 + }, + { + "epoch": 1.7230075187969924, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 33645 + }, + { + "epoch": 1.7230676691729323, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 33646 + }, + { + "epoch": 1.7231278195488722, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 33647 + }, + { + "epoch": 1.723187969924812, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 33648 + }, + { + "epoch": 1.723248120300752, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 33649 + }, + { + "epoch": 1.7233082706766916, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 33650 + }, + { + "epoch": 1.7233684210526317, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 33651 + }, + { + "epoch": 1.7234285714285713, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 33652 + }, + { + "epoch": 1.7234887218045114, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 33653 + }, + { + "epoch": 1.723548872180451, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 33654 + }, + { + "epoch": 1.723609022556391, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 33655 + }, + { + "epoch": 1.7236691729323308, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 33656 + }, + { + "epoch": 1.7237293233082707, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 33657 + }, + { + "epoch": 1.7237894736842105, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 33658 + }, + { + "epoch": 1.7238496240601504, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 33659 + }, + { + "epoch": 1.7239097744360903, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 33660 + }, + { + "epoch": 1.72396992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 33661 + }, + { + "epoch": 1.72403007518797, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 33662 + }, + { + "epoch": 1.7240902255639097, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 33663 + }, + { + "epoch": 1.7241503759398498, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 33664 + }, + { + "epoch": 1.7242105263157894, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 33665 + }, + { + "epoch": 1.7242706766917293, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 33666 + }, + { + "epoch": 1.7243308270676692, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 33667 + }, + { + "epoch": 1.724390977443609, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 33668 + }, + { + "epoch": 1.724451127819549, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 33669 + }, + { + "epoch": 1.7245112781954888, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 33670 + }, + { + "epoch": 1.7245714285714286, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 33671 + }, + { + "epoch": 1.7246315789473683, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 33672 + }, + { + "epoch": 1.7246917293233084, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 33673 + }, + { + "epoch": 1.724751879699248, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 33674 + }, + { + "epoch": 1.7248120300751881, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 33675 + }, + { + "epoch": 1.7248721804511278, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 33676 + }, + { + "epoch": 1.7249323308270677, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 33677 + }, + { + "epoch": 1.7249924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 33678 + }, + { + "epoch": 1.7250526315789474, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 33679 + }, + { + "epoch": 1.7251127819548873, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.3069, + "step": 33680 + }, + { + "epoch": 1.725172932330827, + "grad_norm": 1.421875, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 33681 + }, + { + "epoch": 1.725233082706767, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 33682 + }, + { + "epoch": 1.7252932330827067, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 33683 + }, + { + "epoch": 1.7253533834586467, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 33684 + }, + { + "epoch": 1.7254135338345864, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 33685 + }, + { + "epoch": 1.7254736842105263, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 33686 + }, + { + "epoch": 1.7255338345864661, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 33687 + }, + { + "epoch": 1.725593984962406, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 33688 + }, + { + "epoch": 1.7256541353383459, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 33689 + }, + { + "epoch": 1.7257142857142858, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2059, + "step": 33690 + }, + { + "epoch": 1.7257744360902256, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 33691 + }, + { + "epoch": 1.7258345864661653, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 33692 + }, + { + "epoch": 1.7258947368421054, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 33693 + }, + { + "epoch": 1.725954887218045, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 33694 + }, + { + "epoch": 1.7260150375939851, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 33695 + }, + { + "epoch": 1.7260751879699248, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 33696 + }, + { + "epoch": 1.7261353383458646, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 33697 + }, + { + "epoch": 1.7261954887218045, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 33698 + }, + { + "epoch": 1.7262556390977444, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 33699 + }, + { + "epoch": 1.7263157894736842, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 33700 + }, + { + "epoch": 1.726375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 33701 + }, + { + "epoch": 1.726436090225564, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 33702 + }, + { + "epoch": 1.7264962406015036, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 33703 + }, + { + "epoch": 1.7265563909774437, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 33704 + }, + { + "epoch": 1.7266165413533834, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 33705 + }, + { + "epoch": 1.7266766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 33706 + }, + { + "epoch": 1.7267368421052631, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 33707 + }, + { + "epoch": 1.726796992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 33708 + }, + { + "epoch": 1.7268571428571429, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 33709 + }, + { + "epoch": 1.7269172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 33710 + }, + { + "epoch": 1.7269774436090226, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 33711 + }, + { + "epoch": 1.7270375939849623, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 33712 + }, + { + "epoch": 1.7270977443609024, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.235, + "step": 33713 + }, + { + "epoch": 1.727157894736842, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 33714 + }, + { + "epoch": 1.727218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 33715 + }, + { + "epoch": 1.7272781954887217, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 33716 + }, + { + "epoch": 1.7273383458646616, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 33717 + }, + { + "epoch": 1.7273984962406015, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 33718 + }, + { + "epoch": 1.7274586466165414, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 33719 + }, + { + "epoch": 1.7275187969924812, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 33720 + }, + { + "epoch": 1.727578947368421, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 33721 + }, + { + "epoch": 1.727639097744361, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 33722 + }, + { + "epoch": 1.7276992481203006, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 33723 + }, + { + "epoch": 1.7277593984962407, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 33724 + }, + { + "epoch": 1.7278195488721804, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 33725 + }, + { + "epoch": 1.7278796992481205, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 33726 + }, + { + "epoch": 1.72793984962406, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 33727 + }, + { + "epoch": 1.728, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 33728 + }, + { + "epoch": 1.7280601503759399, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 33729 + }, + { + "epoch": 1.7281203007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 33730 + }, + { + "epoch": 1.7281804511278196, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 33731 + }, + { + "epoch": 1.7282406015037592, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 33732 + }, + { + "epoch": 1.7283007518796993, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 33733 + }, + { + "epoch": 1.728360902255639, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 33734 + }, + { + "epoch": 1.728421052631579, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 33735 + }, + { + "epoch": 1.7284812030075187, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 33736 + }, + { + "epoch": 1.7285413533834586, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 33737 + }, + { + "epoch": 1.7286015037593985, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1073, + "step": 33738 + }, + { + "epoch": 1.7286616541353383, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 33739 + }, + { + "epoch": 1.7287218045112782, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 33740 + }, + { + "epoch": 1.728781954887218, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 33741 + }, + { + "epoch": 1.728842105263158, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.2235, + "step": 33742 + }, + { + "epoch": 1.7289022556390976, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 33743 + }, + { + "epoch": 1.7289624060150377, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 33744 + }, + { + "epoch": 1.7290225563909773, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 33745 + }, + { + "epoch": 1.7290827067669174, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 33746 + }, + { + "epoch": 1.729142857142857, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1897, + "step": 33747 + }, + { + "epoch": 1.729203007518797, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 33748 + }, + { + "epoch": 1.7292631578947368, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 33749 + }, + { + "epoch": 1.7293233082706767, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 33750 + }, + { + "epoch": 1.7293834586466166, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 33751 + }, + { + "epoch": 1.7294436090225562, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 33752 + }, + { + "epoch": 1.7295037593984963, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 33753 + }, + { + "epoch": 1.729563909774436, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 33754 + }, + { + "epoch": 1.729624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 33755 + }, + { + "epoch": 1.7296842105263157, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 33756 + }, + { + "epoch": 1.7297443609022556, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 33757 + }, + { + "epoch": 1.7298045112781955, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 33758 + }, + { + "epoch": 1.7298646616541353, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 33759 + }, + { + "epoch": 1.7299248120300752, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 33760 + }, + { + "epoch": 1.729984962406015, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 33761 + }, + { + "epoch": 1.730045112781955, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 33762 + }, + { + "epoch": 1.7301052631578946, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 33763 + }, + { + "epoch": 1.7301654135338347, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 33764 + }, + { + "epoch": 1.7302255639097743, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 33765 + }, + { + "epoch": 1.7302857142857144, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 33766 + }, + { + "epoch": 1.730345864661654, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 33767 + }, + { + "epoch": 1.730406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 33768 + }, + { + "epoch": 1.7304661654135338, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 33769 + }, + { + "epoch": 1.7305263157894737, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2322, + "step": 33770 + }, + { + "epoch": 1.7305864661654136, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9394, + "step": 33771 + }, + { + "epoch": 1.7306466165413534, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2144, + "step": 33772 + }, + { + "epoch": 1.7307067669172933, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 33773 + }, + { + "epoch": 1.730766917293233, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1722, + "step": 33774 + }, + { + "epoch": 1.730827067669173, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 33775 + }, + { + "epoch": 1.7308872180451127, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 33776 + }, + { + "epoch": 1.7309473684210528, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 33777 + }, + { + "epoch": 1.7310075187969924, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 33778 + }, + { + "epoch": 1.7310676691729323, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 33779 + }, + { + "epoch": 1.7311278195488722, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 33780 + }, + { + "epoch": 1.731187969924812, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 33781 + }, + { + "epoch": 1.731248120300752, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 33782 + }, + { + "epoch": 1.7313082706766916, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 33783 + }, + { + "epoch": 1.7313684210526317, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 33784 + }, + { + "epoch": 1.7314285714285713, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 33785 + }, + { + "epoch": 1.7314887218045114, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 33786 + }, + { + "epoch": 1.731548872180451, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 33787 + }, + { + "epoch": 1.731609022556391, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 33788 + }, + { + "epoch": 1.7316691729323308, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 33789 + }, + { + "epoch": 1.7317293233082707, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 33790 + }, + { + "epoch": 1.7317894736842105, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 33791 + }, + { + "epoch": 1.7318496240601504, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 33792 + }, + { + "epoch": 1.7319097744360903, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 33793 + }, + { + "epoch": 1.73196992481203, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 33794 + }, + { + "epoch": 1.73203007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 33795 + }, + { + "epoch": 1.7320902255639097, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 33796 + }, + { + "epoch": 1.7321503759398498, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 33797 + }, + { + "epoch": 1.7322105263157894, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 33798 + }, + { + "epoch": 1.7322706766917293, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 33799 + }, + { + "epoch": 1.7323308270676692, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 33800 + }, + { + "epoch": 1.732390977443609, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 33801 + }, + { + "epoch": 1.732451127819549, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 33802 + }, + { + "epoch": 1.7325112781954886, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 33803 + }, + { + "epoch": 1.7325714285714287, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 33804 + }, + { + "epoch": 1.7326315789473683, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 33805 + }, + { + "epoch": 1.7326917293233084, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 33806 + }, + { + "epoch": 1.732751879699248, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 33807 + }, + { + "epoch": 1.732812030075188, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 33808 + }, + { + "epoch": 1.7328721804511278, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 33809 + }, + { + "epoch": 1.7329323308270677, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 33810 + }, + { + "epoch": 1.7329924812030075, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 33811 + }, + { + "epoch": 1.7330526315789474, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 33812 + }, + { + "epoch": 1.7331127819548873, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 33813 + }, + { + "epoch": 1.733172932330827, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 33814 + }, + { + "epoch": 1.733233082706767, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 33815 + }, + { + "epoch": 1.7332932330827067, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 33816 + }, + { + "epoch": 1.7333533834586468, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 33817 + }, + { + "epoch": 1.7334135338345864, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 33818 + }, + { + "epoch": 1.7334736842105263, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 33819 + }, + { + "epoch": 1.7335338345864661, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 33820 + }, + { + "epoch": 1.733593984962406, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 33821 + }, + { + "epoch": 1.733654135338346, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 33822 + }, + { + "epoch": 1.7337142857142858, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 33823 + }, + { + "epoch": 1.7337744360902256, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 33824 + }, + { + "epoch": 1.7338345864661653, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 33825 + }, + { + "epoch": 1.7338947368421054, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 33826 + }, + { + "epoch": 1.733954887218045, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 33827 + }, + { + "epoch": 1.7340150375939851, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 33828 + }, + { + "epoch": 1.7340751879699248, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 33829 + }, + { + "epoch": 1.7341353383458646, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 33830 + }, + { + "epoch": 1.7341954887218045, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 33831 + }, + { + "epoch": 1.7342556390977444, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 33832 + }, + { + "epoch": 1.7343157894736843, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 33833 + }, + { + "epoch": 1.734375939849624, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9281, + "step": 33834 + }, + { + "epoch": 1.734436090225564, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 33835 + }, + { + "epoch": 1.7344962406015036, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 33836 + }, + { + "epoch": 1.7345563909774437, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 33837 + }, + { + "epoch": 1.7346165413533834, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 33838 + }, + { + "epoch": 1.7346766917293233, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 33839 + }, + { + "epoch": 1.7347368421052631, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 33840 + }, + { + "epoch": 1.734796992481203, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 33841 + }, + { + "epoch": 1.7348571428571429, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 33842 + }, + { + "epoch": 1.7349172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 33843 + }, + { + "epoch": 1.7349774436090226, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 33844 + }, + { + "epoch": 1.7350375939849623, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 33845 + }, + { + "epoch": 1.7350977443609024, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 33846 + }, + { + "epoch": 1.735157894736842, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 33847 + }, + { + "epoch": 1.735218045112782, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 33848 + }, + { + "epoch": 1.7352781954887218, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 33849 + }, + { + "epoch": 1.7353383458646616, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 33850 + }, + { + "epoch": 1.7353984962406015, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 33851 + }, + { + "epoch": 1.7354586466165414, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 33852 + }, + { + "epoch": 1.7355187969924812, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 33853 + }, + { + "epoch": 1.7355789473684209, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 33854 + }, + { + "epoch": 1.735639097744361, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 33855 + }, + { + "epoch": 1.7356992481203006, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 33856 + }, + { + "epoch": 1.7357593984962407, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 33857 + }, + { + "epoch": 1.7358195488721804, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 33858 + }, + { + "epoch": 1.7358796992481205, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 33859 + }, + { + "epoch": 1.7359398496240601, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 33860 + }, + { + "epoch": 1.736, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 33861 + }, + { + "epoch": 1.7360601503759399, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 33862 + }, + { + "epoch": 1.7361203007518797, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 33863 + }, + { + "epoch": 1.7361804511278196, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 33864 + }, + { + "epoch": 1.7362406015037593, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 33865 + }, + { + "epoch": 1.7363007518796993, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 33866 + }, + { + "epoch": 1.736360902255639, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 33867 + }, + { + "epoch": 1.736421052631579, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 33868 + }, + { + "epoch": 1.7364812030075187, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 33869 + }, + { + "epoch": 1.7365413533834586, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 33870 + }, + { + "epoch": 1.7366015037593985, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 33871 + }, + { + "epoch": 1.7366616541353384, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 33872 + }, + { + "epoch": 1.7367218045112782, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 33873 + }, + { + "epoch": 1.736781954887218, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 33874 + }, + { + "epoch": 1.736842105263158, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.2109, + "step": 33875 + }, + { + "epoch": 1.7369022556390976, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 33876 + }, + { + "epoch": 1.7369624060150377, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 33877 + }, + { + "epoch": 1.7370225563909774, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 33878 + }, + { + "epoch": 1.7370827067669175, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9459, + "step": 33879 + }, + { + "epoch": 1.737142857142857, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 33880 + }, + { + "epoch": 1.737203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 33881 + }, + { + "epoch": 1.7372631578947368, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 33882 + }, + { + "epoch": 1.7373233082706767, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 33883 + }, + { + "epoch": 1.7373834586466166, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 33884 + }, + { + "epoch": 1.7374436090225562, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 33885 + }, + { + "epoch": 1.7375037593984963, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 33886 + }, + { + "epoch": 1.737563909774436, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 33887 + }, + { + "epoch": 1.737624060150376, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 33888 + }, + { + "epoch": 1.7376842105263157, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 33889 + }, + { + "epoch": 1.7377443609022556, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 33890 + }, + { + "epoch": 1.7378045112781955, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 33891 + }, + { + "epoch": 1.7378646616541353, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 33892 + }, + { + "epoch": 1.7379248120300752, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 33893 + }, + { + "epoch": 1.737984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 33894 + }, + { + "epoch": 1.738045112781955, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 33895 + }, + { + "epoch": 1.7381052631578946, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 33896 + }, + { + "epoch": 1.7381654135338347, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 33897 + }, + { + "epoch": 1.7382255639097743, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 33898 + }, + { + "epoch": 1.7382857142857144, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 33899 + }, + { + "epoch": 1.738345864661654, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 33900 + }, + { + "epoch": 1.738406015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 33901 + }, + { + "epoch": 1.7384661654135338, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 33902 + }, + { + "epoch": 1.7385263157894737, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2876, + "step": 33903 + }, + { + "epoch": 1.7385864661654136, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 33904 + }, + { + "epoch": 1.7386466165413532, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 33905 + }, + { + "epoch": 1.7387067669172933, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 33906 + }, + { + "epoch": 1.738766917293233, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 33907 + }, + { + "epoch": 1.738827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 33908 + }, + { + "epoch": 1.7388872180451127, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 33909 + }, + { + "epoch": 1.7389473684210528, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.2603, + "step": 33910 + }, + { + "epoch": 1.7390075187969924, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 33911 + }, + { + "epoch": 1.7390676691729323, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 33912 + }, + { + "epoch": 1.7391278195488722, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 33913 + }, + { + "epoch": 1.739187969924812, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 33914 + }, + { + "epoch": 1.739248120300752, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 33915 + }, + { + "epoch": 1.7393082706766916, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 33916 + }, + { + "epoch": 1.7393684210526317, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 33917 + }, + { + "epoch": 1.7394285714285713, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 33918 + }, + { + "epoch": 1.7394887218045114, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 33919 + }, + { + "epoch": 1.739548872180451, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 33920 + }, + { + "epoch": 1.739609022556391, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 33921 + }, + { + "epoch": 1.7396691729323308, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 33922 + }, + { + "epoch": 1.7397293233082707, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 33923 + }, + { + "epoch": 1.7397894736842106, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 33924 + }, + { + "epoch": 1.7398496240601504, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 33925 + }, + { + "epoch": 1.7399097744360903, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 33926 + }, + { + "epoch": 1.73996992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 33927 + }, + { + "epoch": 1.74003007518797, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 33928 + }, + { + "epoch": 1.7400902255639097, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 33929 + }, + { + "epoch": 1.7401503759398498, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 33930 + }, + { + "epoch": 1.7402105263157894, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 33931 + }, + { + "epoch": 1.7402706766917293, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 33932 + }, + { + "epoch": 1.7403308270676692, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 33933 + }, + { + "epoch": 1.740390977443609, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 33934 + }, + { + "epoch": 1.740451127819549, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 33935 + }, + { + "epoch": 1.7405112781954886, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 33936 + }, + { + "epoch": 1.7405714285714287, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 33937 + }, + { + "epoch": 1.7406315789473683, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 33938 + }, + { + "epoch": 1.7406917293233084, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 33939 + }, + { + "epoch": 1.740751879699248, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 33940 + }, + { + "epoch": 1.740812030075188, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 33941 + }, + { + "epoch": 1.7408721804511278, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 33942 + }, + { + "epoch": 1.7409323308270677, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 33943 + }, + { + "epoch": 1.7409924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 33944 + }, + { + "epoch": 1.7410526315789474, + "grad_norm": 1.8359375, + "learning_rate": 3e-05, + "loss": 1.272, + "step": 33945 + }, + { + "epoch": 1.7411127819548873, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 33946 + }, + { + "epoch": 1.741172932330827, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 33947 + }, + { + "epoch": 1.741233082706767, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 33948 + }, + { + "epoch": 1.7412932330827067, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 33949 + }, + { + "epoch": 1.7413533834586468, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 33950 + }, + { + "epoch": 1.7414135338345864, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 33951 + }, + { + "epoch": 1.7414736842105263, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 33952 + }, + { + "epoch": 1.7415338345864662, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 33953 + }, + { + "epoch": 1.741593984962406, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 33954 + }, + { + "epoch": 1.741654135338346, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 33955 + }, + { + "epoch": 1.7417142857142855, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 33956 + }, + { + "epoch": 1.7417744360902256, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 33957 + }, + { + "epoch": 1.7418345864661653, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 33958 + }, + { + "epoch": 1.7418947368421054, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 33959 + }, + { + "epoch": 1.741954887218045, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 33960 + }, + { + "epoch": 1.7420150375939851, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 33961 + }, + { + "epoch": 1.7420751879699248, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 33962 + }, + { + "epoch": 1.7421353383458646, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 33963 + }, + { + "epoch": 1.7421954887218045, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 33964 + }, + { + "epoch": 1.7422556390977444, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 33965 + }, + { + "epoch": 1.7423157894736843, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 33966 + }, + { + "epoch": 1.742375939849624, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 33967 + }, + { + "epoch": 1.742436090225564, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 33968 + }, + { + "epoch": 1.7424962406015037, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 33969 + }, + { + "epoch": 1.7425563909774437, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 33970 + }, + { + "epoch": 1.7426165413533834, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 33971 + }, + { + "epoch": 1.7426766917293233, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 33972 + }, + { + "epoch": 1.7427368421052631, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 33973 + }, + { + "epoch": 1.742796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 33974 + }, + { + "epoch": 1.7428571428571429, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 33975 + }, + { + "epoch": 1.7429172932330828, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 33976 + }, + { + "epoch": 1.7429774436090226, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2006, + "step": 33977 + }, + { + "epoch": 1.7430375939849623, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 33978 + }, + { + "epoch": 1.7430977443609024, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 33979 + }, + { + "epoch": 1.743157894736842, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 33980 + }, + { + "epoch": 1.743218045112782, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 33981 + }, + { + "epoch": 1.7432781954887218, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 33982 + }, + { + "epoch": 1.7433383458646616, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 33983 + }, + { + "epoch": 1.7433984962406015, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 33984 + }, + { + "epoch": 1.7434586466165414, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 33985 + }, + { + "epoch": 1.7435187969924812, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 33986 + }, + { + "epoch": 1.743578947368421, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 33987 + }, + { + "epoch": 1.743639097744361, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 33988 + }, + { + "epoch": 1.7436992481203006, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 33989 + }, + { + "epoch": 1.7437593984962407, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.2817, + "step": 33990 + }, + { + "epoch": 1.7438195488721804, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 33991 + }, + { + "epoch": 1.7438796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 33992 + }, + { + "epoch": 1.7439398496240601, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 33993 + }, + { + "epoch": 1.744, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 33994 + }, + { + "epoch": 1.7440601503759399, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 33995 + }, + { + "epoch": 1.7441203007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 33996 + }, + { + "epoch": 1.7441804511278196, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 33997 + }, + { + "epoch": 1.7442406015037593, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 33998 + }, + { + "epoch": 1.7443007518796994, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 33999 + }, + { + "epoch": 1.744360902255639, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 34000 + }, + { + "epoch": 1.744421052631579, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 34001 + }, + { + "epoch": 1.7444812030075187, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 34002 + }, + { + "epoch": 1.7445413533834586, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 34003 + }, + { + "epoch": 1.7446015037593985, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 34004 + }, + { + "epoch": 1.7446616541353384, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 34005 + }, + { + "epoch": 1.7447218045112782, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 34006 + }, + { + "epoch": 1.744781954887218, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 34007 + }, + { + "epoch": 1.744842105263158, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 34008 + }, + { + "epoch": 1.7449022556390976, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 34009 + }, + { + "epoch": 1.7449624060150377, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 34010 + }, + { + "epoch": 1.7450225563909774, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 34011 + }, + { + "epoch": 1.7450827067669175, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 34012 + }, + { + "epoch": 1.745142857142857, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 34013 + }, + { + "epoch": 1.745203007518797, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 34014 + }, + { + "epoch": 1.7452631578947368, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 34015 + }, + { + "epoch": 1.7453233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 34016 + }, + { + "epoch": 1.7453834586466166, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 34017 + }, + { + "epoch": 1.7454436090225562, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 34018 + }, + { + "epoch": 1.7455037593984963, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 34019 + }, + { + "epoch": 1.745563909774436, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 34020 + }, + { + "epoch": 1.745624060150376, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 34021 + }, + { + "epoch": 1.7456842105263157, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1729, + "step": 34022 + }, + { + "epoch": 1.7457443609022556, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 34023 + }, + { + "epoch": 1.7458045112781955, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 34024 + }, + { + "epoch": 1.7458646616541353, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 34025 + }, + { + "epoch": 1.7459248120300752, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 34026 + }, + { + "epoch": 1.745984962406015, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 34027 + }, + { + "epoch": 1.746045112781955, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 34028 + }, + { + "epoch": 1.7461052631578946, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 34029 + }, + { + "epoch": 1.7461654135338347, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 34030 + }, + { + "epoch": 1.7462255639097743, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 34031 + }, + { + "epoch": 1.7462857142857144, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 34032 + }, + { + "epoch": 1.746345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9336, + "step": 34033 + }, + { + "epoch": 1.746406015037594, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 34034 + }, + { + "epoch": 1.7464661654135338, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 34035 + }, + { + "epoch": 1.7465263157894737, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 34036 + }, + { + "epoch": 1.7465864661654136, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 34037 + }, + { + "epoch": 1.7466466165413532, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 34038 + }, + { + "epoch": 1.7467067669172933, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 34039 + }, + { + "epoch": 1.746766917293233, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 34040 + }, + { + "epoch": 1.746827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 34041 + }, + { + "epoch": 1.7468872180451127, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 34042 + }, + { + "epoch": 1.7469473684210526, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 34043 + }, + { + "epoch": 1.7470075187969925, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 34044 + }, + { + "epoch": 1.7470676691729323, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 34045 + }, + { + "epoch": 1.7471278195488722, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 34046 + }, + { + "epoch": 1.747187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 34047 + }, + { + "epoch": 1.747248120300752, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 34048 + }, + { + "epoch": 1.7473082706766916, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 34049 + }, + { + "epoch": 1.7473684210526317, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 34050 + }, + { + "epoch": 1.7474285714285713, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 34051 + }, + { + "epoch": 1.7474887218045114, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 34052 + }, + { + "epoch": 1.747548872180451, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 34053 + }, + { + "epoch": 1.747609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 34054 + }, + { + "epoch": 1.7476691729323308, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 34055 + }, + { + "epoch": 1.7477293233082707, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 34056 + }, + { + "epoch": 1.7477894736842106, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1461, + "step": 34057 + }, + { + "epoch": 1.7478496240601504, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2313, + "step": 34058 + }, + { + "epoch": 1.7479097744360903, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 34059 + }, + { + "epoch": 1.74796992481203, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 34060 + }, + { + "epoch": 1.74803007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 34061 + }, + { + "epoch": 1.7480902255639097, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 34062 + }, + { + "epoch": 1.7481503759398498, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 34063 + }, + { + "epoch": 1.7482105263157894, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 34064 + }, + { + "epoch": 1.7482706766917293, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.8912, + "step": 34065 + }, + { + "epoch": 1.7483308270676692, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 34066 + }, + { + "epoch": 1.748390977443609, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 34067 + }, + { + "epoch": 1.748451127819549, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 34068 + }, + { + "epoch": 1.7485112781954886, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 34069 + }, + { + "epoch": 1.7485714285714287, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 34070 + }, + { + "epoch": 1.7486315789473683, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 34071 + }, + { + "epoch": 1.7486917293233084, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 34072 + }, + { + "epoch": 1.748751879699248, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 34073 + }, + { + "epoch": 1.748812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 34074 + }, + { + "epoch": 1.7488721804511278, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 34075 + }, + { + "epoch": 1.7489323308270677, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 34076 + }, + { + "epoch": 1.7489924812030075, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 34077 + }, + { + "epoch": 1.7490526315789474, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 34078 + }, + { + "epoch": 1.7491127819548873, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 34079 + }, + { + "epoch": 1.749172932330827, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 34080 + }, + { + "epoch": 1.749233082706767, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 34081 + }, + { + "epoch": 1.7492932330827067, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 34082 + }, + { + "epoch": 1.7493533834586468, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 34083 + }, + { + "epoch": 1.7494135338345864, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 34084 + }, + { + "epoch": 1.7494736842105263, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.2298, + "step": 34085 + }, + { + "epoch": 1.7495338345864662, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 34086 + }, + { + "epoch": 1.749593984962406, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 34087 + }, + { + "epoch": 1.749654135338346, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 34088 + }, + { + "epoch": 1.7497142857142856, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 34089 + }, + { + "epoch": 1.7497744360902256, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1723, + "step": 34090 + }, + { + "epoch": 1.7498345864661653, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 34091 + }, + { + "epoch": 1.7498947368421054, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 34092 + }, + { + "epoch": 1.749954887218045, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 34093 + }, + { + "epoch": 1.750015037593985, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 34094 + }, + { + "epoch": 1.7500751879699248, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 34095 + }, + { + "epoch": 1.7501353383458647, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 34096 + }, + { + "epoch": 1.7501954887218045, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 34097 + }, + { + "epoch": 1.7502556390977444, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 34098 + }, + { + "epoch": 1.7503157894736843, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 34099 + }, + { + "epoch": 1.750375939849624, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 34100 + }, + { + "epoch": 1.750436090225564, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 34101 + }, + { + "epoch": 1.7504962406015037, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 34102 + }, + { + "epoch": 1.7505563909774438, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 34103 + }, + { + "epoch": 1.7506165413533834, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 34104 + }, + { + "epoch": 1.7506766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 34105 + }, + { + "epoch": 1.7507368421052631, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 34106 + }, + { + "epoch": 1.750796992481203, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 34107 + }, + { + "epoch": 1.750857142857143, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9445, + "step": 34108 + }, + { + "epoch": 1.7509172932330828, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 34109 + }, + { + "epoch": 1.7509774436090226, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 34110 + }, + { + "epoch": 1.7510375939849623, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 34111 + }, + { + "epoch": 1.7510977443609024, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 34112 + }, + { + "epoch": 1.751157894736842, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 34113 + }, + { + "epoch": 1.7512180451127821, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 34114 + }, + { + "epoch": 1.7512781954887218, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 34115 + }, + { + "epoch": 1.7513383458646616, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 34116 + }, + { + "epoch": 1.7513984962406015, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 34117 + }, + { + "epoch": 1.7514586466165414, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 34118 + }, + { + "epoch": 1.7515187969924813, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 34119 + }, + { + "epoch": 1.751578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 34120 + }, + { + "epoch": 1.751639097744361, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 34121 + }, + { + "epoch": 1.7516992481203006, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 34122 + }, + { + "epoch": 1.7517593984962407, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 34123 + }, + { + "epoch": 1.7518195488721804, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 34124 + }, + { + "epoch": 1.7518796992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 34125 + }, + { + "epoch": 1.7519398496240601, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 34126 + }, + { + "epoch": 1.752, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 34127 + }, + { + "epoch": 1.7520601503759399, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 34128 + }, + { + "epoch": 1.7521203007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 34129 + }, + { + "epoch": 1.7521804511278196, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 34130 + }, + { + "epoch": 1.7522406015037593, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 34131 + }, + { + "epoch": 1.7523007518796994, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 34132 + }, + { + "epoch": 1.752360902255639, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 34133 + }, + { + "epoch": 1.752421052631579, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 34134 + }, + { + "epoch": 1.7524812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 34135 + }, + { + "epoch": 1.7525413533834586, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 34136 + }, + { + "epoch": 1.7526015037593985, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 34137 + }, + { + "epoch": 1.7526616541353384, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 34138 + }, + { + "epoch": 1.7527218045112782, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 34139 + }, + { + "epoch": 1.7527819548872179, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 34140 + }, + { + "epoch": 1.752842105263158, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 34141 + }, + { + "epoch": 1.7529022556390976, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 34142 + }, + { + "epoch": 1.7529624060150377, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9173, + "step": 34143 + }, + { + "epoch": 1.7530225563909774, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 34144 + }, + { + "epoch": 1.7530827067669172, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 34145 + }, + { + "epoch": 1.7531428571428571, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 34146 + }, + { + "epoch": 1.753203007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 34147 + }, + { + "epoch": 1.7532631578947369, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 34148 + }, + { + "epoch": 1.7533233082706767, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 34149 + }, + { + "epoch": 1.7533834586466166, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 34150 + }, + { + "epoch": 1.7534436090225562, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 34151 + }, + { + "epoch": 1.7535037593984963, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 34152 + }, + { + "epoch": 1.753563909774436, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 34153 + }, + { + "epoch": 1.753624060150376, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1968, + "step": 34154 + }, + { + "epoch": 1.7536842105263157, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 34155 + }, + { + "epoch": 1.7537443609022556, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 34156 + }, + { + "epoch": 1.7538045112781955, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 34157 + }, + { + "epoch": 1.7538646616541353, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 34158 + }, + { + "epoch": 1.7539248120300752, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 34159 + }, + { + "epoch": 1.753984962406015, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 34160 + }, + { + "epoch": 1.754045112781955, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 34161 + }, + { + "epoch": 1.7541052631578946, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 34162 + }, + { + "epoch": 1.7541654135338347, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 34163 + }, + { + "epoch": 1.7542255639097744, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 34164 + }, + { + "epoch": 1.7542857142857144, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 34165 + }, + { + "epoch": 1.754345864661654, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 34166 + }, + { + "epoch": 1.754406015037594, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 34167 + }, + { + "epoch": 1.7544661654135338, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 34168 + }, + { + "epoch": 1.7545263157894737, + "grad_norm": 1.5078125, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 34169 + }, + { + "epoch": 1.7545864661654136, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2138, + "step": 34170 + }, + { + "epoch": 1.7546466165413532, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 34171 + }, + { + "epoch": 1.7547067669172933, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 34172 + }, + { + "epoch": 1.754766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 34173 + }, + { + "epoch": 1.754827067669173, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 34174 + }, + { + "epoch": 1.7548872180451127, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 34175 + }, + { + "epoch": 1.7549473684210526, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 34176 + }, + { + "epoch": 1.7550075187969925, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 34177 + }, + { + "epoch": 1.7550676691729323, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 34178 + }, + { + "epoch": 1.7551278195488722, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 34179 + }, + { + "epoch": 1.755187969924812, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 34180 + }, + { + "epoch": 1.755248120300752, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 34181 + }, + { + "epoch": 1.7553082706766916, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 34182 + }, + { + "epoch": 1.7553684210526317, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 34183 + }, + { + "epoch": 1.7554285714285713, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 34184 + }, + { + "epoch": 1.7554887218045114, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1913, + "step": 34185 + }, + { + "epoch": 1.755548872180451, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 34186 + }, + { + "epoch": 1.755609022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 34187 + }, + { + "epoch": 1.7556691729323308, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 34188 + }, + { + "epoch": 1.7557293233082707, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 34189 + }, + { + "epoch": 1.7557894736842106, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 34190 + }, + { + "epoch": 1.7558496240601502, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 34191 + }, + { + "epoch": 1.7559097744360903, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 34192 + }, + { + "epoch": 1.75596992481203, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 34193 + }, + { + "epoch": 1.75603007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.138, + "step": 34194 + }, + { + "epoch": 1.7560902255639097, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 34195 + }, + { + "epoch": 1.7561503759398496, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 34196 + }, + { + "epoch": 1.7562105263157894, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 34197 + }, + { + "epoch": 1.7562706766917293, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 34198 + }, + { + "epoch": 1.7563308270676692, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 34199 + }, + { + "epoch": 1.756390977443609, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 34200 + }, + { + "epoch": 1.756451127819549, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 34201 + }, + { + "epoch": 1.7565112781954886, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 34202 + }, + { + "epoch": 1.7565714285714287, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 34203 + }, + { + "epoch": 1.7566315789473683, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 34204 + }, + { + "epoch": 1.7566917293233084, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 34205 + }, + { + "epoch": 1.756751879699248, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 34206 + }, + { + "epoch": 1.756812030075188, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 34207 + }, + { + "epoch": 1.7568721804511278, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 34208 + }, + { + "epoch": 1.7569323308270677, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 34209 + }, + { + "epoch": 1.7569924812030076, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1761, + "step": 34210 + }, + { + "epoch": 1.7570526315789474, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 34211 + }, + { + "epoch": 1.7571127819548873, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 34212 + }, + { + "epoch": 1.757172932330827, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 34213 + }, + { + "epoch": 1.757233082706767, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 34214 + }, + { + "epoch": 1.7572932330827067, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 34215 + }, + { + "epoch": 1.7573533834586468, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 34216 + }, + { + "epoch": 1.7574135338345864, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 34217 + }, + { + "epoch": 1.7574736842105263, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 34218 + }, + { + "epoch": 1.7575338345864662, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 34219 + }, + { + "epoch": 1.757593984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 34220 + }, + { + "epoch": 1.757654135338346, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 34221 + }, + { + "epoch": 1.7577142857142856, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 34222 + }, + { + "epoch": 1.7577744360902257, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 34223 + }, + { + "epoch": 1.7578345864661653, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 34224 + }, + { + "epoch": 1.7578947368421054, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 34225 + }, + { + "epoch": 1.757954887218045, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 34226 + }, + { + "epoch": 1.758015037593985, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 34227 + }, + { + "epoch": 1.7580751879699248, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 34228 + }, + { + "epoch": 1.7581353383458647, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 34229 + }, + { + "epoch": 1.7581954887218045, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 34230 + }, + { + "epoch": 1.7582556390977444, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 34231 + }, + { + "epoch": 1.7583157894736843, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 34232 + }, + { + "epoch": 1.758375939849624, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 34233 + }, + { + "epoch": 1.758436090225564, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 34234 + }, + { + "epoch": 1.7584962406015037, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 34235 + }, + { + "epoch": 1.7585563909774438, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 34236 + }, + { + "epoch": 1.7586165413533834, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 34237 + }, + { + "epoch": 1.7586766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 34238 + }, + { + "epoch": 1.7587368421052632, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 34239 + }, + { + "epoch": 1.758796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 34240 + }, + { + "epoch": 1.758857142857143, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 34241 + }, + { + "epoch": 1.7589172932330825, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 34242 + }, + { + "epoch": 1.7589774436090226, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9382, + "step": 34243 + }, + { + "epoch": 1.7590375939849623, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 34244 + }, + { + "epoch": 1.7590977443609024, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 34245 + }, + { + "epoch": 1.759157894736842, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 34246 + }, + { + "epoch": 1.759218045112782, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 34247 + }, + { + "epoch": 1.7592781954887218, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 34248 + }, + { + "epoch": 1.7593383458646616, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 34249 + }, + { + "epoch": 1.7593984962406015, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 34250 + }, + { + "epoch": 1.7594586466165414, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 34251 + }, + { + "epoch": 1.7595187969924813, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 34252 + }, + { + "epoch": 1.759578947368421, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 34253 + }, + { + "epoch": 1.759639097744361, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 34254 + }, + { + "epoch": 1.7596992481203007, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 34255 + }, + { + "epoch": 1.7597593984962407, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 34256 + }, + { + "epoch": 1.7598195488721804, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 34257 + }, + { + "epoch": 1.7598796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 34258 + }, + { + "epoch": 1.7599398496240601, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 34259 + }, + { + "epoch": 1.76, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 34260 + }, + { + "epoch": 1.7600601503759399, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9272, + "step": 34261 + }, + { + "epoch": 1.7601203007518798, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 34262 + }, + { + "epoch": 1.7601804511278196, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 34263 + }, + { + "epoch": 1.7602406015037593, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 34264 + }, + { + "epoch": 1.7603007518796994, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 34265 + }, + { + "epoch": 1.760360902255639, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 34266 + }, + { + "epoch": 1.760421052631579, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 34267 + }, + { + "epoch": 1.7604812030075188, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 34268 + }, + { + "epoch": 1.7605413533834586, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 34269 + }, + { + "epoch": 1.7606015037593985, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 34270 + }, + { + "epoch": 1.7606616541353384, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 34271 + }, + { + "epoch": 1.7607218045112782, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 34272 + }, + { + "epoch": 1.760781954887218, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 34273 + }, + { + "epoch": 1.760842105263158, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 34274 + }, + { + "epoch": 1.7609022556390976, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 34275 + }, + { + "epoch": 1.7609624060150377, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 34276 + }, + { + "epoch": 1.7610225563909774, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 34277 + }, + { + "epoch": 1.7610827067669172, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 34278 + }, + { + "epoch": 1.7611428571428571, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 34279 + }, + { + "epoch": 1.761203007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 34280 + }, + { + "epoch": 1.7612631578947369, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 34281 + }, + { + "epoch": 1.7613233082706767, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 34282 + }, + { + "epoch": 1.7613834586466166, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 34283 + }, + { + "epoch": 1.7614436090225563, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 34284 + }, + { + "epoch": 1.7615037593984963, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 34285 + }, + { + "epoch": 1.761563909774436, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 34286 + }, + { + "epoch": 1.761624060150376, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 34287 + }, + { + "epoch": 1.7616842105263157, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 34288 + }, + { + "epoch": 1.7617443609022556, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 34289 + }, + { + "epoch": 1.7618045112781955, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 34290 + }, + { + "epoch": 1.7618646616541354, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1641, + "step": 34291 + }, + { + "epoch": 1.7619248120300752, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 34292 + }, + { + "epoch": 1.7619849624060149, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 34293 + }, + { + "epoch": 1.762045112781955, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 34294 + }, + { + "epoch": 1.7621052631578946, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 34295 + }, + { + "epoch": 1.7621654135338347, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 34296 + }, + { + "epoch": 1.7622255639097744, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 34297 + }, + { + "epoch": 1.7622857142857142, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 34298 + }, + { + "epoch": 1.762345864661654, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 34299 + }, + { + "epoch": 1.762406015037594, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 34300 + }, + { + "epoch": 1.7624661654135338, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 34301 + }, + { + "epoch": 1.7625263157894737, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 34302 + }, + { + "epoch": 1.7625864661654136, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 34303 + }, + { + "epoch": 1.7626466165413532, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 34304 + }, + { + "epoch": 1.7627067669172933, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 34305 + }, + { + "epoch": 1.762766917293233, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 34306 + }, + { + "epoch": 1.762827067669173, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 34307 + }, + { + "epoch": 1.7628872180451127, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 34308 + }, + { + "epoch": 1.7629473684210526, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 34309 + }, + { + "epoch": 1.7630075187969925, + "grad_norm": 1.84375, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 34310 + }, + { + "epoch": 1.7630676691729323, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 34311 + }, + { + "epoch": 1.7631278195488722, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 34312 + }, + { + "epoch": 1.763187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 34313 + }, + { + "epoch": 1.763248120300752, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 34314 + }, + { + "epoch": 1.7633082706766916, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 34315 + }, + { + "epoch": 1.7633684210526317, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 34316 + }, + { + "epoch": 1.7634285714285713, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 34317 + }, + { + "epoch": 1.7634887218045114, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 34318 + }, + { + "epoch": 1.763548872180451, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 34319 + }, + { + "epoch": 1.763609022556391, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 34320 + }, + { + "epoch": 1.7636691729323308, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 34321 + }, + { + "epoch": 1.7637293233082707, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 34322 + }, + { + "epoch": 1.7637894736842106, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 34323 + }, + { + "epoch": 1.7638496240601502, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 34324 + }, + { + "epoch": 1.7639097744360903, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 34325 + }, + { + "epoch": 1.76396992481203, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 34326 + }, + { + "epoch": 1.76403007518797, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 34327 + }, + { + "epoch": 1.7640902255639097, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 34328 + }, + { + "epoch": 1.7641503759398496, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 34329 + }, + { + "epoch": 1.7642105263157895, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 34330 + }, + { + "epoch": 1.7642706766917293, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 34331 + }, + { + "epoch": 1.7643308270676692, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 34332 + }, + { + "epoch": 1.764390977443609, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 34333 + }, + { + "epoch": 1.764451127819549, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 34334 + }, + { + "epoch": 1.7645112781954886, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 34335 + }, + { + "epoch": 1.7645714285714287, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 34336 + }, + { + "epoch": 1.7646315789473683, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1718, + "step": 34337 + }, + { + "epoch": 1.7646917293233084, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 34338 + }, + { + "epoch": 1.764751879699248, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 34339 + }, + { + "epoch": 1.764812030075188, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 34340 + }, + { + "epoch": 1.7648721804511278, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 34341 + }, + { + "epoch": 1.7649323308270677, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 34342 + }, + { + "epoch": 1.7649924812030076, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 34343 + }, + { + "epoch": 1.7650526315789472, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 34344 + }, + { + "epoch": 1.7651127819548873, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 34345 + }, + { + "epoch": 1.765172932330827, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 34346 + }, + { + "epoch": 1.765233082706767, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 34347 + }, + { + "epoch": 1.7652932330827067, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 34348 + }, + { + "epoch": 1.7653533834586466, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 34349 + }, + { + "epoch": 1.7654135338345864, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.227, + "step": 34350 + }, + { + "epoch": 1.7654736842105263, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 34351 + }, + { + "epoch": 1.7655338345864662, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 34352 + }, + { + "epoch": 1.765593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 34353 + }, + { + "epoch": 1.765654135338346, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 34354 + }, + { + "epoch": 1.7657142857142856, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 34355 + }, + { + "epoch": 1.7657744360902257, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 34356 + }, + { + "epoch": 1.7658345864661653, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 34357 + }, + { + "epoch": 1.7658947368421054, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 34358 + }, + { + "epoch": 1.765954887218045, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 34359 + }, + { + "epoch": 1.766015037593985, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 34360 + }, + { + "epoch": 1.7660751879699248, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 34361 + }, + { + "epoch": 1.7661353383458647, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 34362 + }, + { + "epoch": 1.7661954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 34363 + }, + { + "epoch": 1.7662556390977444, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 34364 + }, + { + "epoch": 1.7663157894736843, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 34365 + }, + { + "epoch": 1.766375939849624, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 34366 + }, + { + "epoch": 1.766436090225564, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 34367 + }, + { + "epoch": 1.7664962406015037, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 34368 + }, + { + "epoch": 1.7665563909774438, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 34369 + }, + { + "epoch": 1.7666165413533834, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 34370 + }, + { + "epoch": 1.7666766917293233, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 34371 + }, + { + "epoch": 1.7667368421052632, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 34372 + }, + { + "epoch": 1.766796992481203, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 34373 + }, + { + "epoch": 1.766857142857143, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 34374 + }, + { + "epoch": 1.7669172932330826, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 34375 + }, + { + "epoch": 1.7669774436090226, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 34376 + }, + { + "epoch": 1.7670375939849623, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 34377 + }, + { + "epoch": 1.7670977443609024, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 34378 + }, + { + "epoch": 1.767157894736842, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 34379 + }, + { + "epoch": 1.767218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 34380 + }, + { + "epoch": 1.7672781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 34381 + }, + { + "epoch": 1.7673383458646617, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 34382 + }, + { + "epoch": 1.7673984962406015, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 34383 + }, + { + "epoch": 1.7674586466165414, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 34384 + }, + { + "epoch": 1.7675187969924813, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 34385 + }, + { + "epoch": 1.767578947368421, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 34386 + }, + { + "epoch": 1.767639097744361, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 34387 + }, + { + "epoch": 1.7676992481203007, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 34388 + }, + { + "epoch": 1.7677593984962408, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 34389 + }, + { + "epoch": 1.7678195488721804, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 34390 + }, + { + "epoch": 1.7678796992481203, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 34391 + }, + { + "epoch": 1.7679398496240601, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 34392 + }, + { + "epoch": 1.768, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 34393 + }, + { + "epoch": 1.7680601503759399, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 34394 + }, + { + "epoch": 1.7681203007518795, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 34395 + }, + { + "epoch": 1.7681804511278196, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 34396 + }, + { + "epoch": 1.7682406015037593, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 34397 + }, + { + "epoch": 1.7683007518796994, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 34398 + }, + { + "epoch": 1.768360902255639, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 34399 + }, + { + "epoch": 1.768421052631579, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 34400 + }, + { + "epoch": 1.7684812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 34401 + }, + { + "epoch": 1.7685413533834586, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 34402 + }, + { + "epoch": 1.7686015037593985, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 34403 + }, + { + "epoch": 1.7686616541353384, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 34404 + }, + { + "epoch": 1.7687218045112783, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 34405 + }, + { + "epoch": 1.768781954887218, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 34406 + }, + { + "epoch": 1.768842105263158, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 34407 + }, + { + "epoch": 1.7689022556390976, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 34408 + }, + { + "epoch": 1.7689624060150377, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 34409 + }, + { + "epoch": 1.7690225563909774, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 34410 + }, + { + "epoch": 1.7690827067669173, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 34411 + }, + { + "epoch": 1.7691428571428571, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 34412 + }, + { + "epoch": 1.769203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 34413 + }, + { + "epoch": 1.7692631578947369, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 34414 + }, + { + "epoch": 1.7693233082706767, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 34415 + }, + { + "epoch": 1.7693834586466166, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 34416 + }, + { + "epoch": 1.7694436090225563, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 34417 + }, + { + "epoch": 1.7695037593984964, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 34418 + }, + { + "epoch": 1.769563909774436, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 34419 + }, + { + "epoch": 1.769624060150376, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 34420 + }, + { + "epoch": 1.7696842105263157, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 34421 + }, + { + "epoch": 1.7697443609022556, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 34422 + }, + { + "epoch": 1.7698045112781955, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 34423 + }, + { + "epoch": 1.7698646616541354, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 34424 + }, + { + "epoch": 1.7699248120300752, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 34425 + }, + { + "epoch": 1.7699849624060149, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9118, + "step": 34426 + }, + { + "epoch": 1.770045112781955, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 34427 + }, + { + "epoch": 1.7701052631578946, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 34428 + }, + { + "epoch": 1.7701654135338347, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 34429 + }, + { + "epoch": 1.7702255639097744, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 34430 + }, + { + "epoch": 1.7702857142857142, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.3123, + "step": 34431 + }, + { + "epoch": 1.7703458646616541, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 34432 + }, + { + "epoch": 1.770406015037594, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 34433 + }, + { + "epoch": 1.7704661654135339, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 34434 + }, + { + "epoch": 1.7705263157894737, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 34435 + }, + { + "epoch": 1.7705864661654136, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 34436 + }, + { + "epoch": 1.7706466165413532, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 34437 + }, + { + "epoch": 1.7707067669172933, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 34438 + }, + { + "epoch": 1.770766917293233, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1902, + "step": 34439 + }, + { + "epoch": 1.770827067669173, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9245, + "step": 34440 + }, + { + "epoch": 1.7708872180451127, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 34441 + }, + { + "epoch": 1.7709473684210526, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 34442 + }, + { + "epoch": 1.7710075187969925, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 34443 + }, + { + "epoch": 1.7710676691729323, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 34444 + }, + { + "epoch": 1.7711278195488722, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 34445 + }, + { + "epoch": 1.7711879699248119, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 34446 + }, + { + "epoch": 1.771248120300752, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 34447 + }, + { + "epoch": 1.7713082706766916, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 34448 + }, + { + "epoch": 1.7713684210526317, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 34449 + }, + { + "epoch": 1.7714285714285714, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 34450 + }, + { + "epoch": 1.7714887218045112, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1403, + "step": 34451 + }, + { + "epoch": 1.771548872180451, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 34452 + }, + { + "epoch": 1.771609022556391, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 34453 + }, + { + "epoch": 1.7716691729323308, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 34454 + }, + { + "epoch": 1.7717293233082707, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 34455 + }, + { + "epoch": 1.7717894736842106, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 34456 + }, + { + "epoch": 1.7718496240601502, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 34457 + }, + { + "epoch": 1.7719097744360903, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 34458 + }, + { + "epoch": 1.77196992481203, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 34459 + }, + { + "epoch": 1.77203007518797, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 34460 + }, + { + "epoch": 1.7720902255639097, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 34461 + }, + { + "epoch": 1.7721503759398496, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 34462 + }, + { + "epoch": 1.7722105263157895, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 34463 + }, + { + "epoch": 1.7722706766917293, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 34464 + }, + { + "epoch": 1.7723308270676692, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 34465 + }, + { + "epoch": 1.772390977443609, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 34466 + }, + { + "epoch": 1.772451127819549, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 34467 + }, + { + "epoch": 1.7725112781954886, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 34468 + }, + { + "epoch": 1.7725714285714287, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 34469 + }, + { + "epoch": 1.7726315789473683, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 34470 + }, + { + "epoch": 1.7726917293233084, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1794, + "step": 34471 + }, + { + "epoch": 1.772751879699248, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 34472 + }, + { + "epoch": 1.772812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 34473 + }, + { + "epoch": 1.7728721804511278, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 34474 + }, + { + "epoch": 1.7729323308270677, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 34475 + }, + { + "epoch": 1.7729924812030076, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 34476 + }, + { + "epoch": 1.7730526315789472, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 34477 + }, + { + "epoch": 1.7731127819548873, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 34478 + }, + { + "epoch": 1.773172932330827, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 34479 + }, + { + "epoch": 1.773233082706767, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 34480 + }, + { + "epoch": 1.7732932330827067, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 34481 + }, + { + "epoch": 1.7733533834586466, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 34482 + }, + { + "epoch": 1.7734135338345864, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 34483 + }, + { + "epoch": 1.7734736842105263, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 34484 + }, + { + "epoch": 1.7735338345864662, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 34485 + }, + { + "epoch": 1.773593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 34486 + }, + { + "epoch": 1.773654135338346, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 34487 + }, + { + "epoch": 1.7737142857142856, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 34488 + }, + { + "epoch": 1.7737744360902257, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 34489 + }, + { + "epoch": 1.7738345864661653, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 34490 + }, + { + "epoch": 1.7738947368421054, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 34491 + }, + { + "epoch": 1.773954887218045, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9052, + "step": 34492 + }, + { + "epoch": 1.774015037593985, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 34493 + }, + { + "epoch": 1.7740751879699248, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 34494 + }, + { + "epoch": 1.7741353383458647, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 34495 + }, + { + "epoch": 1.7741954887218045, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 34496 + }, + { + "epoch": 1.7742556390977442, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 34497 + }, + { + "epoch": 1.7743157894736843, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 34498 + }, + { + "epoch": 1.774375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 34499 + }, + { + "epoch": 1.774436090225564, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 34500 + }, + { + "epoch": 1.7744962406015037, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 34501 + }, + { + "epoch": 1.7745563909774436, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 34502 + }, + { + "epoch": 1.7746165413533834, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 34503 + }, + { + "epoch": 1.7746766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 34504 + }, + { + "epoch": 1.7747368421052632, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 34505 + }, + { + "epoch": 1.774796992481203, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 34506 + }, + { + "epoch": 1.774857142857143, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 34507 + }, + { + "epoch": 1.7749172932330826, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 34508 + }, + { + "epoch": 1.7749774436090227, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2123, + "step": 34509 + }, + { + "epoch": 1.7750375939849623, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 34510 + }, + { + "epoch": 1.7750977443609024, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 34511 + }, + { + "epoch": 1.775157894736842, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 34512 + }, + { + "epoch": 1.775218045112782, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 34513 + }, + { + "epoch": 1.7752781954887218, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 34514 + }, + { + "epoch": 1.7753383458646617, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 34515 + }, + { + "epoch": 1.7753984962406015, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.1563, + "step": 34516 + }, + { + "epoch": 1.7754586466165414, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 34517 + }, + { + "epoch": 1.7755187969924813, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 34518 + }, + { + "epoch": 1.775578947368421, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1802, + "step": 34519 + }, + { + "epoch": 1.775639097744361, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 34520 + }, + { + "epoch": 1.7756992481203007, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 34521 + }, + { + "epoch": 1.7757593984962408, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 34522 + }, + { + "epoch": 1.7758195488721804, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 34523 + }, + { + "epoch": 1.7758796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 34524 + }, + { + "epoch": 1.7759398496240602, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 34525 + }, + { + "epoch": 1.776, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 34526 + }, + { + "epoch": 1.77606015037594, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 34527 + }, + { + "epoch": 1.7761203007518795, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 34528 + }, + { + "epoch": 1.7761804511278196, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 34529 + }, + { + "epoch": 1.7762406015037593, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 34530 + }, + { + "epoch": 1.7763007518796994, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 34531 + }, + { + "epoch": 1.776360902255639, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 34532 + }, + { + "epoch": 1.776421052631579, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1616, + "step": 34533 + }, + { + "epoch": 1.7764812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 34534 + }, + { + "epoch": 1.7765413533834586, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 34535 + }, + { + "epoch": 1.7766015037593985, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 34536 + }, + { + "epoch": 1.7766616541353384, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 34537 + }, + { + "epoch": 1.7767218045112783, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 34538 + }, + { + "epoch": 1.776781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1543, + "step": 34539 + }, + { + "epoch": 1.776842105263158, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 34540 + }, + { + "epoch": 1.7769022556390976, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 34541 + }, + { + "epoch": 1.7769624060150377, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1847, + "step": 34542 + }, + { + "epoch": 1.7770225563909774, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 34543 + }, + { + "epoch": 1.7770827067669173, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 34544 + }, + { + "epoch": 1.7771428571428571, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 34545 + }, + { + "epoch": 1.777203007518797, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 34546 + }, + { + "epoch": 1.7772631578947369, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9444, + "step": 34547 + }, + { + "epoch": 1.7773233082706765, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 34548 + }, + { + "epoch": 1.7773834586466166, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 34549 + }, + { + "epoch": 1.7774436090225563, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 34550 + }, + { + "epoch": 1.7775037593984964, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1983, + "step": 34551 + }, + { + "epoch": 1.777563909774436, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 34552 + }, + { + "epoch": 1.777624060150376, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 34553 + }, + { + "epoch": 1.7776842105263158, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 34554 + }, + { + "epoch": 1.7777443609022556, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 34555 + }, + { + "epoch": 1.7778045112781955, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 34556 + }, + { + "epoch": 1.7778646616541354, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 34557 + }, + { + "epoch": 1.7779248120300752, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 34558 + }, + { + "epoch": 1.777984962406015, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 34559 + }, + { + "epoch": 1.778045112781955, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 34560 + }, + { + "epoch": 1.7781052631578946, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 34561 + }, + { + "epoch": 1.7781654135338347, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 34562 + }, + { + "epoch": 1.7782255639097744, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1909, + "step": 34563 + }, + { + "epoch": 1.7782857142857142, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 34564 + }, + { + "epoch": 1.7783458646616541, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 34565 + }, + { + "epoch": 1.778406015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 34566 + }, + { + "epoch": 1.7784661654135339, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2182, + "step": 34567 + }, + { + "epoch": 1.7785263157894737, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 34568 + }, + { + "epoch": 1.7785864661654136, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 34569 + }, + { + "epoch": 1.7786466165413533, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 34570 + }, + { + "epoch": 1.7787067669172933, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1859, + "step": 34571 + }, + { + "epoch": 1.778766917293233, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 34572 + }, + { + "epoch": 1.778827067669173, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 34573 + }, + { + "epoch": 1.7788872180451127, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 34574 + }, + { + "epoch": 1.7789473684210526, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 34575 + }, + { + "epoch": 1.7790075187969925, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 34576 + }, + { + "epoch": 1.7790676691729324, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 34577 + }, + { + "epoch": 1.7791278195488722, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 34578 + }, + { + "epoch": 1.7791879699248119, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9216, + "step": 34579 + }, + { + "epoch": 1.779248120300752, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 34580 + }, + { + "epoch": 1.7793082706766916, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 34581 + }, + { + "epoch": 1.7793684210526317, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 34582 + }, + { + "epoch": 1.7794285714285714, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 34583 + }, + { + "epoch": 1.7794887218045112, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 34584 + }, + { + "epoch": 1.779548872180451, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 34585 + }, + { + "epoch": 1.779609022556391, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 34586 + }, + { + "epoch": 1.7796691729323308, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 34587 + }, + { + "epoch": 1.7797293233082707, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 34588 + }, + { + "epoch": 1.7797894736842106, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 34589 + }, + { + "epoch": 1.7798496240601502, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 34590 + }, + { + "epoch": 1.7799097744360903, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 34591 + }, + { + "epoch": 1.77996992481203, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 34592 + }, + { + "epoch": 1.78003007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 34593 + }, + { + "epoch": 1.7800902255639097, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 34594 + }, + { + "epoch": 1.7801503759398496, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 34595 + }, + { + "epoch": 1.7802105263157895, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 34596 + }, + { + "epoch": 1.7802706766917293, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 34597 + }, + { + "epoch": 1.7803308270676692, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 34598 + }, + { + "epoch": 1.7803909774436089, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 34599 + }, + { + "epoch": 1.780451127819549, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 34600 + }, + { + "epoch": 1.7805112781954886, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 34601 + }, + { + "epoch": 1.7805714285714287, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 34602 + }, + { + "epoch": 1.7806315789473683, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 34603 + }, + { + "epoch": 1.7806917293233084, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 34604 + }, + { + "epoch": 1.780751879699248, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 34605 + }, + { + "epoch": 1.780812030075188, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 34606 + }, + { + "epoch": 1.7808721804511278, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2616, + "step": 34607 + }, + { + "epoch": 1.7809323308270677, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 34608 + }, + { + "epoch": 1.7809924812030076, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 34609 + }, + { + "epoch": 1.7810526315789472, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 34610 + }, + { + "epoch": 1.7811127819548873, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 34611 + }, + { + "epoch": 1.781172932330827, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 34612 + }, + { + "epoch": 1.781233082706767, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 34613 + }, + { + "epoch": 1.7812932330827067, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 34614 + }, + { + "epoch": 1.7813533834586466, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 34615 + }, + { + "epoch": 1.7814135338345864, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 34616 + }, + { + "epoch": 1.7814736842105263, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 34617 + }, + { + "epoch": 1.7815338345864662, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 34618 + }, + { + "epoch": 1.781593984962406, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 34619 + }, + { + "epoch": 1.781654135338346, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 34620 + }, + { + "epoch": 1.7817142857142856, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 34621 + }, + { + "epoch": 1.7817744360902257, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 34622 + }, + { + "epoch": 1.7818345864661653, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 34623 + }, + { + "epoch": 1.7818947368421054, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 34624 + }, + { + "epoch": 1.781954887218045, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 34625 + }, + { + "epoch": 1.782015037593985, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 34626 + }, + { + "epoch": 1.7820751879699248, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1742, + "step": 34627 + }, + { + "epoch": 1.7821353383458647, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 34628 + }, + { + "epoch": 1.7821954887218046, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 34629 + }, + { + "epoch": 1.7822556390977442, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 34630 + }, + { + "epoch": 1.7823157894736843, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 34631 + }, + { + "epoch": 1.782375939849624, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 34632 + }, + { + "epoch": 1.782436090225564, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 34633 + }, + { + "epoch": 1.7824962406015037, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 34634 + }, + { + "epoch": 1.7825563909774436, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 34635 + }, + { + "epoch": 1.7826165413533834, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 34636 + }, + { + "epoch": 1.7826766917293233, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 34637 + }, + { + "epoch": 1.7827368421052632, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 34638 + }, + { + "epoch": 1.782796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 34639 + }, + { + "epoch": 1.782857142857143, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 34640 + }, + { + "epoch": 1.7829172932330826, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 34641 + }, + { + "epoch": 1.7829774436090227, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 34642 + }, + { + "epoch": 1.7830375939849623, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 34643 + }, + { + "epoch": 1.7830977443609024, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 34644 + }, + { + "epoch": 1.783157894736842, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 34645 + }, + { + "epoch": 1.783218045112782, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 34646 + }, + { + "epoch": 1.7832781954887218, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 34647 + }, + { + "epoch": 1.7833383458646617, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1879, + "step": 34648 + }, + { + "epoch": 1.7833984962406015, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 34649 + }, + { + "epoch": 1.7834586466165412, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 34650 + }, + { + "epoch": 1.7835187969924813, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 34651 + }, + { + "epoch": 1.783578947368421, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 34652 + }, + { + "epoch": 1.783639097744361, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 34653 + }, + { + "epoch": 1.7836992481203007, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 34654 + }, + { + "epoch": 1.7837593984962408, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 34655 + }, + { + "epoch": 1.7838195488721804, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 34656 + }, + { + "epoch": 1.7838796992481203, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 34657 + }, + { + "epoch": 1.7839398496240602, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 34658 + }, + { + "epoch": 1.784, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9411, + "step": 34659 + }, + { + "epoch": 1.78406015037594, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 34660 + }, + { + "epoch": 1.7841203007518796, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 34661 + }, + { + "epoch": 1.7841804511278196, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 34662 + }, + { + "epoch": 1.7842406015037593, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 34663 + }, + { + "epoch": 1.7843007518796994, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 34664 + }, + { + "epoch": 1.784360902255639, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2537, + "step": 34665 + }, + { + "epoch": 1.784421052631579, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 34666 + }, + { + "epoch": 1.7844812030075188, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 34667 + }, + { + "epoch": 1.7845413533834587, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 34668 + }, + { + "epoch": 1.7846015037593985, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 34669 + }, + { + "epoch": 1.7846616541353384, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 34670 + }, + { + "epoch": 1.7847218045112783, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 34671 + }, + { + "epoch": 1.784781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 34672 + }, + { + "epoch": 1.784842105263158, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 34673 + }, + { + "epoch": 1.7849022556390977, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 34674 + }, + { + "epoch": 1.7849624060150378, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 34675 + }, + { + "epoch": 1.7850225563909774, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 34676 + }, + { + "epoch": 1.7850827067669173, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 34677 + }, + { + "epoch": 1.7851428571428571, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 34678 + }, + { + "epoch": 1.785203007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 34679 + }, + { + "epoch": 1.7852631578947369, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 34680 + }, + { + "epoch": 1.7853233082706765, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 34681 + }, + { + "epoch": 1.7853834586466166, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 34682 + }, + { + "epoch": 1.7854436090225563, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1594, + "step": 34683 + }, + { + "epoch": 1.7855037593984964, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 34684 + }, + { + "epoch": 1.785563909774436, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 34685 + }, + { + "epoch": 1.785624060150376, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 34686 + }, + { + "epoch": 1.7856842105263158, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 34687 + }, + { + "epoch": 1.7857443609022556, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 34688 + }, + { + "epoch": 1.7858045112781955, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 34689 + }, + { + "epoch": 1.7858646616541354, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 34690 + }, + { + "epoch": 1.7859248120300752, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 34691 + }, + { + "epoch": 1.785984962406015, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 34692 + }, + { + "epoch": 1.786045112781955, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 34693 + }, + { + "epoch": 1.7861052631578946, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 34694 + }, + { + "epoch": 1.7861654135338347, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 34695 + }, + { + "epoch": 1.7862255639097744, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 34696 + }, + { + "epoch": 1.7862857142857143, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 34697 + }, + { + "epoch": 1.7863458646616541, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 34698 + }, + { + "epoch": 1.786406015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 34699 + }, + { + "epoch": 1.7864661654135339, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 34700 + }, + { + "epoch": 1.7865263157894735, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 34701 + }, + { + "epoch": 1.7865864661654136, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 34702 + }, + { + "epoch": 1.7866466165413533, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 34703 + }, + { + "epoch": 1.7867067669172934, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 34704 + }, + { + "epoch": 1.786766917293233, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 34705 + }, + { + "epoch": 1.786827067669173, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 34706 + }, + { + "epoch": 1.7868872180451127, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 34707 + }, + { + "epoch": 1.7869473684210526, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 34708 + }, + { + "epoch": 1.7870075187969925, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 34709 + }, + { + "epoch": 1.7870676691729324, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 34710 + }, + { + "epoch": 1.7871278195488722, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 34711 + }, + { + "epoch": 1.7871879699248119, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 34712 + }, + { + "epoch": 1.787248120300752, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 34713 + }, + { + "epoch": 1.7873082706766916, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 34714 + }, + { + "epoch": 1.7873684210526317, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 34715 + }, + { + "epoch": 1.7874285714285714, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 34716 + }, + { + "epoch": 1.7874887218045112, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 34717 + }, + { + "epoch": 1.787548872180451, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 34718 + }, + { + "epoch": 1.787609022556391, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 34719 + }, + { + "epoch": 1.7876691729323309, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 34720 + }, + { + "epoch": 1.7877293233082707, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 34721 + }, + { + "epoch": 1.7877894736842106, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2351, + "step": 34722 + }, + { + "epoch": 1.7878496240601502, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 34723 + }, + { + "epoch": 1.7879097744360903, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 34724 + }, + { + "epoch": 1.78796992481203, + "grad_norm": 1.3671875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 34725 + }, + { + "epoch": 1.78803007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 34726 + }, + { + "epoch": 1.7880902255639097, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 34727 + }, + { + "epoch": 1.7881503759398496, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 34728 + }, + { + "epoch": 1.7882105263157895, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 34729 + }, + { + "epoch": 1.7882706766917293, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 34730 + }, + { + "epoch": 1.7883308270676692, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 34731 + }, + { + "epoch": 1.7883909774436089, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 34732 + }, + { + "epoch": 1.788451127819549, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 34733 + }, + { + "epoch": 1.7885112781954886, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 34734 + }, + { + "epoch": 1.7885714285714287, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 34735 + }, + { + "epoch": 1.7886315789473684, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 34736 + }, + { + "epoch": 1.7886917293233082, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 34737 + }, + { + "epoch": 1.788751879699248, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 34738 + }, + { + "epoch": 1.788812030075188, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 34739 + }, + { + "epoch": 1.7888721804511278, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 34740 + }, + { + "epoch": 1.7889323308270677, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 34741 + }, + { + "epoch": 1.7889924812030076, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 34742 + }, + { + "epoch": 1.7890526315789472, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9488, + "step": 34743 + }, + { + "epoch": 1.7891127819548873, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 34744 + }, + { + "epoch": 1.789172932330827, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 34745 + }, + { + "epoch": 1.789233082706767, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 34746 + }, + { + "epoch": 1.7892932330827067, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 34747 + }, + { + "epoch": 1.7893533834586466, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 34748 + }, + { + "epoch": 1.7894135338345865, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 34749 + }, + { + "epoch": 1.7894736842105263, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 34750 + }, + { + "epoch": 1.7895338345864662, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 34751 + }, + { + "epoch": 1.789593984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 34752 + }, + { + "epoch": 1.789654135338346, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 34753 + }, + { + "epoch": 1.7897142857142856, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 34754 + }, + { + "epoch": 1.7897744360902257, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 34755 + }, + { + "epoch": 1.7898345864661653, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 34756 + }, + { + "epoch": 1.7898947368421054, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 34757 + }, + { + "epoch": 1.789954887218045, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 34758 + }, + { + "epoch": 1.790015037593985, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 34759 + }, + { + "epoch": 1.7900751879699248, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 34760 + }, + { + "epoch": 1.7901353383458647, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2052, + "step": 34761 + }, + { + "epoch": 1.7901954887218046, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 34762 + }, + { + "epoch": 1.7902556390977442, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 34763 + }, + { + "epoch": 1.7903157894736843, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 34764 + }, + { + "epoch": 1.790375939849624, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 34765 + }, + { + "epoch": 1.790436090225564, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 34766 + }, + { + "epoch": 1.7904962406015037, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2467, + "step": 34767 + }, + { + "epoch": 1.7905563909774436, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 34768 + }, + { + "epoch": 1.7906165413533834, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 34769 + }, + { + "epoch": 1.7906766917293233, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 34770 + }, + { + "epoch": 1.7907368421052632, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 34771 + }, + { + "epoch": 1.790796992481203, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 34772 + }, + { + "epoch": 1.790857142857143, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 34773 + }, + { + "epoch": 1.7909172932330826, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 34774 + }, + { + "epoch": 1.7909774436090227, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 34775 + }, + { + "epoch": 1.7910375939849623, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 34776 + }, + { + "epoch": 1.7910977443609024, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 34777 + }, + { + "epoch": 1.791157894736842, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 34778 + }, + { + "epoch": 1.791218045112782, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 34779 + }, + { + "epoch": 1.7912781954887218, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 34780 + }, + { + "epoch": 1.7913383458646617, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 34781 + }, + { + "epoch": 1.7913984962406015, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 34782 + }, + { + "epoch": 1.7914586466165412, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 34783 + }, + { + "epoch": 1.7915187969924813, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 34784 + }, + { + "epoch": 1.791578947368421, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 34785 + }, + { + "epoch": 1.791639097744361, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 34786 + }, + { + "epoch": 1.7916992481203007, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 34787 + }, + { + "epoch": 1.7917593984962406, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 34788 + }, + { + "epoch": 1.7918195488721804, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 34789 + }, + { + "epoch": 1.7918796992481203, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 34790 + }, + { + "epoch": 1.7919398496240602, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 34791 + }, + { + "epoch": 1.792, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 34792 + }, + { + "epoch": 1.79206015037594, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 34793 + }, + { + "epoch": 1.7921203007518796, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 34794 + }, + { + "epoch": 1.7921804511278197, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 34795 + }, + { + "epoch": 1.7922406015037593, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 34796 + }, + { + "epoch": 1.7923007518796994, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 34797 + }, + { + "epoch": 1.792360902255639, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 34798 + }, + { + "epoch": 1.792421052631579, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 34799 + }, + { + "epoch": 1.7924812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 34800 + }, + { + "epoch": 1.7925413533834587, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 34801 + }, + { + "epoch": 1.7926015037593985, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 34802 + }, + { + "epoch": 1.7926616541353384, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9354, + "step": 34803 + }, + { + "epoch": 1.7927218045112783, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 34804 + }, + { + "epoch": 1.792781954887218, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 34805 + }, + { + "epoch": 1.792842105263158, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 34806 + }, + { + "epoch": 1.7929022556390977, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 34807 + }, + { + "epoch": 1.7929624060150378, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 34808 + }, + { + "epoch": 1.7930225563909774, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 34809 + }, + { + "epoch": 1.7930827067669173, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 34810 + }, + { + "epoch": 1.7931428571428571, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 34811 + }, + { + "epoch": 1.793203007518797, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 34812 + }, + { + "epoch": 1.793263157894737, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 34813 + }, + { + "epoch": 1.7933233082706765, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 34814 + }, + { + "epoch": 1.7933834586466166, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 34815 + }, + { + "epoch": 1.7934436090225563, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 34816 + }, + { + "epoch": 1.7935037593984964, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 34817 + }, + { + "epoch": 1.793563909774436, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 34818 + }, + { + "epoch": 1.793624060150376, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 34819 + }, + { + "epoch": 1.7936842105263158, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 34820 + }, + { + "epoch": 1.7937443609022556, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 34821 + }, + { + "epoch": 1.7938045112781955, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 34822 + }, + { + "epoch": 1.7938646616541354, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 34823 + }, + { + "epoch": 1.7939248120300753, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 34824 + }, + { + "epoch": 1.793984962406015, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 34825 + }, + { + "epoch": 1.794045112781955, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 34826 + }, + { + "epoch": 1.7941052631578946, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 34827 + }, + { + "epoch": 1.7941654135338347, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 34828 + }, + { + "epoch": 1.7942255639097744, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 34829 + }, + { + "epoch": 1.7942857142857143, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 34830 + }, + { + "epoch": 1.7943458646616541, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 34831 + }, + { + "epoch": 1.794406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 34832 + }, + { + "epoch": 1.7944661654135339, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 34833 + }, + { + "epoch": 1.7945263157894735, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 34834 + }, + { + "epoch": 1.7945864661654136, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 34835 + }, + { + "epoch": 1.7946466165413533, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 34836 + }, + { + "epoch": 1.7947067669172934, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 34837 + }, + { + "epoch": 1.794766917293233, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.2259, + "step": 34838 + }, + { + "epoch": 1.7948270676691729, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 34839 + }, + { + "epoch": 1.7948872180451128, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 34840 + }, + { + "epoch": 1.7949473684210526, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1264, + "step": 34841 + }, + { + "epoch": 1.7950075187969925, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 34842 + }, + { + "epoch": 1.7950676691729324, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 34843 + }, + { + "epoch": 1.7951278195488722, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 34844 + }, + { + "epoch": 1.7951879699248119, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 34845 + }, + { + "epoch": 1.795248120300752, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 34846 + }, + { + "epoch": 1.7953082706766916, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 34847 + }, + { + "epoch": 1.7953684210526317, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 34848 + }, + { + "epoch": 1.7954285714285714, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 34849 + }, + { + "epoch": 1.7954887218045112, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 34850 + }, + { + "epoch": 1.7955488721804511, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 34851 + }, + { + "epoch": 1.795609022556391, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 34852 + }, + { + "epoch": 1.7956691729323309, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 34853 + }, + { + "epoch": 1.7957293233082707, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 34854 + }, + { + "epoch": 1.7957894736842106, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 34855 + }, + { + "epoch": 1.7958496240601503, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 34856 + }, + { + "epoch": 1.7959097744360903, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 34857 + }, + { + "epoch": 1.79596992481203, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 34858 + }, + { + "epoch": 1.79603007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 34859 + }, + { + "epoch": 1.7960902255639097, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 34860 + }, + { + "epoch": 1.7961503759398496, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 34861 + }, + { + "epoch": 1.7962105263157895, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 34862 + }, + { + "epoch": 1.7962706766917294, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 34863 + }, + { + "epoch": 1.7963308270676692, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 34864 + }, + { + "epoch": 1.7963909774436089, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 34865 + }, + { + "epoch": 1.796451127819549, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 34866 + }, + { + "epoch": 1.7965112781954886, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 34867 + }, + { + "epoch": 1.7965714285714287, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 34868 + }, + { + "epoch": 1.7966315789473684, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 34869 + }, + { + "epoch": 1.7966917293233082, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 34870 + }, + { + "epoch": 1.796751879699248, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 34871 + }, + { + "epoch": 1.796812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 34872 + }, + { + "epoch": 1.7968721804511278, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 34873 + }, + { + "epoch": 1.7969323308270677, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 34874 + }, + { + "epoch": 1.7969924812030076, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 34875 + }, + { + "epoch": 1.7970526315789472, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 34876 + }, + { + "epoch": 1.7971127819548873, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1985, + "step": 34877 + }, + { + "epoch": 1.797172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 34878 + }, + { + "epoch": 1.797233082706767, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 34879 + }, + { + "epoch": 1.7972932330827067, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 34880 + }, + { + "epoch": 1.7973533834586466, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 34881 + }, + { + "epoch": 1.7974135338345865, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 34882 + }, + { + "epoch": 1.7974736842105263, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 34883 + }, + { + "epoch": 1.7975338345864662, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 34884 + }, + { + "epoch": 1.7975939849624059, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 34885 + }, + { + "epoch": 1.797654135338346, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 34886 + }, + { + "epoch": 1.7977142857142856, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 34887 + }, + { + "epoch": 1.7977744360902257, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 34888 + }, + { + "epoch": 1.7978345864661653, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 34889 + }, + { + "epoch": 1.7978947368421052, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 34890 + }, + { + "epoch": 1.797954887218045, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 34891 + }, + { + "epoch": 1.798015037593985, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 34892 + }, + { + "epoch": 1.7980751879699248, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 34893 + }, + { + "epoch": 1.7981353383458647, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 34894 + }, + { + "epoch": 1.7981954887218046, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 34895 + }, + { + "epoch": 1.7982556390977442, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 34896 + }, + { + "epoch": 1.7983157894736843, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 34897 + }, + { + "epoch": 1.798375939849624, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 34898 + }, + { + "epoch": 1.798436090225564, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 34899 + }, + { + "epoch": 1.7984962406015037, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 34900 + }, + { + "epoch": 1.7985563909774436, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 34901 + }, + { + "epoch": 1.7986165413533834, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 34902 + }, + { + "epoch": 1.7986766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 34903 + }, + { + "epoch": 1.7987368421052632, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 34904 + }, + { + "epoch": 1.798796992481203, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.2633, + "step": 34905 + }, + { + "epoch": 1.798857142857143, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 34906 + }, + { + "epoch": 1.7989172932330826, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 34907 + }, + { + "epoch": 1.7989774436090227, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 34908 + }, + { + "epoch": 1.7990375939849623, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 34909 + }, + { + "epoch": 1.7990977443609024, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 34910 + }, + { + "epoch": 1.799157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 34911 + }, + { + "epoch": 1.799218045112782, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 34912 + }, + { + "epoch": 1.7992781954887218, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 34913 + }, + { + "epoch": 1.7993383458646617, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 34914 + }, + { + "epoch": 1.7993984962406016, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 34915 + }, + { + "epoch": 1.7994586466165412, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 34916 + }, + { + "epoch": 1.7995187969924813, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 34917 + }, + { + "epoch": 1.799578947368421, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 34918 + }, + { + "epoch": 1.799639097744361, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 34919 + }, + { + "epoch": 1.7996992481203007, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 34920 + }, + { + "epoch": 1.7997593984962406, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 34921 + }, + { + "epoch": 1.7998195488721804, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 34922 + }, + { + "epoch": 1.7998796992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 34923 + }, + { + "epoch": 1.7999398496240602, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 34924 + }, + { + "epoch": 1.8, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 34925 + }, + { + "epoch": 1.80006015037594, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 34926 + }, + { + "epoch": 1.8001203007518796, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 34927 + }, + { + "epoch": 1.8001804511278197, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 34928 + }, + { + "epoch": 1.8002406015037593, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 34929 + }, + { + "epoch": 1.8003007518796994, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 34930 + }, + { + "epoch": 1.800360902255639, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 34931 + }, + { + "epoch": 1.800421052631579, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.202, + "step": 34932 + }, + { + "epoch": 1.8004812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 34933 + }, + { + "epoch": 1.8005413533834587, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 34934 + }, + { + "epoch": 1.8006015037593985, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 34935 + }, + { + "epoch": 1.8006616541353382, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 34936 + }, + { + "epoch": 1.8007218045112783, + "grad_norm": 1.65625, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 34937 + }, + { + "epoch": 1.800781954887218, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 34938 + }, + { + "epoch": 1.800842105263158, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 34939 + }, + { + "epoch": 1.8009022556390977, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 34940 + }, + { + "epoch": 1.8009624060150375, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 34941 + }, + { + "epoch": 1.8010225563909774, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 34942 + }, + { + "epoch": 1.8010827067669173, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 34943 + }, + { + "epoch": 1.8011428571428572, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 34944 + }, + { + "epoch": 1.801203007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 34945 + }, + { + "epoch": 1.801263157894737, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 34946 + }, + { + "epoch": 1.8013233082706765, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 34947 + }, + { + "epoch": 1.8013834586466166, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 34948 + }, + { + "epoch": 1.8014436090225563, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 34949 + }, + { + "epoch": 1.8015037593984964, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 34950 + }, + { + "epoch": 1.801563909774436, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 34951 + }, + { + "epoch": 1.801624060150376, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 34952 + }, + { + "epoch": 1.8016842105263158, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 34953 + }, + { + "epoch": 1.8017443609022556, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 34954 + }, + { + "epoch": 1.8018045112781955, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 34955 + }, + { + "epoch": 1.8018646616541354, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 34956 + }, + { + "epoch": 1.8019248120300753, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 34957 + }, + { + "epoch": 1.801984962406015, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 34958 + }, + { + "epoch": 1.802045112781955, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 34959 + }, + { + "epoch": 1.8021052631578947, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 34960 + }, + { + "epoch": 1.8021654135338347, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 34961 + }, + { + "epoch": 1.8022255639097744, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 34962 + }, + { + "epoch": 1.8022857142857143, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 34963 + }, + { + "epoch": 1.8023458646616541, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 34964 + }, + { + "epoch": 1.802406015037594, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 34965 + }, + { + "epoch": 1.8024661654135339, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 34966 + }, + { + "epoch": 1.8025263157894735, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 34967 + }, + { + "epoch": 1.8025864661654136, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 34968 + }, + { + "epoch": 1.8026466165413533, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9267, + "step": 34969 + }, + { + "epoch": 1.8027067669172934, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 34970 + }, + { + "epoch": 1.802766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 34971 + }, + { + "epoch": 1.802827067669173, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 34972 + }, + { + "epoch": 1.8028872180451128, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 34973 + }, + { + "epoch": 1.8029473684210526, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 34974 + }, + { + "epoch": 1.8030075187969925, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 34975 + }, + { + "epoch": 1.8030676691729324, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 34976 + }, + { + "epoch": 1.8031278195488722, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 34977 + }, + { + "epoch": 1.803187969924812, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 34978 + }, + { + "epoch": 1.803248120300752, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 34979 + }, + { + "epoch": 1.8033082706766916, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 34980 + }, + { + "epoch": 1.8033684210526317, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 34981 + }, + { + "epoch": 1.8034285714285714, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 34982 + }, + { + "epoch": 1.8034887218045113, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 34983 + }, + { + "epoch": 1.8035488721804511, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 34984 + }, + { + "epoch": 1.803609022556391, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 34985 + }, + { + "epoch": 1.8036691729323309, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 34986 + }, + { + "epoch": 1.8037293233082705, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 34987 + }, + { + "epoch": 1.8037894736842106, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 34988 + }, + { + "epoch": 1.8038496240601503, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 34989 + }, + { + "epoch": 1.8039097744360904, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9247, + "step": 34990 + }, + { + "epoch": 1.80396992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 34991 + }, + { + "epoch": 1.8040300751879699, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 34992 + }, + { + "epoch": 1.8040902255639097, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 34993 + }, + { + "epoch": 1.8041503759398496, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 34994 + }, + { + "epoch": 1.8042105263157895, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 34995 + }, + { + "epoch": 1.8042706766917294, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 34996 + }, + { + "epoch": 1.8043308270676692, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 34997 + }, + { + "epoch": 1.8043909774436089, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 34998 + }, + { + "epoch": 1.804451127819549, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 34999 + }, + { + "epoch": 1.8045112781954886, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 35000 + }, + { + "epoch": 1.8045714285714287, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 35001 + }, + { + "epoch": 1.8046315789473684, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 35002 + }, + { + "epoch": 1.8046917293233082, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 35003 + }, + { + "epoch": 1.804751879699248, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 35004 + }, + { + "epoch": 1.804812030075188, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 35005 + }, + { + "epoch": 1.8048721804511278, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 35006 + }, + { + "epoch": 1.8049323308270677, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 35007 + }, + { + "epoch": 1.8049924812030076, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 35008 + }, + { + "epoch": 1.8050526315789472, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 35009 + }, + { + "epoch": 1.8051127819548873, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 35010 + }, + { + "epoch": 1.805172932330827, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 35011 + }, + { + "epoch": 1.805233082706767, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 35012 + }, + { + "epoch": 1.8052932330827067, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 35013 + }, + { + "epoch": 1.8053533834586466, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 35014 + }, + { + "epoch": 1.8054135338345865, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 35015 + }, + { + "epoch": 1.8054736842105263, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 35016 + }, + { + "epoch": 1.8055338345864662, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 35017 + }, + { + "epoch": 1.8055939849624059, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 35018 + }, + { + "epoch": 1.805654135338346, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 35019 + }, + { + "epoch": 1.8057142857142856, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 35020 + }, + { + "epoch": 1.8057744360902257, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 35021 + }, + { + "epoch": 1.8058345864661653, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 35022 + }, + { + "epoch": 1.8058947368421052, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 35023 + }, + { + "epoch": 1.805954887218045, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 35024 + }, + { + "epoch": 1.806015037593985, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 35025 + }, + { + "epoch": 1.8060751879699248, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 35026 + }, + { + "epoch": 1.8061353383458647, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 35027 + }, + { + "epoch": 1.8061954887218046, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 35028 + }, + { + "epoch": 1.8062556390977442, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 35029 + }, + { + "epoch": 1.8063157894736843, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 35030 + }, + { + "epoch": 1.806375939849624, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 35031 + }, + { + "epoch": 1.806436090225564, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 35032 + }, + { + "epoch": 1.8064962406015037, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 35033 + }, + { + "epoch": 1.8065563909774436, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 35034 + }, + { + "epoch": 1.8066165413533835, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 35035 + }, + { + "epoch": 1.8066766917293233, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 35036 + }, + { + "epoch": 1.8067368421052632, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 35037 + }, + { + "epoch": 1.8067969924812028, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 35038 + }, + { + "epoch": 1.806857142857143, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 35039 + }, + { + "epoch": 1.8069172932330826, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 35040 + }, + { + "epoch": 1.8069774436090227, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 35041 + }, + { + "epoch": 1.8070375939849623, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 35042 + }, + { + "epoch": 1.8070977443609022, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 35043 + }, + { + "epoch": 1.807157894736842, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 35044 + }, + { + "epoch": 1.807218045112782, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 35045 + }, + { + "epoch": 1.8072781954887218, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 35046 + }, + { + "epoch": 1.8073383458646617, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9373, + "step": 35047 + }, + { + "epoch": 1.8073984962406016, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 35048 + }, + { + "epoch": 1.8074586466165412, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 35049 + }, + { + "epoch": 1.8075187969924813, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 35050 + }, + { + "epoch": 1.807578947368421, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 35051 + }, + { + "epoch": 1.807639097744361, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 35052 + }, + { + "epoch": 1.8076992481203007, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 35053 + }, + { + "epoch": 1.8077593984962406, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9101, + "step": 35054 + }, + { + "epoch": 1.8078195488721804, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 35055 + }, + { + "epoch": 1.8078796992481203, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 35056 + }, + { + "epoch": 1.8079398496240602, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 35057 + }, + { + "epoch": 1.808, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1768, + "step": 35058 + }, + { + "epoch": 1.80806015037594, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 35059 + }, + { + "epoch": 1.8081203007518796, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 35060 + }, + { + "epoch": 1.8081804511278197, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 35061 + }, + { + "epoch": 1.8082406015037593, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 35062 + }, + { + "epoch": 1.8083007518796994, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 35063 + }, + { + "epoch": 1.808360902255639, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 35064 + }, + { + "epoch": 1.808421052631579, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 35065 + }, + { + "epoch": 1.8084812030075188, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 35066 + }, + { + "epoch": 1.8085413533834587, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 35067 + }, + { + "epoch": 1.8086015037593985, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 35068 + }, + { + "epoch": 1.8086616541353382, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 35069 + }, + { + "epoch": 1.8087218045112783, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 35070 + }, + { + "epoch": 1.808781954887218, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1971, + "step": 35071 + }, + { + "epoch": 1.808842105263158, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 35072 + }, + { + "epoch": 1.8089022556390977, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 35073 + }, + { + "epoch": 1.8089624060150375, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 35074 + }, + { + "epoch": 1.8090225563909774, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 35075 + }, + { + "epoch": 1.8090827067669173, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 35076 + }, + { + "epoch": 1.8091428571428572, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 35077 + }, + { + "epoch": 1.809203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 35078 + }, + { + "epoch": 1.809263157894737, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 35079 + }, + { + "epoch": 1.8093233082706766, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 35080 + }, + { + "epoch": 1.8093834586466166, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 35081 + }, + { + "epoch": 1.8094436090225563, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 35082 + }, + { + "epoch": 1.8095037593984964, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 35083 + }, + { + "epoch": 1.809563909774436, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 35084 + }, + { + "epoch": 1.809624060150376, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 35085 + }, + { + "epoch": 1.8096842105263158, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 35086 + }, + { + "epoch": 1.8097443609022557, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 35087 + }, + { + "epoch": 1.8098045112781955, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 35088 + }, + { + "epoch": 1.8098646616541352, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 35089 + }, + { + "epoch": 1.8099248120300753, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 35090 + }, + { + "epoch": 1.809984962406015, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 35091 + }, + { + "epoch": 1.810045112781955, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 35092 + }, + { + "epoch": 1.8101052631578947, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 35093 + }, + { + "epoch": 1.8101654135338345, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 35094 + }, + { + "epoch": 1.8102255639097744, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 35095 + }, + { + "epoch": 1.8102857142857143, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 35096 + }, + { + "epoch": 1.8103458646616541, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 35097 + }, + { + "epoch": 1.810406015037594, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 35098 + }, + { + "epoch": 1.810466165413534, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 35099 + }, + { + "epoch": 1.8105263157894735, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 35100 + }, + { + "epoch": 1.8105864661654136, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 35101 + }, + { + "epoch": 1.8106466165413533, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 35102 + }, + { + "epoch": 1.8107067669172934, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 35103 + }, + { + "epoch": 1.810766917293233, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 35104 + }, + { + "epoch": 1.810827067669173, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 35105 + }, + { + "epoch": 1.8108872180451128, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 35106 + }, + { + "epoch": 1.8109473684210526, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 35107 + }, + { + "epoch": 1.8110075187969925, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 35108 + }, + { + "epoch": 1.8110676691729324, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 35109 + }, + { + "epoch": 1.8111278195488723, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 35110 + }, + { + "epoch": 1.811187969924812, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 35111 + }, + { + "epoch": 1.811248120300752, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 35112 + }, + { + "epoch": 1.8113082706766916, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 35113 + }, + { + "epoch": 1.8113684210526317, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 35114 + }, + { + "epoch": 1.8114285714285714, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 35115 + }, + { + "epoch": 1.8114887218045113, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 35116 + }, + { + "epoch": 1.8115488721804511, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 35117 + }, + { + "epoch": 1.811609022556391, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 35118 + }, + { + "epoch": 1.8116691729323309, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 35119 + }, + { + "epoch": 1.8117293233082705, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 35120 + }, + { + "epoch": 1.8117894736842106, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 35121 + }, + { + "epoch": 1.8118496240601503, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 35122 + }, + { + "epoch": 1.8119097744360904, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 35123 + }, + { + "epoch": 1.81196992481203, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 35124 + }, + { + "epoch": 1.8120300751879699, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 35125 + }, + { + "epoch": 1.8120902255639098, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 35126 + }, + { + "epoch": 1.8121503759398496, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 35127 + }, + { + "epoch": 1.8122105263157895, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 35128 + }, + { + "epoch": 1.8122706766917294, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 35129 + }, + { + "epoch": 1.8123308270676692, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 35130 + }, + { + "epoch": 1.8123909774436089, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 35131 + }, + { + "epoch": 1.812451127819549, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 35132 + }, + { + "epoch": 1.8125112781954886, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 35133 + }, + { + "epoch": 1.8125714285714287, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 35134 + }, + { + "epoch": 1.8126315789473684, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 35135 + }, + { + "epoch": 1.8126917293233082, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 35136 + }, + { + "epoch": 1.8127518796992481, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 35137 + }, + { + "epoch": 1.812812030075188, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 35138 + }, + { + "epoch": 1.8128721804511279, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 35139 + }, + { + "epoch": 1.8129323308270675, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 35140 + }, + { + "epoch": 1.8129924812030076, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 35141 + }, + { + "epoch": 1.8130526315789472, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 35142 + }, + { + "epoch": 1.8131127819548873, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 35143 + }, + { + "epoch": 1.813172932330827, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 35144 + }, + { + "epoch": 1.8132330827067669, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 35145 + }, + { + "epoch": 1.8132932330827067, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 35146 + }, + { + "epoch": 1.8133533834586466, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 35147 + }, + { + "epoch": 1.8134135338345865, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 35148 + }, + { + "epoch": 1.8134736842105263, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 35149 + }, + { + "epoch": 1.8135338345864662, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 35150 + }, + { + "epoch": 1.8135939849624059, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 35151 + }, + { + "epoch": 1.813654135338346, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 35152 + }, + { + "epoch": 1.8137142857142856, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 35153 + }, + { + "epoch": 1.8137744360902257, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 35154 + }, + { + "epoch": 1.8138345864661654, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 35155 + }, + { + "epoch": 1.8138947368421052, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 35156 + }, + { + "epoch": 1.813954887218045, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 35157 + }, + { + "epoch": 1.814015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 35158 + }, + { + "epoch": 1.8140751879699248, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 35159 + }, + { + "epoch": 1.8141353383458647, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 35160 + }, + { + "epoch": 1.8141954887218046, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 35161 + }, + { + "epoch": 1.8142556390977442, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 35162 + }, + { + "epoch": 1.8143157894736843, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 35163 + }, + { + "epoch": 1.814375939849624, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 35164 + }, + { + "epoch": 1.814436090225564, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 35165 + }, + { + "epoch": 1.8144962406015037, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 35166 + }, + { + "epoch": 1.8145563909774436, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 35167 + }, + { + "epoch": 1.8146165413533835, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 35168 + }, + { + "epoch": 1.8146766917293233, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 35169 + }, + { + "epoch": 1.8147368421052632, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 35170 + }, + { + "epoch": 1.8147969924812029, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 35171 + }, + { + "epoch": 1.814857142857143, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 35172 + }, + { + "epoch": 1.8149172932330826, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 35173 + }, + { + "epoch": 1.8149774436090227, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 35174 + }, + { + "epoch": 1.8150375939849623, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 35175 + }, + { + "epoch": 1.8150977443609022, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 35176 + }, + { + "epoch": 1.815157894736842, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 35177 + }, + { + "epoch": 1.815218045112782, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 35178 + }, + { + "epoch": 1.8152781954887218, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 35179 + }, + { + "epoch": 1.8153383458646617, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 35180 + }, + { + "epoch": 1.8153984962406016, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 35181 + }, + { + "epoch": 1.8154586466165412, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 35182 + }, + { + "epoch": 1.8155187969924813, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 35183 + }, + { + "epoch": 1.815578947368421, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 35184 + }, + { + "epoch": 1.815639097744361, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 35185 + }, + { + "epoch": 1.8156992481203007, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9333, + "step": 35186 + }, + { + "epoch": 1.8157593984962406, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 35187 + }, + { + "epoch": 1.8158195488721804, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 35188 + }, + { + "epoch": 1.8158796992481203, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 35189 + }, + { + "epoch": 1.8159398496240602, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 35190 + }, + { + "epoch": 1.8159999999999998, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 35191 + }, + { + "epoch": 1.81606015037594, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 35192 + }, + { + "epoch": 1.8161203007518796, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 35193 + }, + { + "epoch": 1.8161804511278197, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 35194 + }, + { + "epoch": 1.8162406015037593, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 35195 + }, + { + "epoch": 1.8163007518796992, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 35196 + }, + { + "epoch": 1.816360902255639, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 35197 + }, + { + "epoch": 1.816421052631579, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1418, + "step": 35198 + }, + { + "epoch": 1.8164812030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 35199 + }, + { + "epoch": 1.8165413533834587, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 35200 + }, + { + "epoch": 1.8166015037593986, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 35201 + }, + { + "epoch": 1.8166616541353382, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 35202 + }, + { + "epoch": 1.8167218045112783, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 35203 + }, + { + "epoch": 1.816781954887218, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 35204 + }, + { + "epoch": 1.816842105263158, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 35205 + }, + { + "epoch": 1.8169022556390977, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 35206 + }, + { + "epoch": 1.8169624060150376, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 35207 + }, + { + "epoch": 1.8170225563909774, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 35208 + }, + { + "epoch": 1.8170827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 35209 + }, + { + "epoch": 1.8171428571428572, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1774, + "step": 35210 + }, + { + "epoch": 1.817203007518797, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 35211 + }, + { + "epoch": 1.817263157894737, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 35212 + }, + { + "epoch": 1.8173233082706766, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 35213 + }, + { + "epoch": 1.8173834586466167, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 35214 + }, + { + "epoch": 1.8174436090225563, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 35215 + }, + { + "epoch": 1.8175037593984964, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 35216 + }, + { + "epoch": 1.817563909774436, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 35217 + }, + { + "epoch": 1.817624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 35218 + }, + { + "epoch": 1.8176842105263158, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 35219 + }, + { + "epoch": 1.8177443609022557, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 35220 + }, + { + "epoch": 1.8178045112781955, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1906, + "step": 35221 + }, + { + "epoch": 1.8178646616541352, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 35222 + }, + { + "epoch": 1.8179248120300753, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 35223 + }, + { + "epoch": 1.817984962406015, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 35224 + }, + { + "epoch": 1.818045112781955, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 35225 + }, + { + "epoch": 1.8181052631578947, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 35226 + }, + { + "epoch": 1.8181654135338345, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 35227 + }, + { + "epoch": 1.8182255639097744, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 35228 + }, + { + "epoch": 1.8182857142857143, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 35229 + }, + { + "epoch": 1.8183458646616542, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 35230 + }, + { + "epoch": 1.818406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 35231 + }, + { + "epoch": 1.818466165413534, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 35232 + }, + { + "epoch": 1.8185263157894735, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 35233 + }, + { + "epoch": 1.8185864661654136, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 35234 + }, + { + "epoch": 1.8186466165413533, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 35235 + }, + { + "epoch": 1.8187067669172934, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 35236 + }, + { + "epoch": 1.818766917293233, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 35237 + }, + { + "epoch": 1.818827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 35238 + }, + { + "epoch": 1.8188872180451128, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9403, + "step": 35239 + }, + { + "epoch": 1.8189473684210526, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1418, + "step": 35240 + }, + { + "epoch": 1.8190075187969925, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 35241 + }, + { + "epoch": 1.8190676691729322, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 35242 + }, + { + "epoch": 1.8191278195488723, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 35243 + }, + { + "epoch": 1.819187969924812, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 35244 + }, + { + "epoch": 1.819248120300752, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 35245 + }, + { + "epoch": 1.8193082706766917, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 35246 + }, + { + "epoch": 1.8193684210526315, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 35247 + }, + { + "epoch": 1.8194285714285714, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 35248 + }, + { + "epoch": 1.8194887218045113, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 35249 + }, + { + "epoch": 1.8195488721804511, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 35250 + }, + { + "epoch": 1.819609022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 35251 + }, + { + "epoch": 1.8196691729323309, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 35252 + }, + { + "epoch": 1.8197293233082705, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9251, + "step": 35253 + }, + { + "epoch": 1.8197894736842106, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 35254 + }, + { + "epoch": 1.8198496240601503, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 35255 + }, + { + "epoch": 1.8199097744360904, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 35256 + }, + { + "epoch": 1.81996992481203, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 35257 + }, + { + "epoch": 1.8200300751879699, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 35258 + }, + { + "epoch": 1.8200902255639098, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 35259 + }, + { + "epoch": 1.8201503759398496, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 35260 + }, + { + "epoch": 1.8202105263157895, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 35261 + }, + { + "epoch": 1.8202706766917294, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 35262 + }, + { + "epoch": 1.8203308270676692, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 35263 + }, + { + "epoch": 1.820390977443609, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 35264 + }, + { + "epoch": 1.820451127819549, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 35265 + }, + { + "epoch": 1.8205112781954886, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 35266 + }, + { + "epoch": 1.8205714285714287, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 35267 + }, + { + "epoch": 1.8206315789473684, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 35268 + }, + { + "epoch": 1.8206917293233082, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 35269 + }, + { + "epoch": 1.8207518796992481, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 35270 + }, + { + "epoch": 1.820812030075188, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 35271 + }, + { + "epoch": 1.8208721804511279, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 35272 + }, + { + "epoch": 1.8209323308270675, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.225, + "step": 35273 + }, + { + "epoch": 1.8209924812030076, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 35274 + }, + { + "epoch": 1.8210526315789473, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 35275 + }, + { + "epoch": 1.8211127819548873, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 35276 + }, + { + "epoch": 1.821172932330827, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 35277 + }, + { + "epoch": 1.8212330827067669, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 35278 + }, + { + "epoch": 1.8212932330827067, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 35279 + }, + { + "epoch": 1.8213533834586466, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.3341, + "step": 35280 + }, + { + "epoch": 1.8214135338345865, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 35281 + }, + { + "epoch": 1.8214736842105264, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 35282 + }, + { + "epoch": 1.8215338345864662, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 35283 + }, + { + "epoch": 1.8215939849624059, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1597, + "step": 35284 + }, + { + "epoch": 1.821654135338346, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 35285 + }, + { + "epoch": 1.8217142857142856, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 35286 + }, + { + "epoch": 1.8217744360902257, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 35287 + }, + { + "epoch": 1.8218345864661654, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 35288 + }, + { + "epoch": 1.8218947368421052, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 35289 + }, + { + "epoch": 1.821954887218045, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 35290 + }, + { + "epoch": 1.822015037593985, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 35291 + }, + { + "epoch": 1.8220751879699248, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 35292 + }, + { + "epoch": 1.8221353383458645, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 35293 + }, + { + "epoch": 1.8221954887218046, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 35294 + }, + { + "epoch": 1.8222556390977442, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 35295 + }, + { + "epoch": 1.8223157894736843, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 35296 + }, + { + "epoch": 1.822375939849624, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 35297 + }, + { + "epoch": 1.822436090225564, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 35298 + }, + { + "epoch": 1.8224962406015037, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 35299 + }, + { + "epoch": 1.8225563909774436, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 35300 + }, + { + "epoch": 1.8226165413533835, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 35301 + }, + { + "epoch": 1.8226766917293233, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 35302 + }, + { + "epoch": 1.8227368421052632, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 35303 + }, + { + "epoch": 1.8227969924812029, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 35304 + }, + { + "epoch": 1.822857142857143, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 35305 + }, + { + "epoch": 1.8229172932330826, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 35306 + }, + { + "epoch": 1.8229774436090227, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 35307 + }, + { + "epoch": 1.8230375939849623, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 35308 + }, + { + "epoch": 1.8230977443609022, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 35309 + }, + { + "epoch": 1.823157894736842, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 35310 + }, + { + "epoch": 1.823218045112782, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 35311 + }, + { + "epoch": 1.8232781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 35312 + }, + { + "epoch": 1.8233383458646617, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 35313 + }, + { + "epoch": 1.8233984962406016, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 35314 + }, + { + "epoch": 1.8234586466165412, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 35315 + }, + { + "epoch": 1.8235187969924813, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 35316 + }, + { + "epoch": 1.823578947368421, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 35317 + }, + { + "epoch": 1.823639097744361, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 35318 + }, + { + "epoch": 1.8236992481203007, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 35319 + }, + { + "epoch": 1.8237593984962406, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 35320 + }, + { + "epoch": 1.8238195488721805, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 35321 + }, + { + "epoch": 1.8238796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 35322 + }, + { + "epoch": 1.8239398496240602, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 35323 + }, + { + "epoch": 1.8239999999999998, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.3324, + "step": 35324 + }, + { + "epoch": 1.82406015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 35325 + }, + { + "epoch": 1.8241203007518796, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.2118, + "step": 35326 + }, + { + "epoch": 1.8241804511278197, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 35327 + }, + { + "epoch": 1.8242406015037593, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 35328 + }, + { + "epoch": 1.8243007518796992, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 35329 + }, + { + "epoch": 1.824360902255639, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 35330 + }, + { + "epoch": 1.824421052631579, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 35331 + }, + { + "epoch": 1.8244812030075188, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 35332 + }, + { + "epoch": 1.8245413533834587, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 35333 + }, + { + "epoch": 1.8246015037593986, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 35334 + }, + { + "epoch": 1.8246616541353382, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 35335 + }, + { + "epoch": 1.8247218045112783, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 35336 + }, + { + "epoch": 1.824781954887218, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 35337 + }, + { + "epoch": 1.824842105263158, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 35338 + }, + { + "epoch": 1.8249022556390977, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 35339 + }, + { + "epoch": 1.8249624060150376, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 35340 + }, + { + "epoch": 1.8250225563909774, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 35341 + }, + { + "epoch": 1.8250827067669173, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 35342 + }, + { + "epoch": 1.8251428571428572, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 35343 + }, + { + "epoch": 1.8252030075187968, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 35344 + }, + { + "epoch": 1.825263157894737, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 35345 + }, + { + "epoch": 1.8253233082706766, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 35346 + }, + { + "epoch": 1.8253834586466167, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 35347 + }, + { + "epoch": 1.8254436090225563, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 35348 + }, + { + "epoch": 1.8255037593984964, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 35349 + }, + { + "epoch": 1.825563909774436, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 35350 + }, + { + "epoch": 1.825624060150376, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 35351 + }, + { + "epoch": 1.8256842105263158, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 35352 + }, + { + "epoch": 1.8257443609022557, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 35353 + }, + { + "epoch": 1.8258045112781955, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 35354 + }, + { + "epoch": 1.8258646616541352, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 35355 + }, + { + "epoch": 1.8259248120300753, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 35356 + }, + { + "epoch": 1.825984962406015, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 35357 + }, + { + "epoch": 1.826045112781955, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1705, + "step": 35358 + }, + { + "epoch": 1.8261052631578947, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 35359 + }, + { + "epoch": 1.8261654135338345, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 35360 + }, + { + "epoch": 1.8262255639097744, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 35361 + }, + { + "epoch": 1.8262857142857143, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 35362 + }, + { + "epoch": 1.8263458646616542, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 35363 + }, + { + "epoch": 1.826406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 35364 + }, + { + "epoch": 1.826466165413534, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 35365 + }, + { + "epoch": 1.8265263157894736, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 35366 + }, + { + "epoch": 1.8265864661654136, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 35367 + }, + { + "epoch": 1.8266466165413533, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 35368 + }, + { + "epoch": 1.8267067669172934, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 35369 + }, + { + "epoch": 1.826766917293233, + "grad_norm": 2.1875, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 35370 + }, + { + "epoch": 1.826827067669173, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 35371 + }, + { + "epoch": 1.8268872180451128, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 35372 + }, + { + "epoch": 1.8269473684210527, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 35373 + }, + { + "epoch": 1.8270075187969925, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 35374 + }, + { + "epoch": 1.8270676691729322, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 35375 + }, + { + "epoch": 1.8271278195488723, + "grad_norm": 1.4375, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 35376 + }, + { + "epoch": 1.827187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 35377 + }, + { + "epoch": 1.827248120300752, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 35378 + }, + { + "epoch": 1.8273082706766917, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 35379 + }, + { + "epoch": 1.8273684210526315, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 35380 + }, + { + "epoch": 1.8274285714285714, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 35381 + }, + { + "epoch": 1.8274887218045113, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 35382 + }, + { + "epoch": 1.8275488721804511, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 35383 + }, + { + "epoch": 1.827609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 35384 + }, + { + "epoch": 1.8276691729323309, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 35385 + }, + { + "epoch": 1.8277293233082705, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 35386 + }, + { + "epoch": 1.8277894736842106, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 35387 + }, + { + "epoch": 1.8278496240601503, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 35388 + }, + { + "epoch": 1.8279097744360904, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9099, + "step": 35389 + }, + { + "epoch": 1.82796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 35390 + }, + { + "epoch": 1.82803007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 35391 + }, + { + "epoch": 1.8280902255639098, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.2101, + "step": 35392 + }, + { + "epoch": 1.8281503759398496, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 35393 + }, + { + "epoch": 1.8282105263157895, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 35394 + }, + { + "epoch": 1.8282706766917292, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 35395 + }, + { + "epoch": 1.8283308270676693, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 35396 + }, + { + "epoch": 1.828390977443609, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 35397 + }, + { + "epoch": 1.828451127819549, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 35398 + }, + { + "epoch": 1.8285112781954886, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 35399 + }, + { + "epoch": 1.8285714285714287, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 35400 + }, + { + "epoch": 1.8286315789473684, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 35401 + }, + { + "epoch": 1.8286917293233083, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 35402 + }, + { + "epoch": 1.8287518796992481, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 35403 + }, + { + "epoch": 1.828812030075188, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 35404 + }, + { + "epoch": 1.8288721804511279, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 35405 + }, + { + "epoch": 1.8289323308270675, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 35406 + }, + { + "epoch": 1.8289924812030076, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 35407 + }, + { + "epoch": 1.8290526315789473, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 35408 + }, + { + "epoch": 1.8291127819548874, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 35409 + }, + { + "epoch": 1.829172932330827, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 35410 + }, + { + "epoch": 1.8292330827067669, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 35411 + }, + { + "epoch": 1.8292932330827067, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 35412 + }, + { + "epoch": 1.8293533834586466, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 35413 + }, + { + "epoch": 1.8294135338345865, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 35414 + }, + { + "epoch": 1.8294736842105264, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 35415 + }, + { + "epoch": 1.8295338345864662, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 35416 + }, + { + "epoch": 1.8295939849624059, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.8818, + "step": 35417 + }, + { + "epoch": 1.829654135338346, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 35418 + }, + { + "epoch": 1.8297142857142856, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 35419 + }, + { + "epoch": 1.8297744360902257, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 35420 + }, + { + "epoch": 1.8298345864661654, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 35421 + }, + { + "epoch": 1.8298947368421052, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 35422 + }, + { + "epoch": 1.8299548872180451, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 35423 + }, + { + "epoch": 1.830015037593985, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.248, + "step": 35424 + }, + { + "epoch": 1.8300751879699249, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 35425 + }, + { + "epoch": 1.8301353383458645, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 35426 + }, + { + "epoch": 1.8301954887218046, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 35427 + }, + { + "epoch": 1.8302556390977442, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 35428 + }, + { + "epoch": 1.8303157894736843, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 35429 + }, + { + "epoch": 1.830375939849624, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1667, + "step": 35430 + }, + { + "epoch": 1.8304360902255639, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 35431 + }, + { + "epoch": 1.8304962406015037, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 35432 + }, + { + "epoch": 1.8305563909774436, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 35433 + }, + { + "epoch": 1.8306165413533835, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 35434 + }, + { + "epoch": 1.8306766917293233, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 35435 + }, + { + "epoch": 1.8307368421052632, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 35436 + }, + { + "epoch": 1.8307969924812029, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9015, + "step": 35437 + }, + { + "epoch": 1.830857142857143, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.156, + "step": 35438 + }, + { + "epoch": 1.8309172932330826, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 35439 + }, + { + "epoch": 1.8309774436090227, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 35440 + }, + { + "epoch": 1.8310375939849624, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 35441 + }, + { + "epoch": 1.8310977443609022, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 35442 + }, + { + "epoch": 1.831157894736842, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 35443 + }, + { + "epoch": 1.831218045112782, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 35444 + }, + { + "epoch": 1.8312781954887218, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 35445 + }, + { + "epoch": 1.8313383458646615, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 35446 + }, + { + "epoch": 1.8313984962406016, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 35447 + }, + { + "epoch": 1.8314586466165412, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 35448 + }, + { + "epoch": 1.8315187969924813, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 35449 + }, + { + "epoch": 1.831578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 35450 + }, + { + "epoch": 1.831639097744361, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 35451 + }, + { + "epoch": 1.8316992481203007, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 35452 + }, + { + "epoch": 1.8317593984962406, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 35453 + }, + { + "epoch": 1.8318195488721805, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 35454 + }, + { + "epoch": 1.8318796992481203, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 35455 + }, + { + "epoch": 1.8319398496240602, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 35456 + }, + { + "epoch": 1.8319999999999999, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 35457 + }, + { + "epoch": 1.83206015037594, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 35458 + }, + { + "epoch": 1.8321203007518796, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 35459 + }, + { + "epoch": 1.8321804511278197, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 35460 + }, + { + "epoch": 1.8322406015037593, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 35461 + }, + { + "epoch": 1.8323007518796992, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 35462 + }, + { + "epoch": 1.832360902255639, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 35463 + }, + { + "epoch": 1.832421052631579, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 35464 + }, + { + "epoch": 1.8324812030075188, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 35465 + }, + { + "epoch": 1.8325413533834587, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 35466 + }, + { + "epoch": 1.8326015037593986, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 35467 + }, + { + "epoch": 1.8326616541353382, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 35468 + }, + { + "epoch": 1.8327218045112783, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 35469 + }, + { + "epoch": 1.832781954887218, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 35470 + }, + { + "epoch": 1.832842105263158, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 35471 + }, + { + "epoch": 1.8329022556390977, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 35472 + }, + { + "epoch": 1.8329624060150376, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 35473 + }, + { + "epoch": 1.8330225563909774, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 35474 + }, + { + "epoch": 1.8330827067669173, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 35475 + }, + { + "epoch": 1.8331428571428572, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 35476 + }, + { + "epoch": 1.8332030075187968, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 35477 + }, + { + "epoch": 1.833263157894737, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 35478 + }, + { + "epoch": 1.8333233082706766, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 35479 + }, + { + "epoch": 1.8333834586466167, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 35480 + }, + { + "epoch": 1.8334436090225563, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 35481 + }, + { + "epoch": 1.8335037593984962, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 35482 + }, + { + "epoch": 1.833563909774436, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 35483 + }, + { + "epoch": 1.833624060150376, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 35484 + }, + { + "epoch": 1.8336842105263158, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 35485 + }, + { + "epoch": 1.8337443609022557, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1313, + "step": 35486 + }, + { + "epoch": 1.8338045112781955, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 35487 + }, + { + "epoch": 1.8338646616541352, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 35488 + }, + { + "epoch": 1.8339248120300753, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 35489 + }, + { + "epoch": 1.833984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 35490 + }, + { + "epoch": 1.834045112781955, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 35491 + }, + { + "epoch": 1.8341052631578947, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.2295, + "step": 35492 + }, + { + "epoch": 1.8341654135338346, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 35493 + }, + { + "epoch": 1.8342255639097744, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 35494 + }, + { + "epoch": 1.8342857142857143, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 35495 + }, + { + "epoch": 1.8343458646616542, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 35496 + }, + { + "epoch": 1.834406015037594, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1315, + "step": 35497 + }, + { + "epoch": 1.834466165413534, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 35498 + }, + { + "epoch": 1.8345263157894736, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 35499 + }, + { + "epoch": 1.8345864661654137, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9372, + "step": 35500 + }, + { + "epoch": 1.8346466165413533, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 35501 + }, + { + "epoch": 1.8347067669172934, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 35502 + }, + { + "epoch": 1.834766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 35503 + }, + { + "epoch": 1.834827067669173, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 35504 + }, + { + "epoch": 1.8348872180451128, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 35505 + }, + { + "epoch": 1.8349473684210527, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 35506 + }, + { + "epoch": 1.8350075187969925, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 35507 + }, + { + "epoch": 1.8350676691729322, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2771, + "step": 35508 + }, + { + "epoch": 1.8351278195488723, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 35509 + }, + { + "epoch": 1.835187969924812, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 35510 + }, + { + "epoch": 1.835248120300752, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 35511 + }, + { + "epoch": 1.8353082706766917, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 35512 + }, + { + "epoch": 1.8353684210526315, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 35513 + }, + { + "epoch": 1.8354285714285714, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 35514 + }, + { + "epoch": 1.8354887218045113, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 35515 + }, + { + "epoch": 1.8355488721804512, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 35516 + }, + { + "epoch": 1.835609022556391, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 35517 + }, + { + "epoch": 1.835669172932331, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 35518 + }, + { + "epoch": 1.8357293233082705, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9345, + "step": 35519 + }, + { + "epoch": 1.8357894736842106, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 35520 + }, + { + "epoch": 1.8358496240601503, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 35521 + }, + { + "epoch": 1.8359097744360904, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 35522 + }, + { + "epoch": 1.83596992481203, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 35523 + }, + { + "epoch": 1.83603007518797, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 35524 + }, + { + "epoch": 1.8360902255639098, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 35525 + }, + { + "epoch": 1.8361503759398496, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 35526 + }, + { + "epoch": 1.8362105263157895, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 35527 + }, + { + "epoch": 1.8362706766917292, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 35528 + }, + { + "epoch": 1.8363308270676693, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 35529 + }, + { + "epoch": 1.836390977443609, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 35530 + }, + { + "epoch": 1.836451127819549, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 35531 + }, + { + "epoch": 1.8365112781954887, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 35532 + }, + { + "epoch": 1.8365714285714285, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 35533 + }, + { + "epoch": 1.8366315789473684, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 35534 + }, + { + "epoch": 1.8366917293233083, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 35535 + }, + { + "epoch": 1.8367518796992481, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 35536 + }, + { + "epoch": 1.836812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 35537 + }, + { + "epoch": 1.8368721804511279, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 35538 + }, + { + "epoch": 1.8369323308270675, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 35539 + }, + { + "epoch": 1.8369924812030076, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 35540 + }, + { + "epoch": 1.8370526315789473, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 35541 + }, + { + "epoch": 1.8371127819548874, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 35542 + }, + { + "epoch": 1.837172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 35543 + }, + { + "epoch": 1.8372330827067669, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 35544 + }, + { + "epoch": 1.8372932330827068, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 35545 + }, + { + "epoch": 1.8373533834586466, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 35546 + }, + { + "epoch": 1.8374135338345865, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 35547 + }, + { + "epoch": 1.8374736842105264, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 35548 + }, + { + "epoch": 1.8375338345864662, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 35549 + }, + { + "epoch": 1.837593984962406, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.2357, + "step": 35550 + }, + { + "epoch": 1.837654135338346, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 35551 + }, + { + "epoch": 1.8377142857142856, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 35552 + }, + { + "epoch": 1.8377744360902257, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1603, + "step": 35553 + }, + { + "epoch": 1.8378345864661654, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 35554 + }, + { + "epoch": 1.8378947368421052, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 35555 + }, + { + "epoch": 1.8379548872180451, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 35556 + }, + { + "epoch": 1.838015037593985, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 35557 + }, + { + "epoch": 1.8380751879699249, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 35558 + }, + { + "epoch": 1.8381353383458645, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 35559 + }, + { + "epoch": 1.8381954887218046, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 35560 + }, + { + "epoch": 1.8382556390977443, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 35561 + }, + { + "epoch": 1.8383157894736843, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 35562 + }, + { + "epoch": 1.838375939849624, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 35563 + }, + { + "epoch": 1.8384360902255639, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 35564 + }, + { + "epoch": 1.8384962406015037, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 35565 + }, + { + "epoch": 1.8385563909774436, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 35566 + }, + { + "epoch": 1.8386165413533835, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 35567 + }, + { + "epoch": 1.8386766917293234, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 35568 + }, + { + "epoch": 1.8387368421052632, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 35569 + }, + { + "epoch": 1.8387969924812029, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 35570 + }, + { + "epoch": 1.838857142857143, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 35571 + }, + { + "epoch": 1.8389172932330826, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 35572 + }, + { + "epoch": 1.8389774436090227, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 35573 + }, + { + "epoch": 1.8390375939849624, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 35574 + }, + { + "epoch": 1.8390977443609022, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 35575 + }, + { + "epoch": 1.839157894736842, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 35576 + }, + { + "epoch": 1.839218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 35577 + }, + { + "epoch": 1.8392781954887218, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 35578 + }, + { + "epoch": 1.8393383458646615, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 35579 + }, + { + "epoch": 1.8393984962406016, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 35580 + }, + { + "epoch": 1.8394586466165412, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 35581 + }, + { + "epoch": 1.8395187969924813, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 35582 + }, + { + "epoch": 1.839578947368421, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 35583 + }, + { + "epoch": 1.8396390977443609, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 35584 + }, + { + "epoch": 1.8396992481203007, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 35585 + }, + { + "epoch": 1.8397593984962406, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 35586 + }, + { + "epoch": 1.8398195488721805, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 35587 + }, + { + "epoch": 1.8398796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 35588 + }, + { + "epoch": 1.8399398496240602, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 35589 + }, + { + "epoch": 1.8399999999999999, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 35590 + }, + { + "epoch": 1.84006015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 35591 + }, + { + "epoch": 1.8401203007518796, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 35592 + }, + { + "epoch": 1.8401804511278197, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 35593 + }, + { + "epoch": 1.8402406015037593, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 35594 + }, + { + "epoch": 1.8403007518796992, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 35595 + }, + { + "epoch": 1.840360902255639, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 35596 + }, + { + "epoch": 1.840421052631579, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 35597 + }, + { + "epoch": 1.8404812030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 35598 + }, + { + "epoch": 1.8405413533834587, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 35599 + }, + { + "epoch": 1.8406015037593986, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 35600 + }, + { + "epoch": 1.8406616541353382, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 35601 + }, + { + "epoch": 1.8407218045112783, + "grad_norm": 1.7578125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 35602 + }, + { + "epoch": 1.840781954887218, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 35603 + }, + { + "epoch": 1.840842105263158, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 35604 + }, + { + "epoch": 1.8409022556390977, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 35605 + }, + { + "epoch": 1.8409624060150376, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 35606 + }, + { + "epoch": 1.8410225563909774, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 35607 + }, + { + "epoch": 1.8410827067669173, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 35608 + }, + { + "epoch": 1.8411428571428572, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 35609 + }, + { + "epoch": 1.8412030075187968, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1876, + "step": 35610 + }, + { + "epoch": 1.841263157894737, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 35611 + }, + { + "epoch": 1.8413233082706766, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 35612 + }, + { + "epoch": 1.8413834586466167, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 35613 + }, + { + "epoch": 1.8414436090225563, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.1942, + "step": 35614 + }, + { + "epoch": 1.8415037593984962, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 35615 + }, + { + "epoch": 1.841563909774436, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 35616 + }, + { + "epoch": 1.841624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 35617 + }, + { + "epoch": 1.8416842105263158, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 35618 + }, + { + "epoch": 1.8417443609022557, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 35619 + }, + { + "epoch": 1.8418045112781956, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 35620 + }, + { + "epoch": 1.8418646616541352, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1929, + "step": 35621 + }, + { + "epoch": 1.8419248120300753, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 35622 + }, + { + "epoch": 1.841984962406015, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 35623 + }, + { + "epoch": 1.842045112781955, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 35624 + }, + { + "epoch": 1.8421052631578947, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 35625 + }, + { + "epoch": 1.8421654135338346, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 35626 + }, + { + "epoch": 1.8422255639097744, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 35627 + }, + { + "epoch": 1.8422857142857143, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1214, + "step": 35628 + }, + { + "epoch": 1.8423458646616542, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 35629 + }, + { + "epoch": 1.8424060150375938, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 35630 + }, + { + "epoch": 1.842466165413534, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 35631 + }, + { + "epoch": 1.8425263157894736, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 35632 + }, + { + "epoch": 1.8425864661654137, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 35633 + }, + { + "epoch": 1.8426466165413533, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 35634 + }, + { + "epoch": 1.8427067669172932, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 35635 + }, + { + "epoch": 1.842766917293233, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 35636 + }, + { + "epoch": 1.842827067669173, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 35637 + }, + { + "epoch": 1.8428872180451128, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 35638 + }, + { + "epoch": 1.8429473684210527, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 35639 + }, + { + "epoch": 1.8430075187969925, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 35640 + }, + { + "epoch": 1.8430676691729322, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 35641 + }, + { + "epoch": 1.8431278195488723, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 35642 + }, + { + "epoch": 1.843187969924812, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 35643 + }, + { + "epoch": 1.843248120300752, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 35644 + }, + { + "epoch": 1.8433082706766917, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9181, + "step": 35645 + }, + { + "epoch": 1.8433684210526315, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 35646 + }, + { + "epoch": 1.8434285714285714, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 35647 + }, + { + "epoch": 1.8434887218045113, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 35648 + }, + { + "epoch": 1.8435488721804512, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 35649 + }, + { + "epoch": 1.843609022556391, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 35650 + }, + { + "epoch": 1.843669172932331, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 35651 + }, + { + "epoch": 1.8437293233082706, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 35652 + }, + { + "epoch": 1.8437894736842106, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 35653 + }, + { + "epoch": 1.8438496240601503, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 35654 + }, + { + "epoch": 1.8439097744360904, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 35655 + }, + { + "epoch": 1.84396992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 35656 + }, + { + "epoch": 1.84403007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 35657 + }, + { + "epoch": 1.8440902255639098, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 35658 + }, + { + "epoch": 1.8441503759398497, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9339, + "step": 35659 + }, + { + "epoch": 1.8442105263157895, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 35660 + }, + { + "epoch": 1.8442706766917292, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 35661 + }, + { + "epoch": 1.8443308270676693, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 35662 + }, + { + "epoch": 1.844390977443609, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 35663 + }, + { + "epoch": 1.844451127819549, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 35664 + }, + { + "epoch": 1.8445112781954887, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 35665 + }, + { + "epoch": 1.8445714285714285, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 35666 + }, + { + "epoch": 1.8446315789473684, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 35667 + }, + { + "epoch": 1.8446917293233083, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 35668 + }, + { + "epoch": 1.8447518796992481, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 35669 + }, + { + "epoch": 1.844812030075188, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 35670 + }, + { + "epoch": 1.8448721804511279, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 35671 + }, + { + "epoch": 1.8449323308270675, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 35672 + }, + { + "epoch": 1.8449924812030076, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1409, + "step": 35673 + }, + { + "epoch": 1.8450526315789473, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 35674 + }, + { + "epoch": 1.8451127819548874, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 35675 + }, + { + "epoch": 1.845172932330827, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 35676 + }, + { + "epoch": 1.845233082706767, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 35677 + }, + { + "epoch": 1.8452932330827068, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 35678 + }, + { + "epoch": 1.8453533834586466, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 35679 + }, + { + "epoch": 1.8454135338345865, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 35680 + }, + { + "epoch": 1.8454736842105262, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 35681 + }, + { + "epoch": 1.8455338345864662, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 35682 + }, + { + "epoch": 1.845593984962406, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9519, + "step": 35683 + }, + { + "epoch": 1.845654135338346, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 35684 + }, + { + "epoch": 1.8457142857142856, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 35685 + }, + { + "epoch": 1.8457744360902255, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 35686 + }, + { + "epoch": 1.8458345864661654, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 35687 + }, + { + "epoch": 1.8458947368421053, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.3413, + "step": 35688 + }, + { + "epoch": 1.8459548872180451, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 35689 + }, + { + "epoch": 1.846015037593985, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1485, + "step": 35690 + }, + { + "epoch": 1.8460751879699249, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 35691 + }, + { + "epoch": 1.8461353383458645, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 35692 + }, + { + "epoch": 1.8461954887218046, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 35693 + }, + { + "epoch": 1.8462556390977443, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 35694 + }, + { + "epoch": 1.8463157894736844, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 35695 + }, + { + "epoch": 1.846375939849624, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 35696 + }, + { + "epoch": 1.8464360902255639, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 35697 + }, + { + "epoch": 1.8464962406015037, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 35698 + }, + { + "epoch": 1.8465563909774436, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 35699 + }, + { + "epoch": 1.8466165413533835, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 35700 + }, + { + "epoch": 1.8466766917293234, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 35701 + }, + { + "epoch": 1.8467368421052632, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 35702 + }, + { + "epoch": 1.8467969924812029, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 35703 + }, + { + "epoch": 1.846857142857143, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.92, + "step": 35704 + }, + { + "epoch": 1.8469172932330826, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 35705 + }, + { + "epoch": 1.8469774436090227, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 35706 + }, + { + "epoch": 1.8470375939849624, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 35707 + }, + { + "epoch": 1.8470977443609022, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 35708 + }, + { + "epoch": 1.847157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 35709 + }, + { + "epoch": 1.847218045112782, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 35710 + }, + { + "epoch": 1.8472781954887219, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 35711 + }, + { + "epoch": 1.8473383458646615, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 35712 + }, + { + "epoch": 1.8473984962406016, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 35713 + }, + { + "epoch": 1.8474586466165412, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 35714 + }, + { + "epoch": 1.8475187969924813, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 35715 + }, + { + "epoch": 1.847578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 35716 + }, + { + "epoch": 1.8476390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 35717 + }, + { + "epoch": 1.8476992481203007, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 35718 + }, + { + "epoch": 1.8477593984962406, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 35719 + }, + { + "epoch": 1.8478195488721805, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 35720 + }, + { + "epoch": 1.8478796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 35721 + }, + { + "epoch": 1.8479398496240602, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 35722 + }, + { + "epoch": 1.8479999999999999, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 35723 + }, + { + "epoch": 1.84806015037594, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 35724 + }, + { + "epoch": 1.8481203007518796, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 35725 + }, + { + "epoch": 1.8481804511278197, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9407, + "step": 35726 + }, + { + "epoch": 1.8482406015037594, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 35727 + }, + { + "epoch": 1.8483007518796992, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 35728 + }, + { + "epoch": 1.848360902255639, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 35729 + }, + { + "epoch": 1.848421052631579, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 35730 + }, + { + "epoch": 1.8484812030075188, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 35731 + }, + { + "epoch": 1.8485413533834585, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 35732 + }, + { + "epoch": 1.8486015037593986, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 35733 + }, + { + "epoch": 1.8486616541353382, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 35734 + }, + { + "epoch": 1.8487218045112783, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 35735 + }, + { + "epoch": 1.848781954887218, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 35736 + }, + { + "epoch": 1.8488421052631578, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 35737 + }, + { + "epoch": 1.8489022556390977, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 35738 + }, + { + "epoch": 1.8489624060150376, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 35739 + }, + { + "epoch": 1.8490225563909775, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 35740 + }, + { + "epoch": 1.8490827067669173, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 35741 + }, + { + "epoch": 1.8491428571428572, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 35742 + }, + { + "epoch": 1.8492030075187968, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 35743 + }, + { + "epoch": 1.849263157894737, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 35744 + }, + { + "epoch": 1.8493233082706766, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9374, + "step": 35745 + }, + { + "epoch": 1.8493834586466167, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 35746 + }, + { + "epoch": 1.8494436090225563, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 35747 + }, + { + "epoch": 1.8495037593984962, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 35748 + }, + { + "epoch": 1.849563909774436, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 35749 + }, + { + "epoch": 1.849624060150376, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 35750 + }, + { + "epoch": 1.8496842105263158, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 35751 + }, + { + "epoch": 1.8497443609022557, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 35752 + }, + { + "epoch": 1.8498045112781956, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 35753 + }, + { + "epoch": 1.8498646616541352, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 35754 + }, + { + "epoch": 1.8499248120300753, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 35755 + }, + { + "epoch": 1.849984962406015, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 35756 + }, + { + "epoch": 1.850045112781955, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 35757 + }, + { + "epoch": 1.8501052631578947, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 35758 + }, + { + "epoch": 1.8501654135338346, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 35759 + }, + { + "epoch": 1.8502255639097744, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 35760 + }, + { + "epoch": 1.8502857142857143, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9222, + "step": 35761 + }, + { + "epoch": 1.8503458646616542, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 35762 + }, + { + "epoch": 1.8504060150375938, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 35763 + }, + { + "epoch": 1.850466165413534, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 35764 + }, + { + "epoch": 1.8505263157894736, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 35765 + }, + { + "epoch": 1.8505864661654137, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 35766 + }, + { + "epoch": 1.8506466165413533, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 35767 + }, + { + "epoch": 1.8507067669172932, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 35768 + }, + { + "epoch": 1.850766917293233, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 35769 + }, + { + "epoch": 1.850827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 35770 + }, + { + "epoch": 1.8508872180451128, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1338, + "step": 35771 + }, + { + "epoch": 1.8509473684210527, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 35772 + }, + { + "epoch": 1.8510075187969925, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 35773 + }, + { + "epoch": 1.8510676691729322, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 35774 + }, + { + "epoch": 1.8511278195488723, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 35775 + }, + { + "epoch": 1.851187969924812, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 35776 + }, + { + "epoch": 1.851248120300752, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 35777 + }, + { + "epoch": 1.8513082706766917, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 35778 + }, + { + "epoch": 1.8513684210526316, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 35779 + }, + { + "epoch": 1.8514285714285714, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 35780 + }, + { + "epoch": 1.8514887218045113, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 35781 + }, + { + "epoch": 1.8515488721804512, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 35782 + }, + { + "epoch": 1.8516090225563908, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 35783 + }, + { + "epoch": 1.851669172932331, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 35784 + }, + { + "epoch": 1.8517293233082706, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 35785 + }, + { + "epoch": 1.8517894736842107, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 35786 + }, + { + "epoch": 1.8518496240601503, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 35787 + }, + { + "epoch": 1.8519097744360902, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9291, + "step": 35788 + }, + { + "epoch": 1.85196992481203, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 35789 + }, + { + "epoch": 1.85203007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 35790 + }, + { + "epoch": 1.8520902255639098, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 35791 + }, + { + "epoch": 1.8521503759398497, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 35792 + }, + { + "epoch": 1.8522105263157895, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 35793 + }, + { + "epoch": 1.8522706766917292, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 35794 + }, + { + "epoch": 1.8523308270676693, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 35795 + }, + { + "epoch": 1.852390977443609, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 35796 + }, + { + "epoch": 1.852451127819549, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 35797 + }, + { + "epoch": 1.8525112781954887, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 35798 + }, + { + "epoch": 1.8525714285714285, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 35799 + }, + { + "epoch": 1.8526315789473684, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 35800 + }, + { + "epoch": 1.8526917293233083, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 35801 + }, + { + "epoch": 1.8527518796992481, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 35802 + }, + { + "epoch": 1.852812030075188, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 35803 + }, + { + "epoch": 1.852872180451128, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 35804 + }, + { + "epoch": 1.8529323308270675, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 35805 + }, + { + "epoch": 1.8529924812030076, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 35806 + }, + { + "epoch": 1.8530526315789473, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 35807 + }, + { + "epoch": 1.8531127819548874, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 35808 + }, + { + "epoch": 1.853172932330827, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 35809 + }, + { + "epoch": 1.853233082706767, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 35810 + }, + { + "epoch": 1.8532932330827068, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 35811 + }, + { + "epoch": 1.8533533834586466, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 35812 + }, + { + "epoch": 1.8534135338345865, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 35813 + }, + { + "epoch": 1.8534736842105262, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 35814 + }, + { + "epoch": 1.8535338345864663, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 35815 + }, + { + "epoch": 1.853593984962406, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 35816 + }, + { + "epoch": 1.853654135338346, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 35817 + }, + { + "epoch": 1.8537142857142856, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 35818 + }, + { + "epoch": 1.8537744360902255, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 35819 + }, + { + "epoch": 1.8538345864661654, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 35820 + }, + { + "epoch": 1.8538947368421053, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 35821 + }, + { + "epoch": 1.8539548872180451, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 35822 + }, + { + "epoch": 1.854015037593985, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 35823 + }, + { + "epoch": 1.8540751879699249, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 35824 + }, + { + "epoch": 1.8541353383458645, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 35825 + }, + { + "epoch": 1.8541954887218046, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 35826 + }, + { + "epoch": 1.8542556390977443, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 35827 + }, + { + "epoch": 1.8543157894736844, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 35828 + }, + { + "epoch": 1.854375939849624, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 35829 + }, + { + "epoch": 1.8544360902255639, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 35830 + }, + { + "epoch": 1.8544962406015038, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 35831 + }, + { + "epoch": 1.8545563909774436, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 35832 + }, + { + "epoch": 1.8546165413533835, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 35833 + }, + { + "epoch": 1.8546766917293231, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1793, + "step": 35834 + }, + { + "epoch": 1.8547368421052632, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 35835 + }, + { + "epoch": 1.8547969924812029, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 35836 + }, + { + "epoch": 1.854857142857143, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 35837 + }, + { + "epoch": 1.8549172932330826, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 35838 + }, + { + "epoch": 1.8549774436090225, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 35839 + }, + { + "epoch": 1.8550375939849624, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 35840 + }, + { + "epoch": 1.8550977443609022, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 35841 + }, + { + "epoch": 1.8551578947368421, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 35842 + }, + { + "epoch": 1.855218045112782, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 35843 + }, + { + "epoch": 1.8552781954887219, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 35844 + }, + { + "epoch": 1.8553383458646615, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 35845 + }, + { + "epoch": 1.8553984962406016, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 35846 + }, + { + "epoch": 1.8554586466165413, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 35847 + }, + { + "epoch": 1.8555187969924813, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 35848 + }, + { + "epoch": 1.855578947368421, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 35849 + }, + { + "epoch": 1.8556390977443609, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 35850 + }, + { + "epoch": 1.8556992481203007, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 35851 + }, + { + "epoch": 1.8557593984962406, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 35852 + }, + { + "epoch": 1.8558195488721805, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 35853 + }, + { + "epoch": 1.8558796992481204, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 35854 + }, + { + "epoch": 1.8559398496240602, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 35855 + }, + { + "epoch": 1.8559999999999999, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 35856 + }, + { + "epoch": 1.85606015037594, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 35857 + }, + { + "epoch": 1.8561203007518796, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 35858 + }, + { + "epoch": 1.8561804511278197, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2958, + "step": 35859 + }, + { + "epoch": 1.8562406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.263, + "step": 35860 + }, + { + "epoch": 1.8563007518796992, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 35861 + }, + { + "epoch": 1.856360902255639, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 35862 + }, + { + "epoch": 1.856421052631579, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 35863 + }, + { + "epoch": 1.8564812030075188, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 35864 + }, + { + "epoch": 1.8565413533834585, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 35865 + }, + { + "epoch": 1.8566015037593986, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 35866 + }, + { + "epoch": 1.8566616541353382, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 35867 + }, + { + "epoch": 1.8567218045112783, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 35868 + }, + { + "epoch": 1.856781954887218, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 35869 + }, + { + "epoch": 1.8568421052631578, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 35870 + }, + { + "epoch": 1.8569022556390977, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 35871 + }, + { + "epoch": 1.8569624060150376, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 35872 + }, + { + "epoch": 1.8570225563909775, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 35873 + }, + { + "epoch": 1.8570827067669173, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 35874 + }, + { + "epoch": 1.8571428571428572, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 35875 + }, + { + "epoch": 1.8572030075187969, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 35876 + }, + { + "epoch": 1.857263157894737, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 35877 + }, + { + "epoch": 1.8573233082706766, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 35878 + }, + { + "epoch": 1.8573834586466167, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 35879 + }, + { + "epoch": 1.8574436090225563, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 35880 + }, + { + "epoch": 1.8575037593984962, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 35881 + }, + { + "epoch": 1.857563909774436, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 35882 + }, + { + "epoch": 1.857624060150376, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 35883 + }, + { + "epoch": 1.8576842105263158, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 35884 + }, + { + "epoch": 1.8577443609022555, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 35885 + }, + { + "epoch": 1.8578045112781956, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 35886 + }, + { + "epoch": 1.8578646616541352, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 35887 + }, + { + "epoch": 1.8579248120300753, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 35888 + }, + { + "epoch": 1.857984962406015, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 35889 + }, + { + "epoch": 1.8580451127819548, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 35890 + }, + { + "epoch": 1.8581052631578947, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 35891 + }, + { + "epoch": 1.8581654135338346, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 35892 + }, + { + "epoch": 1.8582255639097744, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 35893 + }, + { + "epoch": 1.8582857142857143, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 35894 + }, + { + "epoch": 1.8583458646616542, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 35895 + }, + { + "epoch": 1.8584060150375938, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 35896 + }, + { + "epoch": 1.858466165413534, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 35897 + }, + { + "epoch": 1.8585263157894736, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 35898 + }, + { + "epoch": 1.8585864661654137, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 35899 + }, + { + "epoch": 1.8586466165413533, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 35900 + }, + { + "epoch": 1.8587067669172932, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 35901 + }, + { + "epoch": 1.858766917293233, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 35902 + }, + { + "epoch": 1.858827067669173, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 35903 + }, + { + "epoch": 1.8588872180451128, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 35904 + }, + { + "epoch": 1.8589473684210527, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 35905 + }, + { + "epoch": 1.8590075187969926, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 35906 + }, + { + "epoch": 1.8590676691729322, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 35907 + }, + { + "epoch": 1.8591278195488723, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 35908 + }, + { + "epoch": 1.859187969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 35909 + }, + { + "epoch": 1.859248120300752, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 35910 + }, + { + "epoch": 1.8593082706766917, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 35911 + }, + { + "epoch": 1.8593684210526316, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 35912 + }, + { + "epoch": 1.8594285714285714, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 35913 + }, + { + "epoch": 1.8594887218045113, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 35914 + }, + { + "epoch": 1.8595488721804512, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 35915 + }, + { + "epoch": 1.8596090225563908, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 35916 + }, + { + "epoch": 1.859669172932331, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 35917 + }, + { + "epoch": 1.8597293233082706, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 35918 + }, + { + "epoch": 1.8597894736842107, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 35919 + }, + { + "epoch": 1.8598496240601503, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 35920 + }, + { + "epoch": 1.8599097744360902, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 35921 + }, + { + "epoch": 1.85996992481203, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 35922 + }, + { + "epoch": 1.86003007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 35923 + }, + { + "epoch": 1.8600902255639098, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 35924 + }, + { + "epoch": 1.8601503759398497, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 35925 + }, + { + "epoch": 1.8602105263157895, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 35926 + }, + { + "epoch": 1.8602706766917292, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 35927 + }, + { + "epoch": 1.8603308270676693, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 35928 + }, + { + "epoch": 1.860390977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 35929 + }, + { + "epoch": 1.860451127819549, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 35930 + }, + { + "epoch": 1.8605112781954887, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 35931 + }, + { + "epoch": 1.8605714285714285, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 35932 + }, + { + "epoch": 1.8606315789473684, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 35933 + }, + { + "epoch": 1.8606917293233083, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 35934 + }, + { + "epoch": 1.8607518796992482, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 35935 + }, + { + "epoch": 1.8608120300751878, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9529, + "step": 35936 + }, + { + "epoch": 1.860872180451128, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 35937 + }, + { + "epoch": 1.8609323308270675, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 35938 + }, + { + "epoch": 1.8609924812030076, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1234, + "step": 35939 + }, + { + "epoch": 1.8610526315789473, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 35940 + }, + { + "epoch": 1.8611127819548872, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 35941 + }, + { + "epoch": 1.861172932330827, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 35942 + }, + { + "epoch": 1.861233082706767, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 35943 + }, + { + "epoch": 1.8612932330827068, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 35944 + }, + { + "epoch": 1.8613533834586466, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 35945 + }, + { + "epoch": 1.8614135338345865, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 35946 + }, + { + "epoch": 1.8614736842105262, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 35947 + }, + { + "epoch": 1.8615338345864663, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 35948 + }, + { + "epoch": 1.861593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 35949 + }, + { + "epoch": 1.861654135338346, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 35950 + }, + { + "epoch": 1.8617142857142857, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 35951 + }, + { + "epoch": 1.8617744360902255, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 35952 + }, + { + "epoch": 1.8618345864661654, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 35953 + }, + { + "epoch": 1.8618947368421053, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 35954 + }, + { + "epoch": 1.8619548872180451, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 35955 + }, + { + "epoch": 1.862015037593985, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 35956 + }, + { + "epoch": 1.8620751879699249, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 35957 + }, + { + "epoch": 1.8621353383458645, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 35958 + }, + { + "epoch": 1.8621954887218046, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 35959 + }, + { + "epoch": 1.8622556390977443, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 35960 + }, + { + "epoch": 1.8623157894736844, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 35961 + }, + { + "epoch": 1.862375939849624, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 35962 + }, + { + "epoch": 1.862436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 35963 + }, + { + "epoch": 1.8624962406015038, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 35964 + }, + { + "epoch": 1.8625563909774436, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 35965 + }, + { + "epoch": 1.8626165413533835, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 35966 + }, + { + "epoch": 1.8626766917293232, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 35967 + }, + { + "epoch": 1.8627368421052632, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 35968 + }, + { + "epoch": 1.862796992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 35969 + }, + { + "epoch": 1.862857142857143, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 35970 + }, + { + "epoch": 1.8629172932330826, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 35971 + }, + { + "epoch": 1.8629774436090225, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 35972 + }, + { + "epoch": 1.8630375939849624, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 35973 + }, + { + "epoch": 1.8630977443609023, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 35974 + }, + { + "epoch": 1.8631578947368421, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 35975 + }, + { + "epoch": 1.863218045112782, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 35976 + }, + { + "epoch": 1.8632781954887219, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1283, + "step": 35977 + }, + { + "epoch": 1.8633383458646615, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 35978 + }, + { + "epoch": 1.8633984962406016, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 35979 + }, + { + "epoch": 1.8634586466165413, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 35980 + }, + { + "epoch": 1.8635187969924814, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 35981 + }, + { + "epoch": 1.863578947368421, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 35982 + }, + { + "epoch": 1.8636390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 35983 + }, + { + "epoch": 1.8636992481203007, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 35984 + }, + { + "epoch": 1.8637593984962406, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2094, + "step": 35985 + }, + { + "epoch": 1.8638195488721805, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 35986 + }, + { + "epoch": 1.8638796992481201, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 35987 + }, + { + "epoch": 1.8639398496240602, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 35988 + }, + { + "epoch": 1.8639999999999999, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 35989 + }, + { + "epoch": 1.86406015037594, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.2337, + "step": 35990 + }, + { + "epoch": 1.8641203007518796, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 35991 + }, + { + "epoch": 1.8641804511278197, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.931, + "step": 35992 + }, + { + "epoch": 1.8642406015037594, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 35993 + }, + { + "epoch": 1.8643007518796992, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 35994 + }, + { + "epoch": 1.864360902255639, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 35995 + }, + { + "epoch": 1.864421052631579, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 35996 + }, + { + "epoch": 1.8644812030075189, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 35997 + }, + { + "epoch": 1.8645413533834585, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 35998 + }, + { + "epoch": 1.8646015037593986, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 35999 + }, + { + "epoch": 1.8646616541353382, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 36000 + }, + { + "epoch": 1.8647218045112783, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 36001 + }, + { + "epoch": 1.864781954887218, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 36002 + }, + { + "epoch": 1.8648421052631579, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 36003 + }, + { + "epoch": 1.8649022556390977, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 36004 + }, + { + "epoch": 1.8649624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 36005 + }, + { + "epoch": 1.8650225563909775, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 36006 + }, + { + "epoch": 1.8650827067669173, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 36007 + }, + { + "epoch": 1.8651428571428572, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 36008 + }, + { + "epoch": 1.8652030075187969, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 36009 + }, + { + "epoch": 1.865263157894737, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 36010 + }, + { + "epoch": 1.8653233082706766, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 36011 + }, + { + "epoch": 1.8653834586466167, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 36012 + }, + { + "epoch": 1.8654436090225563, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 36013 + }, + { + "epoch": 1.8655037593984962, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 36014 + }, + { + "epoch": 1.865563909774436, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 36015 + }, + { + "epoch": 1.865624060150376, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 36016 + }, + { + "epoch": 1.8656842105263158, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 36017 + }, + { + "epoch": 1.8657443609022555, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 36018 + }, + { + "epoch": 1.8658045112781956, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 36019 + }, + { + "epoch": 1.8658646616541352, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 36020 + }, + { + "epoch": 1.8659248120300753, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 36021 + }, + { + "epoch": 1.865984962406015, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 36022 + }, + { + "epoch": 1.8660451127819548, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 36023 + }, + { + "epoch": 1.8661052631578947, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 36024 + }, + { + "epoch": 1.8661654135338346, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 36025 + }, + { + "epoch": 1.8662255639097745, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 36026 + }, + { + "epoch": 1.8662857142857143, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 36027 + }, + { + "epoch": 1.8663458646616542, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 36028 + }, + { + "epoch": 1.8664060150375938, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 36029 + }, + { + "epoch": 1.866466165413534, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 36030 + }, + { + "epoch": 1.8665263157894736, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 36031 + }, + { + "epoch": 1.8665864661654137, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 36032 + }, + { + "epoch": 1.8666466165413533, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 36033 + }, + { + "epoch": 1.8667067669172932, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 36034 + }, + { + "epoch": 1.866766917293233, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 36035 + }, + { + "epoch": 1.866827067669173, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 36036 + }, + { + "epoch": 1.8668872180451128, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9326, + "step": 36037 + }, + { + "epoch": 1.8669473684210525, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 36038 + }, + { + "epoch": 1.8670075187969926, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 36039 + }, + { + "epoch": 1.8670676691729322, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 36040 + }, + { + "epoch": 1.8671278195488723, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 36041 + }, + { + "epoch": 1.867187969924812, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 36042 + }, + { + "epoch": 1.867248120300752, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 36043 + }, + { + "epoch": 1.8673082706766917, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 36044 + }, + { + "epoch": 1.8673684210526316, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 36045 + }, + { + "epoch": 1.8674285714285714, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 36046 + }, + { + "epoch": 1.8674887218045113, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 36047 + }, + { + "epoch": 1.8675488721804512, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9286, + "step": 36048 + }, + { + "epoch": 1.8676090225563908, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 36049 + }, + { + "epoch": 1.867669172932331, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 36050 + }, + { + "epoch": 1.8677293233082706, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 36051 + }, + { + "epoch": 1.8677894736842107, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 36052 + }, + { + "epoch": 1.8678496240601503, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 36053 + }, + { + "epoch": 1.8679097744360902, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 36054 + }, + { + "epoch": 1.86796992481203, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9453, + "step": 36055 + }, + { + "epoch": 1.86803007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 36056 + }, + { + "epoch": 1.8680902255639098, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 36057 + }, + { + "epoch": 1.8681503759398497, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 36058 + }, + { + "epoch": 1.8682105263157895, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 36059 + }, + { + "epoch": 1.8682706766917292, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 36060 + }, + { + "epoch": 1.8683308270676693, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 36061 + }, + { + "epoch": 1.868390977443609, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 36062 + }, + { + "epoch": 1.868451127819549, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 36063 + }, + { + "epoch": 1.8685112781954887, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 36064 + }, + { + "epoch": 1.8685714285714285, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 36065 + }, + { + "epoch": 1.8686315789473684, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 36066 + }, + { + "epoch": 1.8686917293233083, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 36067 + }, + { + "epoch": 1.8687518796992482, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 36068 + }, + { + "epoch": 1.8688120300751878, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 36069 + }, + { + "epoch": 1.868872180451128, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 36070 + }, + { + "epoch": 1.8689323308270676, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 36071 + }, + { + "epoch": 1.8689924812030076, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 36072 + }, + { + "epoch": 1.8690526315789473, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 36073 + }, + { + "epoch": 1.8691127819548872, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 36074 + }, + { + "epoch": 1.869172932330827, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 36075 + }, + { + "epoch": 1.869233082706767, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 36076 + }, + { + "epoch": 1.8692932330827068, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 36077 + }, + { + "epoch": 1.8693533834586467, + "grad_norm": 1.578125, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 36078 + }, + { + "epoch": 1.8694135338345865, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 36079 + }, + { + "epoch": 1.8694736842105262, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 36080 + }, + { + "epoch": 1.8695338345864663, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 36081 + }, + { + "epoch": 1.869593984962406, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 36082 + }, + { + "epoch": 1.869654135338346, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 36083 + }, + { + "epoch": 1.8697142857142857, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 36084 + }, + { + "epoch": 1.8697744360902255, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 36085 + }, + { + "epoch": 1.8698345864661654, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 36086 + }, + { + "epoch": 1.8698947368421053, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 36087 + }, + { + "epoch": 1.8699548872180451, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 36088 + }, + { + "epoch": 1.8700150375939848, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 36089 + }, + { + "epoch": 1.870075187969925, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.2851, + "step": 36090 + }, + { + "epoch": 1.8701353383458645, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 36091 + }, + { + "epoch": 1.8701954887218046, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 36092 + }, + { + "epoch": 1.8702556390977443, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 36093 + }, + { + "epoch": 1.8703157894736844, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 36094 + }, + { + "epoch": 1.870375939849624, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 36095 + }, + { + "epoch": 1.870436090225564, + "grad_norm": 1.6015625, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 36096 + }, + { + "epoch": 1.8704962406015038, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 36097 + }, + { + "epoch": 1.8705563909774436, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 36098 + }, + { + "epoch": 1.8706165413533835, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 36099 + }, + { + "epoch": 1.8706766917293232, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 36100 + }, + { + "epoch": 1.8707368421052633, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 36101 + }, + { + "epoch": 1.870796992481203, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 36102 + }, + { + "epoch": 1.870857142857143, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 36103 + }, + { + "epoch": 1.8709172932330826, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 36104 + }, + { + "epoch": 1.8709774436090225, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 36105 + }, + { + "epoch": 1.8710375939849624, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 36106 + }, + { + "epoch": 1.8710977443609023, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 36107 + }, + { + "epoch": 1.8711578947368421, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 36108 + }, + { + "epoch": 1.871218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 36109 + }, + { + "epoch": 1.8712781954887219, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 36110 + }, + { + "epoch": 1.8713383458646615, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 36111 + }, + { + "epoch": 1.8713984962406016, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 36112 + }, + { + "epoch": 1.8714586466165413, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 36113 + }, + { + "epoch": 1.8715187969924814, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 36114 + }, + { + "epoch": 1.871578947368421, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 36115 + }, + { + "epoch": 1.8716390977443609, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 36116 + }, + { + "epoch": 1.8716992481203008, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 36117 + }, + { + "epoch": 1.8717593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 36118 + }, + { + "epoch": 1.8718195488721805, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 36119 + }, + { + "epoch": 1.8718796992481201, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.3347, + "step": 36120 + }, + { + "epoch": 1.8719398496240602, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 36121 + }, + { + "epoch": 1.8719999999999999, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 36122 + }, + { + "epoch": 1.87206015037594, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 36123 + }, + { + "epoch": 1.8721203007518796, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 36124 + }, + { + "epoch": 1.8721804511278195, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 36125 + }, + { + "epoch": 1.8722406015037594, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 36126 + }, + { + "epoch": 1.8723007518796992, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 36127 + }, + { + "epoch": 1.8723609022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 36128 + }, + { + "epoch": 1.872421052631579, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 36129 + }, + { + "epoch": 1.8724812030075189, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 36130 + }, + { + "epoch": 1.8725413533834585, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1914, + "step": 36131 + }, + { + "epoch": 1.8726015037593986, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 36132 + }, + { + "epoch": 1.8726616541353382, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 36133 + }, + { + "epoch": 1.8727218045112783, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 36134 + }, + { + "epoch": 1.872781954887218, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 36135 + }, + { + "epoch": 1.8728421052631579, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 36136 + }, + { + "epoch": 1.8729022556390977, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9635, + "step": 36137 + }, + { + "epoch": 1.8729624060150376, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 36138 + }, + { + "epoch": 1.8730225563909775, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 36139 + }, + { + "epoch": 1.8730827067669171, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 36140 + }, + { + "epoch": 1.8731428571428572, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 36141 + }, + { + "epoch": 1.8732030075187969, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 36142 + }, + { + "epoch": 1.873263157894737, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 36143 + }, + { + "epoch": 1.8733233082706766, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 36144 + }, + { + "epoch": 1.8733834586466167, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 36145 + }, + { + "epoch": 1.8734436090225564, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9411, + "step": 36146 + }, + { + "epoch": 1.8735037593984962, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 36147 + }, + { + "epoch": 1.873563909774436, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 36148 + }, + { + "epoch": 1.873624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 36149 + }, + { + "epoch": 1.8736842105263158, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 36150 + }, + { + "epoch": 1.8737443609022555, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 36151 + }, + { + "epoch": 1.8738045112781956, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 36152 + }, + { + "epoch": 1.8738646616541352, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 36153 + }, + { + "epoch": 1.8739248120300753, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 36154 + }, + { + "epoch": 1.873984962406015, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 36155 + }, + { + "epoch": 1.8740451127819548, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 36156 + }, + { + "epoch": 1.8741052631578947, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 36157 + }, + { + "epoch": 1.8741654135338346, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 36158 + }, + { + "epoch": 1.8742255639097745, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 36159 + }, + { + "epoch": 1.8742857142857143, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 36160 + }, + { + "epoch": 1.8743458646616542, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 36161 + }, + { + "epoch": 1.8744060150375939, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 36162 + }, + { + "epoch": 1.874466165413534, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 36163 + }, + { + "epoch": 1.8745263157894736, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 36164 + }, + { + "epoch": 1.8745864661654137, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 36165 + }, + { + "epoch": 1.8746466165413533, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 36166 + }, + { + "epoch": 1.8747067669172932, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 36167 + }, + { + "epoch": 1.874766917293233, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 36168 + }, + { + "epoch": 1.874827067669173, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 36169 + }, + { + "epoch": 1.8748872180451128, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 36170 + }, + { + "epoch": 1.8749473684210525, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 36171 + }, + { + "epoch": 1.8750075187969926, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 36172 + }, + { + "epoch": 1.8750676691729322, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 36173 + }, + { + "epoch": 1.8751278195488723, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 36174 + }, + { + "epoch": 1.875187969924812, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 36175 + }, + { + "epoch": 1.8752481203007518, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 36176 + }, + { + "epoch": 1.8753082706766917, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.136, + "step": 36177 + }, + { + "epoch": 1.8753684210526316, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 36178 + }, + { + "epoch": 1.8754285714285714, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 36179 + }, + { + "epoch": 1.8754887218045113, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 36180 + }, + { + "epoch": 1.8755488721804512, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 36181 + }, + { + "epoch": 1.8756090225563908, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 36182 + }, + { + "epoch": 1.875669172932331, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 36183 + }, + { + "epoch": 1.8757293233082706, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 36184 + }, + { + "epoch": 1.8757894736842107, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 36185 + }, + { + "epoch": 1.8758496240601503, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 36186 + }, + { + "epoch": 1.8759097744360902, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 36187 + }, + { + "epoch": 1.87596992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 36188 + }, + { + "epoch": 1.87603007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 36189 + }, + { + "epoch": 1.8760902255639098, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 36190 + }, + { + "epoch": 1.8761503759398497, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 36191 + }, + { + "epoch": 1.8762105263157896, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 36192 + }, + { + "epoch": 1.8762706766917292, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 36193 + }, + { + "epoch": 1.8763308270676693, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 36194 + }, + { + "epoch": 1.876390977443609, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 36195 + }, + { + "epoch": 1.876451127819549, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 36196 + }, + { + "epoch": 1.8765112781954887, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 36197 + }, + { + "epoch": 1.8765714285714286, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 36198 + }, + { + "epoch": 1.8766315789473684, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 36199 + }, + { + "epoch": 1.8766917293233083, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 36200 + }, + { + "epoch": 1.8767518796992482, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9014, + "step": 36201 + }, + { + "epoch": 1.8768120300751878, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 36202 + }, + { + "epoch": 1.876872180451128, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 36203 + }, + { + "epoch": 1.8769323308270676, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 36204 + }, + { + "epoch": 1.8769924812030077, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 36205 + }, + { + "epoch": 1.8770526315789473, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 36206 + }, + { + "epoch": 1.8771127819548872, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 36207 + }, + { + "epoch": 1.877172932330827, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 36208 + }, + { + "epoch": 1.877233082706767, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 36209 + }, + { + "epoch": 1.8772932330827068, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 36210 + }, + { + "epoch": 1.8773533834586467, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 36211 + }, + { + "epoch": 1.8774135338345865, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 36212 + }, + { + "epoch": 1.8774736842105262, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 36213 + }, + { + "epoch": 1.8775338345864663, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 36214 + }, + { + "epoch": 1.877593984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 36215 + }, + { + "epoch": 1.877654135338346, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 36216 + }, + { + "epoch": 1.8777142857142857, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 36217 + }, + { + "epoch": 1.8777744360902255, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 36218 + }, + { + "epoch": 1.8778345864661654, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 36219 + }, + { + "epoch": 1.8778947368421053, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 36220 + }, + { + "epoch": 1.8779548872180452, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 36221 + }, + { + "epoch": 1.8780150375939848, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 36222 + }, + { + "epoch": 1.878075187969925, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 36223 + }, + { + "epoch": 1.8781353383458645, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 36224 + }, + { + "epoch": 1.8781954887218046, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 36225 + }, + { + "epoch": 1.8782556390977443, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 36226 + }, + { + "epoch": 1.8783157894736842, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 36227 + }, + { + "epoch": 1.878375939849624, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 36228 + }, + { + "epoch": 1.878436090225564, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 36229 + }, + { + "epoch": 1.8784962406015038, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 36230 + }, + { + "epoch": 1.8785563909774436, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 36231 + }, + { + "epoch": 1.8786165413533835, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 36232 + }, + { + "epoch": 1.8786766917293232, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 36233 + }, + { + "epoch": 1.8787368421052633, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 36234 + }, + { + "epoch": 1.878796992481203, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 36235 + }, + { + "epoch": 1.878857142857143, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 36236 + }, + { + "epoch": 1.8789172932330827, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 36237 + }, + { + "epoch": 1.8789774436090225, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 36238 + }, + { + "epoch": 1.8790375939849624, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 36239 + }, + { + "epoch": 1.8790977443609023, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 36240 + }, + { + "epoch": 1.8791578947368421, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 36241 + }, + { + "epoch": 1.879218045112782, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 36242 + }, + { + "epoch": 1.8792781954887219, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 36243 + }, + { + "epoch": 1.8793383458646615, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 36244 + }, + { + "epoch": 1.8793984962406016, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2637, + "step": 36245 + }, + { + "epoch": 1.8794586466165413, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 36246 + }, + { + "epoch": 1.8795187969924814, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 36247 + }, + { + "epoch": 1.879578947368421, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 36248 + }, + { + "epoch": 1.8796390977443609, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 36249 + }, + { + "epoch": 1.8796992481203008, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 36250 + }, + { + "epoch": 1.8797593984962406, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 36251 + }, + { + "epoch": 1.8798195488721805, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 36252 + }, + { + "epoch": 1.8798796992481202, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 36253 + }, + { + "epoch": 1.8799398496240602, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 36254 + }, + { + "epoch": 1.88, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 36255 + }, + { + "epoch": 1.88006015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 36256 + }, + { + "epoch": 1.8801203007518796, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 36257 + }, + { + "epoch": 1.8801804511278195, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 36258 + }, + { + "epoch": 1.8802406015037594, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 36259 + }, + { + "epoch": 1.8803007518796993, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 36260 + }, + { + "epoch": 1.8803609022556391, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 36261 + }, + { + "epoch": 1.880421052631579, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 36262 + }, + { + "epoch": 1.8804812030075189, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 36263 + }, + { + "epoch": 1.8805413533834585, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 36264 + }, + { + "epoch": 1.8806015037593986, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 36265 + }, + { + "epoch": 1.8806616541353383, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 36266 + }, + { + "epoch": 1.8807218045112783, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 36267 + }, + { + "epoch": 1.880781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 36268 + }, + { + "epoch": 1.8808421052631579, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 36269 + }, + { + "epoch": 1.8809022556390977, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 36270 + }, + { + "epoch": 1.8809624060150376, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 36271 + }, + { + "epoch": 1.8810225563909775, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 36272 + }, + { + "epoch": 1.8810827067669171, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 36273 + }, + { + "epoch": 1.8811428571428572, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 36274 + }, + { + "epoch": 1.8812030075187969, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 36275 + }, + { + "epoch": 1.881263157894737, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 36276 + }, + { + "epoch": 1.8813233082706766, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 36277 + }, + { + "epoch": 1.8813834586466165, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 36278 + }, + { + "epoch": 1.8814436090225564, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 36279 + }, + { + "epoch": 1.8815037593984962, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 36280 + }, + { + "epoch": 1.881563909774436, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9266, + "step": 36281 + }, + { + "epoch": 1.881624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 36282 + }, + { + "epoch": 1.8816842105263158, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 36283 + }, + { + "epoch": 1.8817443609022555, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 36284 + }, + { + "epoch": 1.8818045112781956, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 36285 + }, + { + "epoch": 1.8818646616541352, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 36286 + }, + { + "epoch": 1.8819248120300753, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 36287 + }, + { + "epoch": 1.881984962406015, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 36288 + }, + { + "epoch": 1.8820451127819549, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9248, + "step": 36289 + }, + { + "epoch": 1.8821052631578947, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 36290 + }, + { + "epoch": 1.8821654135338346, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 36291 + }, + { + "epoch": 1.8822255639097745, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 36292 + }, + { + "epoch": 1.8822857142857143, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 36293 + }, + { + "epoch": 1.8823458646616542, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 36294 + }, + { + "epoch": 1.8824060150375939, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 36295 + }, + { + "epoch": 1.882466165413534, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 36296 + }, + { + "epoch": 1.8825263157894736, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 36297 + }, + { + "epoch": 1.8825864661654137, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 36298 + }, + { + "epoch": 1.8826466165413533, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 36299 + }, + { + "epoch": 1.8827067669172932, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 36300 + }, + { + "epoch": 1.882766917293233, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 36301 + }, + { + "epoch": 1.882827067669173, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9516, + "step": 36302 + }, + { + "epoch": 1.8828872180451128, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 36303 + }, + { + "epoch": 1.8829473684210525, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 36304 + }, + { + "epoch": 1.8830075187969926, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 36305 + }, + { + "epoch": 1.8830676691729322, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 36306 + }, + { + "epoch": 1.8831278195488723, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 36307 + }, + { + "epoch": 1.883187969924812, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 36308 + }, + { + "epoch": 1.8832481203007518, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 36309 + }, + { + "epoch": 1.8833082706766917, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9348, + "step": 36310 + }, + { + "epoch": 1.8833684210526316, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 36311 + }, + { + "epoch": 1.8834285714285715, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 36312 + }, + { + "epoch": 1.8834887218045113, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 36313 + }, + { + "epoch": 1.8835488721804512, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 36314 + }, + { + "epoch": 1.8836090225563908, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 36315 + }, + { + "epoch": 1.883669172932331, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 36316 + }, + { + "epoch": 1.8837293233082706, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 36317 + }, + { + "epoch": 1.8837894736842107, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1703, + "step": 36318 + }, + { + "epoch": 1.8838496240601503, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 36319 + }, + { + "epoch": 1.8839097744360902, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1472, + "step": 36320 + }, + { + "epoch": 1.88396992481203, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 36321 + }, + { + "epoch": 1.88403007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 36322 + }, + { + "epoch": 1.8840902255639098, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 36323 + }, + { + "epoch": 1.8841503759398495, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 36324 + }, + { + "epoch": 1.8842105263157896, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 36325 + }, + { + "epoch": 1.8842706766917292, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 36326 + }, + { + "epoch": 1.8843308270676693, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1941, + "step": 36327 + }, + { + "epoch": 1.884390977443609, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 36328 + }, + { + "epoch": 1.8844511278195488, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 36329 + }, + { + "epoch": 1.8845112781954887, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 36330 + }, + { + "epoch": 1.8845714285714286, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 36331 + }, + { + "epoch": 1.8846315789473684, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 36332 + }, + { + "epoch": 1.8846917293233083, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 36333 + }, + { + "epoch": 1.8847518796992482, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 36334 + }, + { + "epoch": 1.8848120300751878, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 36335 + }, + { + "epoch": 1.884872180451128, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 36336 + }, + { + "epoch": 1.8849323308270676, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 36337 + }, + { + "epoch": 1.8849924812030077, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 36338 + }, + { + "epoch": 1.8850526315789473, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 36339 + }, + { + "epoch": 1.8851127819548872, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 36340 + }, + { + "epoch": 1.885172932330827, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 36341 + }, + { + "epoch": 1.885233082706767, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 36342 + }, + { + "epoch": 1.8852932330827068, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 36343 + }, + { + "epoch": 1.8853533834586467, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 36344 + }, + { + "epoch": 1.8854135338345865, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 36345 + }, + { + "epoch": 1.8854736842105262, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 36346 + }, + { + "epoch": 1.8855338345864663, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 36347 + }, + { + "epoch": 1.885593984962406, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 36348 + }, + { + "epoch": 1.885654135338346, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 36349 + }, + { + "epoch": 1.8857142857142857, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 36350 + }, + { + "epoch": 1.8857744360902255, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 36351 + }, + { + "epoch": 1.8858345864661654, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 36352 + }, + { + "epoch": 1.8858947368421053, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 36353 + }, + { + "epoch": 1.8859548872180452, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 36354 + }, + { + "epoch": 1.8860150375939848, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 36355 + }, + { + "epoch": 1.886075187969925, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 36356 + }, + { + "epoch": 1.8861353383458646, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 36357 + }, + { + "epoch": 1.8861954887218046, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 36358 + }, + { + "epoch": 1.8862556390977443, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 36359 + }, + { + "epoch": 1.8863157894736842, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1589, + "step": 36360 + }, + { + "epoch": 1.886375939849624, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 36361 + }, + { + "epoch": 1.886436090225564, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 36362 + }, + { + "epoch": 1.8864962406015038, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 36363 + }, + { + "epoch": 1.8865563909774437, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 36364 + }, + { + "epoch": 1.8866165413533835, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 36365 + }, + { + "epoch": 1.8866766917293232, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 36366 + }, + { + "epoch": 1.8867368421052633, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 36367 + }, + { + "epoch": 1.886796992481203, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.8662, + "step": 36368 + }, + { + "epoch": 1.886857142857143, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 36369 + }, + { + "epoch": 1.8869172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 36370 + }, + { + "epoch": 1.8869774436090225, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 36371 + }, + { + "epoch": 1.8870375939849624, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 36372 + }, + { + "epoch": 1.8870977443609023, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 36373 + }, + { + "epoch": 1.8871578947368421, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 36374 + }, + { + "epoch": 1.8872180451127818, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 36375 + }, + { + "epoch": 1.8872781954887219, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 36376 + }, + { + "epoch": 1.8873383458646615, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 36377 + }, + { + "epoch": 1.8873984962406016, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 36378 + }, + { + "epoch": 1.8874586466165413, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 36379 + }, + { + "epoch": 1.8875187969924812, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 36380 + }, + { + "epoch": 1.887578947368421, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 36381 + }, + { + "epoch": 1.887639097744361, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 36382 + }, + { + "epoch": 1.8876992481203008, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 36383 + }, + { + "epoch": 1.8877593984962406, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1975, + "step": 36384 + }, + { + "epoch": 1.8878195488721805, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 36385 + }, + { + "epoch": 1.8878796992481202, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 36386 + }, + { + "epoch": 1.8879398496240603, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 36387 + }, + { + "epoch": 1.888, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2518, + "step": 36388 + }, + { + "epoch": 1.88806015037594, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 36389 + }, + { + "epoch": 1.8881203007518796, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 36390 + }, + { + "epoch": 1.8881804511278195, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 36391 + }, + { + "epoch": 1.8882406015037594, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 36392 + }, + { + "epoch": 1.8883007518796993, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 36393 + }, + { + "epoch": 1.8883609022556391, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 36394 + }, + { + "epoch": 1.888421052631579, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 36395 + }, + { + "epoch": 1.8884812030075189, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 36396 + }, + { + "epoch": 1.8885413533834585, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 36397 + }, + { + "epoch": 1.8886015037593986, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 36398 + }, + { + "epoch": 1.8886616541353383, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 36399 + }, + { + "epoch": 1.8887218045112784, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 36400 + }, + { + "epoch": 1.888781954887218, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 36401 + }, + { + "epoch": 1.8888421052631579, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 36402 + }, + { + "epoch": 1.8889022556390977, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 36403 + }, + { + "epoch": 1.8889624060150376, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 36404 + }, + { + "epoch": 1.8890225563909775, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 36405 + }, + { + "epoch": 1.8890827067669171, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 36406 + }, + { + "epoch": 1.8891428571428572, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 36407 + }, + { + "epoch": 1.8892030075187969, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 36408 + }, + { + "epoch": 1.889263157894737, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 36409 + }, + { + "epoch": 1.8893233082706766, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 36410 + }, + { + "epoch": 1.8893834586466165, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 36411 + }, + { + "epoch": 1.8894436090225564, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 36412 + }, + { + "epoch": 1.8895037593984962, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 36413 + }, + { + "epoch": 1.8895639097744361, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 36414 + }, + { + "epoch": 1.889624060150376, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 36415 + }, + { + "epoch": 1.8896842105263159, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 36416 + }, + { + "epoch": 1.8897443609022555, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 36417 + }, + { + "epoch": 1.8898045112781956, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 36418 + }, + { + "epoch": 1.8898646616541352, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 36419 + }, + { + "epoch": 1.8899248120300753, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 36420 + }, + { + "epoch": 1.889984962406015, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 36421 + }, + { + "epoch": 1.8900451127819549, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 36422 + }, + { + "epoch": 1.8901052631578947, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 36423 + }, + { + "epoch": 1.8901654135338346, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 36424 + }, + { + "epoch": 1.8902255639097745, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 36425 + }, + { + "epoch": 1.8902857142857141, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 36426 + }, + { + "epoch": 1.8903458646616542, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1875, + "step": 36427 + }, + { + "epoch": 1.8904060150375939, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 36428 + }, + { + "epoch": 1.890466165413534, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 36429 + }, + { + "epoch": 1.8905263157894736, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 36430 + }, + { + "epoch": 1.8905864661654135, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 36431 + }, + { + "epoch": 1.8906466165413534, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 36432 + }, + { + "epoch": 1.8907067669172932, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 36433 + }, + { + "epoch": 1.890766917293233, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 36434 + }, + { + "epoch": 1.890827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 36435 + }, + { + "epoch": 1.8908872180451128, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 36436 + }, + { + "epoch": 1.8909473684210525, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 36437 + }, + { + "epoch": 1.8910075187969926, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 36438 + }, + { + "epoch": 1.8910676691729322, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 36439 + }, + { + "epoch": 1.8911278195488723, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 36440 + }, + { + "epoch": 1.891187969924812, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 36441 + }, + { + "epoch": 1.8912481203007518, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 36442 + }, + { + "epoch": 1.8913082706766917, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 36443 + }, + { + "epoch": 1.8913684210526316, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 36444 + }, + { + "epoch": 1.8914285714285715, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 36445 + }, + { + "epoch": 1.8914887218045113, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 36446 + }, + { + "epoch": 1.8915488721804512, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 36447 + }, + { + "epoch": 1.8916090225563909, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 36448 + }, + { + "epoch": 1.891669172932331, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 36449 + }, + { + "epoch": 1.8917293233082706, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 36450 + }, + { + "epoch": 1.8917894736842107, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 36451 + }, + { + "epoch": 1.8918496240601503, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 36452 + }, + { + "epoch": 1.8919097744360902, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 36453 + }, + { + "epoch": 1.89196992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 36454 + }, + { + "epoch": 1.89203007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 36455 + }, + { + "epoch": 1.8920902255639098, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 36456 + }, + { + "epoch": 1.8921503759398495, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 36457 + }, + { + "epoch": 1.8922105263157896, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 36458 + }, + { + "epoch": 1.8922706766917292, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 36459 + }, + { + "epoch": 1.8923308270676693, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 36460 + }, + { + "epoch": 1.892390977443609, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 36461 + }, + { + "epoch": 1.8924511278195488, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 36462 + }, + { + "epoch": 1.8925112781954887, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 36463 + }, + { + "epoch": 1.8925714285714286, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 36464 + }, + { + "epoch": 1.8926315789473684, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 36465 + }, + { + "epoch": 1.8926917293233083, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 36466 + }, + { + "epoch": 1.8927518796992482, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 36467 + }, + { + "epoch": 1.8928120300751878, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 36468 + }, + { + "epoch": 1.892872180451128, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 36469 + }, + { + "epoch": 1.8929323308270676, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 36470 + }, + { + "epoch": 1.8929924812030077, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 36471 + }, + { + "epoch": 1.8930526315789473, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 36472 + }, + { + "epoch": 1.8931127819548872, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 36473 + }, + { + "epoch": 1.893172932330827, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 36474 + }, + { + "epoch": 1.893233082706767, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 36475 + }, + { + "epoch": 1.8932932330827068, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 36476 + }, + { + "epoch": 1.8933533834586465, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 36477 + }, + { + "epoch": 1.8934135338345865, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 36478 + }, + { + "epoch": 1.8934736842105262, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 36479 + }, + { + "epoch": 1.8935338345864663, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 36480 + }, + { + "epoch": 1.893593984962406, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 36481 + }, + { + "epoch": 1.8936541353383458, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 36482 + }, + { + "epoch": 1.8937142857142857, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 36483 + }, + { + "epoch": 1.8937744360902256, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 36484 + }, + { + "epoch": 1.8938345864661654, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 36485 + }, + { + "epoch": 1.8938947368421053, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 36486 + }, + { + "epoch": 1.8939548872180452, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 36487 + }, + { + "epoch": 1.8940150375939848, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 36488 + }, + { + "epoch": 1.894075187969925, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 36489 + }, + { + "epoch": 1.8941353383458646, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 36490 + }, + { + "epoch": 1.8941954887218047, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 36491 + }, + { + "epoch": 1.8942556390977443, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 36492 + }, + { + "epoch": 1.8943157894736842, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 36493 + }, + { + "epoch": 1.894375939849624, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 36494 + }, + { + "epoch": 1.894436090225564, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 36495 + }, + { + "epoch": 1.8944962406015038, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 36496 + }, + { + "epoch": 1.8945563909774437, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 36497 + }, + { + "epoch": 1.8946165413533835, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 36498 + }, + { + "epoch": 1.8946766917293232, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 36499 + }, + { + "epoch": 1.8947368421052633, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 36500 + }, + { + "epoch": 1.894796992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 36501 + }, + { + "epoch": 1.894857142857143, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 36502 + }, + { + "epoch": 1.8949172932330827, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 36503 + }, + { + "epoch": 1.8949774436090225, + "grad_norm": 0.46875, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 36504 + }, + { + "epoch": 1.8950375939849624, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 36505 + }, + { + "epoch": 1.8950977443609023, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 36506 + }, + { + "epoch": 1.8951578947368422, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1893, + "step": 36507 + }, + { + "epoch": 1.8952180451127818, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 36508 + }, + { + "epoch": 1.895278195488722, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 36509 + }, + { + "epoch": 1.8953383458646615, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 36510 + }, + { + "epoch": 1.8953984962406016, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 36511 + }, + { + "epoch": 1.8954586466165413, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 36512 + }, + { + "epoch": 1.8955187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 36513 + }, + { + "epoch": 1.895578947368421, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 36514 + }, + { + "epoch": 1.895639097744361, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 36515 + }, + { + "epoch": 1.8956992481203008, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 36516 + }, + { + "epoch": 1.8957593984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 36517 + }, + { + "epoch": 1.8958195488721805, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 36518 + }, + { + "epoch": 1.8958796992481202, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 36519 + }, + { + "epoch": 1.8959398496240603, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 36520 + }, + { + "epoch": 1.896, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 36521 + }, + { + "epoch": 1.89606015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 36522 + }, + { + "epoch": 1.8961203007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 36523 + }, + { + "epoch": 1.8961804511278195, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 36524 + }, + { + "epoch": 1.8962406015037594, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 36525 + }, + { + "epoch": 1.8963007518796993, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 36526 + }, + { + "epoch": 1.8963609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 36527 + }, + { + "epoch": 1.8964210526315788, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 36528 + }, + { + "epoch": 1.8964812030075189, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 36529 + }, + { + "epoch": 1.8965413533834585, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 36530 + }, + { + "epoch": 1.8966015037593986, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 36531 + }, + { + "epoch": 1.8966616541353383, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 36532 + }, + { + "epoch": 1.8967218045112781, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 36533 + }, + { + "epoch": 1.896781954887218, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 36534 + }, + { + "epoch": 1.8968421052631579, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 36535 + }, + { + "epoch": 1.8969022556390978, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 36536 + }, + { + "epoch": 1.8969624060150376, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 36537 + }, + { + "epoch": 1.8970225563909775, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9315, + "step": 36538 + }, + { + "epoch": 1.8970827067669171, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 36539 + }, + { + "epoch": 1.8971428571428572, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 36540 + }, + { + "epoch": 1.897203007518797, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 36541 + }, + { + "epoch": 1.897263157894737, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 36542 + }, + { + "epoch": 1.8973233082706766, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 36543 + }, + { + "epoch": 1.8973834586466165, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 36544 + }, + { + "epoch": 1.8974436090225564, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 36545 + }, + { + "epoch": 1.8975037593984962, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 36546 + }, + { + "epoch": 1.8975639097744361, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 36547 + }, + { + "epoch": 1.897624060150376, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 36548 + }, + { + "epoch": 1.8976842105263159, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 36549 + }, + { + "epoch": 1.8977443609022555, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 36550 + }, + { + "epoch": 1.8978045112781956, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 36551 + }, + { + "epoch": 1.8978646616541353, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 36552 + }, + { + "epoch": 1.8979248120300753, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 36553 + }, + { + "epoch": 1.897984962406015, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 36554 + }, + { + "epoch": 1.8980451127819549, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 36555 + }, + { + "epoch": 1.8981052631578947, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 36556 + }, + { + "epoch": 1.8981654135338346, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 36557 + }, + { + "epoch": 1.8982255639097745, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 36558 + }, + { + "epoch": 1.8982857142857141, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 36559 + }, + { + "epoch": 1.8983458646616542, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 36560 + }, + { + "epoch": 1.8984060150375939, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 36561 + }, + { + "epoch": 1.898466165413534, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 36562 + }, + { + "epoch": 1.8985263157894736, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 36563 + }, + { + "epoch": 1.8985864661654135, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 36564 + }, + { + "epoch": 1.8986466165413534, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 36565 + }, + { + "epoch": 1.8987067669172932, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 36566 + }, + { + "epoch": 1.898766917293233, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 36567 + }, + { + "epoch": 1.898827067669173, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 36568 + }, + { + "epoch": 1.8988872180451128, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 36569 + }, + { + "epoch": 1.8989473684210525, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1719, + "step": 36570 + }, + { + "epoch": 1.8990075187969926, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 36571 + }, + { + "epoch": 1.8990676691729322, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 36572 + }, + { + "epoch": 1.8991278195488723, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 36573 + }, + { + "epoch": 1.899187969924812, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 36574 + }, + { + "epoch": 1.8992481203007519, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 36575 + }, + { + "epoch": 1.8993082706766917, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 36576 + }, + { + "epoch": 1.8993684210526316, + "grad_norm": 1.515625, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 36577 + }, + { + "epoch": 1.8994285714285715, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 36578 + }, + { + "epoch": 1.8994887218045111, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 36579 + }, + { + "epoch": 1.8995488721804512, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 36580 + }, + { + "epoch": 1.8996090225563909, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9395, + "step": 36581 + }, + { + "epoch": 1.899669172932331, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 36582 + }, + { + "epoch": 1.8997293233082706, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 36583 + }, + { + "epoch": 1.8997894736842105, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 36584 + }, + { + "epoch": 1.8998496240601503, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 36585 + }, + { + "epoch": 1.8999097744360902, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 36586 + }, + { + "epoch": 1.89996992481203, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 36587 + }, + { + "epoch": 1.90003007518797, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 36588 + }, + { + "epoch": 1.9000902255639098, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 36589 + }, + { + "epoch": 1.9001503759398495, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 36590 + }, + { + "epoch": 1.9002105263157896, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 36591 + }, + { + "epoch": 1.9002706766917292, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 36592 + }, + { + "epoch": 1.9003308270676693, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 36593 + }, + { + "epoch": 1.900390977443609, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 36594 + }, + { + "epoch": 1.9004511278195488, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 36595 + }, + { + "epoch": 1.9005112781954887, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 36596 + }, + { + "epoch": 1.9005714285714286, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 36597 + }, + { + "epoch": 1.9006315789473684, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 36598 + }, + { + "epoch": 1.9006917293233083, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 36599 + }, + { + "epoch": 1.9007518796992482, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 36600 + }, + { + "epoch": 1.9008120300751878, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 36601 + }, + { + "epoch": 1.900872180451128, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 36602 + }, + { + "epoch": 1.9009323308270676, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 36603 + }, + { + "epoch": 1.9009924812030077, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 36604 + }, + { + "epoch": 1.9010526315789473, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 36605 + }, + { + "epoch": 1.9011127819548872, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 36606 + }, + { + "epoch": 1.901172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 36607 + }, + { + "epoch": 1.901233082706767, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 36608 + }, + { + "epoch": 1.9012932330827068, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1496, + "step": 36609 + }, + { + "epoch": 1.9013533834586465, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9417, + "step": 36610 + }, + { + "epoch": 1.9014135338345866, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 36611 + }, + { + "epoch": 1.9014736842105262, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 36612 + }, + { + "epoch": 1.9015338345864663, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 36613 + }, + { + "epoch": 1.901593984962406, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 36614 + }, + { + "epoch": 1.9016541353383458, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 36615 + }, + { + "epoch": 1.9017142857142857, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 36616 + }, + { + "epoch": 1.9017744360902256, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 36617 + }, + { + "epoch": 1.9018345864661654, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 36618 + }, + { + "epoch": 1.9018947368421053, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 36619 + }, + { + "epoch": 1.9019548872180452, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 36620 + }, + { + "epoch": 1.9020150375939848, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 36621 + }, + { + "epoch": 1.902075187969925, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 36622 + }, + { + "epoch": 1.9021353383458646, + "grad_norm": 1.390625, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 36623 + }, + { + "epoch": 1.9021954887218047, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 36624 + }, + { + "epoch": 1.9022556390977443, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 36625 + }, + { + "epoch": 1.9023157894736842, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 36626 + }, + { + "epoch": 1.902375939849624, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 36627 + }, + { + "epoch": 1.902436090225564, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 36628 + }, + { + "epoch": 1.9024962406015038, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 36629 + }, + { + "epoch": 1.9025563909774434, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 36630 + }, + { + "epoch": 1.9026165413533835, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 0.8605, + "step": 36631 + }, + { + "epoch": 1.9026766917293232, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 36632 + }, + { + "epoch": 1.9027368421052633, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 36633 + }, + { + "epoch": 1.902796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 36634 + }, + { + "epoch": 1.9028571428571428, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 36635 + }, + { + "epoch": 1.9029172932330827, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 36636 + }, + { + "epoch": 1.9029774436090225, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 36637 + }, + { + "epoch": 1.9030375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 36638 + }, + { + "epoch": 1.9030977443609023, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 36639 + }, + { + "epoch": 1.9031578947368422, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 36640 + }, + { + "epoch": 1.9032180451127818, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 36641 + }, + { + "epoch": 1.903278195488722, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 36642 + }, + { + "epoch": 1.9033383458646616, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 36643 + }, + { + "epoch": 1.9033984962406016, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 36644 + }, + { + "epoch": 1.9034586466165413, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.8732, + "step": 36645 + }, + { + "epoch": 1.9035187969924812, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 36646 + }, + { + "epoch": 1.903578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 36647 + }, + { + "epoch": 1.903639097744361, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 36648 + }, + { + "epoch": 1.9036992481203008, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 36649 + }, + { + "epoch": 1.9037593984962407, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 36650 + }, + { + "epoch": 1.9038195488721805, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 36651 + }, + { + "epoch": 1.9038796992481202, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 36652 + }, + { + "epoch": 1.9039398496240603, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 36653 + }, + { + "epoch": 1.904, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 36654 + }, + { + "epoch": 1.90406015037594, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 36655 + }, + { + "epoch": 1.9041203007518797, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 36656 + }, + { + "epoch": 1.9041804511278195, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 36657 + }, + { + "epoch": 1.9042406015037594, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 36658 + }, + { + "epoch": 1.9043007518796993, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 36659 + }, + { + "epoch": 1.9043609022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 36660 + }, + { + "epoch": 1.9044210526315788, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 36661 + }, + { + "epoch": 1.9044812030075189, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 36662 + }, + { + "epoch": 1.9045413533834585, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 36663 + }, + { + "epoch": 1.9046015037593986, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1795, + "step": 36664 + }, + { + "epoch": 1.9046616541353383, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.146, + "step": 36665 + }, + { + "epoch": 1.9047218045112781, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 36666 + }, + { + "epoch": 1.904781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 36667 + }, + { + "epoch": 1.904842105263158, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 36668 + }, + { + "epoch": 1.9049022556390978, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 36669 + }, + { + "epoch": 1.9049624060150376, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 36670 + }, + { + "epoch": 1.9050225563909775, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 36671 + }, + { + "epoch": 1.9050827067669172, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 36672 + }, + { + "epoch": 1.9051428571428572, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 36673 + }, + { + "epoch": 1.905203007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.144, + "step": 36674 + }, + { + "epoch": 1.905263157894737, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 36675 + }, + { + "epoch": 1.9053233082706766, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.265, + "step": 36676 + }, + { + "epoch": 1.9053834586466165, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 36677 + }, + { + "epoch": 1.9054436090225564, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 36678 + }, + { + "epoch": 1.9055037593984963, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9105, + "step": 36679 + }, + { + "epoch": 1.9055639097744361, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 36680 + }, + { + "epoch": 1.9056240601503758, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 36681 + }, + { + "epoch": 1.9056842105263159, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 36682 + }, + { + "epoch": 1.9057443609022555, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 36683 + }, + { + "epoch": 1.9058045112781956, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 36684 + }, + { + "epoch": 1.9058646616541353, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 36685 + }, + { + "epoch": 1.9059248120300751, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 36686 + }, + { + "epoch": 1.905984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 36687 + }, + { + "epoch": 1.9060451127819549, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 36688 + }, + { + "epoch": 1.9061052631578947, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 36689 + }, + { + "epoch": 1.9061654135338346, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 36690 + }, + { + "epoch": 1.9062255639097745, + "grad_norm": 1.8984375, + "learning_rate": 3e-05, + "loss": 0.9292, + "step": 36691 + }, + { + "epoch": 1.9062857142857141, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 36692 + }, + { + "epoch": 1.9063458646616542, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 36693 + }, + { + "epoch": 1.9064060150375939, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 36694 + }, + { + "epoch": 1.906466165413534, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 36695 + }, + { + "epoch": 1.9065263157894736, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 36696 + }, + { + "epoch": 1.9065864661654135, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1982, + "step": 36697 + }, + { + "epoch": 1.9066466165413534, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 36698 + }, + { + "epoch": 1.9067067669172932, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 36699 + }, + { + "epoch": 1.906766917293233, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 36700 + }, + { + "epoch": 1.906827067669173, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 36701 + }, + { + "epoch": 1.9068872180451129, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 36702 + }, + { + "epoch": 1.9069473684210525, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 36703 + }, + { + "epoch": 1.9070075187969926, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 36704 + }, + { + "epoch": 1.9070676691729322, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 36705 + }, + { + "epoch": 1.9071278195488723, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 36706 + }, + { + "epoch": 1.907187969924812, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 36707 + }, + { + "epoch": 1.9072481203007519, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 36708 + }, + { + "epoch": 1.9073082706766917, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 36709 + }, + { + "epoch": 1.9073684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 36710 + }, + { + "epoch": 1.9074285714285715, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 36711 + }, + { + "epoch": 1.9074887218045111, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 36712 + }, + { + "epoch": 1.9075488721804512, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 36713 + }, + { + "epoch": 1.9076090225563909, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 36714 + }, + { + "epoch": 1.907669172932331, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 36715 + }, + { + "epoch": 1.9077293233082706, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.947, + "step": 36716 + }, + { + "epoch": 1.9077894736842105, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 36717 + }, + { + "epoch": 1.9078496240601504, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 36718 + }, + { + "epoch": 1.9079097744360902, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 36719 + }, + { + "epoch": 1.90796992481203, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 36720 + }, + { + "epoch": 1.90803007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 36721 + }, + { + "epoch": 1.9080902255639098, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 36722 + }, + { + "epoch": 1.9081503759398495, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 36723 + }, + { + "epoch": 1.9082105263157896, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 36724 + }, + { + "epoch": 1.9082706766917292, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 36725 + }, + { + "epoch": 1.9083308270676693, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 36726 + }, + { + "epoch": 1.908390977443609, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 36727 + }, + { + "epoch": 1.9084511278195488, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1892, + "step": 36728 + }, + { + "epoch": 1.9085112781954887, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9286, + "step": 36729 + }, + { + "epoch": 1.9085714285714286, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 36730 + }, + { + "epoch": 1.9086315789473685, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 36731 + }, + { + "epoch": 1.908691729323308, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 36732 + }, + { + "epoch": 1.9087518796992482, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 36733 + }, + { + "epoch": 1.9088120300751878, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 36734 + }, + { + "epoch": 1.908872180451128, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 36735 + }, + { + "epoch": 1.9089323308270676, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9312, + "step": 36736 + }, + { + "epoch": 1.9089924812030077, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 36737 + }, + { + "epoch": 1.9090526315789473, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 36738 + }, + { + "epoch": 1.9091127819548872, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 36739 + }, + { + "epoch": 1.909172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 36740 + }, + { + "epoch": 1.909233082706767, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 36741 + }, + { + "epoch": 1.9092932330827068, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 36742 + }, + { + "epoch": 1.9093533834586465, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 36743 + }, + { + "epoch": 1.9094135338345866, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 36744 + }, + { + "epoch": 1.9094736842105262, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 36745 + }, + { + "epoch": 1.9095338345864663, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 36746 + }, + { + "epoch": 1.909593984962406, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 36747 + }, + { + "epoch": 1.9096541353383458, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 36748 + }, + { + "epoch": 1.9097142857142857, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 36749 + }, + { + "epoch": 1.9097744360902256, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 36750 + }, + { + "epoch": 1.9098345864661654, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 36751 + }, + { + "epoch": 1.9098947368421053, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 36752 + }, + { + "epoch": 1.9099548872180452, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 36753 + }, + { + "epoch": 1.9100150375939848, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 36754 + }, + { + "epoch": 1.910075187969925, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 36755 + }, + { + "epoch": 1.9101353383458646, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 36756 + }, + { + "epoch": 1.9101954887218047, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 36757 + }, + { + "epoch": 1.9102556390977443, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 36758 + }, + { + "epoch": 1.9103157894736842, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9479, + "step": 36759 + }, + { + "epoch": 1.910375939849624, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 36760 + }, + { + "epoch": 1.910436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 36761 + }, + { + "epoch": 1.9104962406015038, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 36762 + }, + { + "epoch": 1.9105563909774435, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 36763 + }, + { + "epoch": 1.9106165413533835, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 36764 + }, + { + "epoch": 1.9106766917293232, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 36765 + }, + { + "epoch": 1.9107368421052633, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 36766 + }, + { + "epoch": 1.910796992481203, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 36767 + }, + { + "epoch": 1.9108571428571428, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 36768 + }, + { + "epoch": 1.9109172932330827, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 36769 + }, + { + "epoch": 1.9109774436090226, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 36770 + }, + { + "epoch": 1.9110375939849624, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 36771 + }, + { + "epoch": 1.9110977443609023, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 36772 + }, + { + "epoch": 1.9111578947368422, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 36773 + }, + { + "epoch": 1.9112180451127818, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 36774 + }, + { + "epoch": 1.911278195488722, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9313, + "step": 36775 + }, + { + "epoch": 1.9113383458646616, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 36776 + }, + { + "epoch": 1.9113984962406017, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 36777 + }, + { + "epoch": 1.9114586466165413, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 36778 + }, + { + "epoch": 1.9115187969924812, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 36779 + }, + { + "epoch": 1.911578947368421, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 36780 + }, + { + "epoch": 1.911639097744361, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 36781 + }, + { + "epoch": 1.9116992481203008, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 36782 + }, + { + "epoch": 1.9117593984962404, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 36783 + }, + { + "epoch": 1.9118195488721805, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 36784 + }, + { + "epoch": 1.9118796992481202, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 36785 + }, + { + "epoch": 1.9119398496240603, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 36786 + }, + { + "epoch": 1.912, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 36787 + }, + { + "epoch": 1.91206015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 36788 + }, + { + "epoch": 1.9121203007518797, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 36789 + }, + { + "epoch": 1.9121804511278195, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 36790 + }, + { + "epoch": 1.9122406015037594, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 36791 + }, + { + "epoch": 1.9123007518796993, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 36792 + }, + { + "epoch": 1.9123609022556392, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 36793 + }, + { + "epoch": 1.9124210526315788, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 36794 + }, + { + "epoch": 1.912481203007519, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 36795 + }, + { + "epoch": 1.9125413533834585, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 36796 + }, + { + "epoch": 1.9126015037593986, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 36797 + }, + { + "epoch": 1.9126616541353383, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 36798 + }, + { + "epoch": 1.9127218045112782, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 36799 + }, + { + "epoch": 1.912781954887218, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 36800 + }, + { + "epoch": 1.912842105263158, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 36801 + }, + { + "epoch": 1.9129022556390978, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 36802 + }, + { + "epoch": 1.9129624060150376, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 36803 + }, + { + "epoch": 1.9130225563909775, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 36804 + }, + { + "epoch": 1.9130827067669172, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 36805 + }, + { + "epoch": 1.9131428571428573, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 36806 + }, + { + "epoch": 1.913203007518797, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 36807 + }, + { + "epoch": 1.913263157894737, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 36808 + }, + { + "epoch": 1.9133233082706766, + "grad_norm": 1.421875, + "learning_rate": 3e-05, + "loss": 1.1858, + "step": 36809 + }, + { + "epoch": 1.9133834586466165, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 36810 + }, + { + "epoch": 1.9134436090225564, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 36811 + }, + { + "epoch": 1.9135037593984963, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 36812 + }, + { + "epoch": 1.9135639097744361, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 36813 + }, + { + "epoch": 1.9136240601503758, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 36814 + }, + { + "epoch": 1.9136842105263159, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 36815 + }, + { + "epoch": 1.9137443609022555, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 36816 + }, + { + "epoch": 1.9138045112781956, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 36817 + }, + { + "epoch": 1.9138646616541353, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 36818 + }, + { + "epoch": 1.9139248120300751, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 36819 + }, + { + "epoch": 1.913984962406015, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 36820 + }, + { + "epoch": 1.9140451127819549, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 36821 + }, + { + "epoch": 1.9141052631578948, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 36822 + }, + { + "epoch": 1.9141654135338346, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1627, + "step": 36823 + }, + { + "epoch": 1.9142255639097745, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 36824 + }, + { + "epoch": 1.9142857142857141, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 36825 + }, + { + "epoch": 1.9143458646616542, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 36826 + }, + { + "epoch": 1.9144060150375939, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 36827 + }, + { + "epoch": 1.914466165413534, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 36828 + }, + { + "epoch": 1.9145263157894736, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 36829 + }, + { + "epoch": 1.9145864661654135, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 36830 + }, + { + "epoch": 1.9146466165413534, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 36831 + }, + { + "epoch": 1.9147067669172932, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 36832 + }, + { + "epoch": 1.9147669172932331, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 36833 + }, + { + "epoch": 1.9148270676691728, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 36834 + }, + { + "epoch": 1.9148872180451129, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 36835 + }, + { + "epoch": 1.9149473684210525, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9416, + "step": 36836 + }, + { + "epoch": 1.9150075187969926, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 36837 + }, + { + "epoch": 1.9150676691729323, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 36838 + }, + { + "epoch": 1.9151278195488723, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 36839 + }, + { + "epoch": 1.915187969924812, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 36840 + }, + { + "epoch": 1.9152481203007519, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 36841 + }, + { + "epoch": 1.9153082706766917, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 36842 + }, + { + "epoch": 1.9153684210526316, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 36843 + }, + { + "epoch": 1.9154285714285715, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 36844 + }, + { + "epoch": 1.9154887218045111, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 36845 + }, + { + "epoch": 1.9155488721804512, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 36846 + }, + { + "epoch": 1.9156090225563909, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 36847 + }, + { + "epoch": 1.915669172932331, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 36848 + }, + { + "epoch": 1.9157293233082706, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 36849 + }, + { + "epoch": 1.9157894736842105, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 36850 + }, + { + "epoch": 1.9158496240601504, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 36851 + }, + { + "epoch": 1.9159097744360902, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 36852 + }, + { + "epoch": 1.91596992481203, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 36853 + }, + { + "epoch": 1.91603007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 36854 + }, + { + "epoch": 1.9160902255639098, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 36855 + }, + { + "epoch": 1.9161503759398495, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 36856 + }, + { + "epoch": 1.9162105263157896, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 36857 + }, + { + "epoch": 1.9162706766917292, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 36858 + }, + { + "epoch": 1.9163308270676693, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 36859 + }, + { + "epoch": 1.916390977443609, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9045, + "step": 36860 + }, + { + "epoch": 1.9164511278195488, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 36861 + }, + { + "epoch": 1.9165112781954887, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 36862 + }, + { + "epoch": 1.9165714285714286, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 36863 + }, + { + "epoch": 1.9166315789473685, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 36864 + }, + { + "epoch": 1.9166917293233081, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 36865 + }, + { + "epoch": 1.9167518796992482, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 36866 + }, + { + "epoch": 1.9168120300751879, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 36867 + }, + { + "epoch": 1.916872180451128, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 36868 + }, + { + "epoch": 1.9169323308270676, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 36869 + }, + { + "epoch": 1.9169924812030075, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 36870 + }, + { + "epoch": 1.9170526315789473, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 36871 + }, + { + "epoch": 1.9171127819548872, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 36872 + }, + { + "epoch": 1.917172932330827, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 36873 + }, + { + "epoch": 1.917233082706767, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 36874 + }, + { + "epoch": 1.9172932330827068, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 36875 + }, + { + "epoch": 1.9173533834586465, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 36876 + }, + { + "epoch": 1.9174135338345866, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 36877 + }, + { + "epoch": 1.9174736842105262, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 36878 + }, + { + "epoch": 1.9175338345864663, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 36879 + }, + { + "epoch": 1.917593984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 36880 + }, + { + "epoch": 1.9176541353383458, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 36881 + }, + { + "epoch": 1.9177142857142857, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 36882 + }, + { + "epoch": 1.9177744360902256, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 36883 + }, + { + "epoch": 1.9178345864661654, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 36884 + }, + { + "epoch": 1.917894736842105, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 36885 + }, + { + "epoch": 1.9179548872180452, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 36886 + }, + { + "epoch": 1.9180150375939848, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 36887 + }, + { + "epoch": 1.918075187969925, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2353, + "step": 36888 + }, + { + "epoch": 1.9181353383458646, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 36889 + }, + { + "epoch": 1.9181954887218047, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 36890 + }, + { + "epoch": 1.9182556390977443, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 36891 + }, + { + "epoch": 1.9183157894736842, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 36892 + }, + { + "epoch": 1.918375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 36893 + }, + { + "epoch": 1.918436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 36894 + }, + { + "epoch": 1.9184962406015038, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 36895 + }, + { + "epoch": 1.9185563909774435, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 36896 + }, + { + "epoch": 1.9186165413533836, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 36897 + }, + { + "epoch": 1.9186766917293232, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 36898 + }, + { + "epoch": 1.9187368421052633, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 36899 + }, + { + "epoch": 1.918796992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 36900 + }, + { + "epoch": 1.9188571428571428, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 36901 + }, + { + "epoch": 1.9189172932330827, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 36902 + }, + { + "epoch": 1.9189774436090226, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 36903 + }, + { + "epoch": 1.9190375939849624, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 36904 + }, + { + "epoch": 1.9190977443609023, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 36905 + }, + { + "epoch": 1.9191578947368422, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 36906 + }, + { + "epoch": 1.9192180451127818, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 36907 + }, + { + "epoch": 1.919278195488722, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 36908 + }, + { + "epoch": 1.9193383458646616, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.8782, + "step": 36909 + }, + { + "epoch": 1.9193984962406017, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9396, + "step": 36910 + }, + { + "epoch": 1.9194586466165413, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 36911 + }, + { + "epoch": 1.9195187969924812, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2441, + "step": 36912 + }, + { + "epoch": 1.919578947368421, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 36913 + }, + { + "epoch": 1.919639097744361, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.187, + "step": 36914 + }, + { + "epoch": 1.9196992481203008, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 36915 + }, + { + "epoch": 1.9197593984962404, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 36916 + }, + { + "epoch": 1.9198195488721805, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 36917 + }, + { + "epoch": 1.9198796992481202, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 36918 + }, + { + "epoch": 1.9199398496240603, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 36919 + }, + { + "epoch": 1.92, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 36920 + }, + { + "epoch": 1.9200601503759398, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 36921 + }, + { + "epoch": 1.9201203007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 36922 + }, + { + "epoch": 1.9201804511278195, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 36923 + }, + { + "epoch": 1.9202406015037594, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 36924 + }, + { + "epoch": 1.9203007518796993, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1434, + "step": 36925 + }, + { + "epoch": 1.9203609022556392, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 36926 + }, + { + "epoch": 1.9204210526315788, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 36927 + }, + { + "epoch": 1.920481203007519, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 36928 + }, + { + "epoch": 1.9205413533834585, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 36929 + }, + { + "epoch": 1.9206015037593986, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 36930 + }, + { + "epoch": 1.9206616541353383, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 36931 + }, + { + "epoch": 1.9207218045112782, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 36932 + }, + { + "epoch": 1.920781954887218, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 36933 + }, + { + "epoch": 1.920842105263158, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 36934 + }, + { + "epoch": 1.9209022556390978, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 36935 + }, + { + "epoch": 1.9209624060150376, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1478, + "step": 36936 + }, + { + "epoch": 1.9210225563909775, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 36937 + }, + { + "epoch": 1.9210827067669172, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 36938 + }, + { + "epoch": 1.9211428571428573, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 36939 + }, + { + "epoch": 1.921203007518797, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 36940 + }, + { + "epoch": 1.921263157894737, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 36941 + }, + { + "epoch": 1.9213233082706767, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 36942 + }, + { + "epoch": 1.9213834586466165, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 36943 + }, + { + "epoch": 1.9214436090225564, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 36944 + }, + { + "epoch": 1.9215037593984963, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 36945 + }, + { + "epoch": 1.9215639097744361, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 36946 + }, + { + "epoch": 1.9216240601503758, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 36947 + }, + { + "epoch": 1.9216842105263159, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 36948 + }, + { + "epoch": 1.9217443609022555, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9217, + "step": 36949 + }, + { + "epoch": 1.9218045112781956, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 36950 + }, + { + "epoch": 1.9218646616541353, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 36951 + }, + { + "epoch": 1.9219248120300751, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 36952 + }, + { + "epoch": 1.921984962406015, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 36953 + }, + { + "epoch": 1.922045112781955, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 36954 + }, + { + "epoch": 1.9221052631578948, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 36955 + }, + { + "epoch": 1.9221654135338346, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 36956 + }, + { + "epoch": 1.9222255639097745, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 36957 + }, + { + "epoch": 1.9222857142857142, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 36958 + }, + { + "epoch": 1.9223458646616542, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1946, + "step": 36959 + }, + { + "epoch": 1.922406015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 36960 + }, + { + "epoch": 1.922466165413534, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 36961 + }, + { + "epoch": 1.9225263157894736, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 36962 + }, + { + "epoch": 1.9225864661654135, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 36963 + }, + { + "epoch": 1.9226466165413534, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 36964 + }, + { + "epoch": 1.9227067669172933, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 36965 + }, + { + "epoch": 1.9227669172932331, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9487, + "step": 36966 + }, + { + "epoch": 1.9228270676691728, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 36967 + }, + { + "epoch": 1.9228872180451129, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 36968 + }, + { + "epoch": 1.9229473684210525, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 36969 + }, + { + "epoch": 1.9230075187969926, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 36970 + }, + { + "epoch": 1.9230676691729323, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 36971 + }, + { + "epoch": 1.9231278195488721, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 36972 + }, + { + "epoch": 1.923187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 36973 + }, + { + "epoch": 1.9232481203007519, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 36974 + }, + { + "epoch": 1.9233082706766917, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 36975 + }, + { + "epoch": 1.9233684210526316, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 36976 + }, + { + "epoch": 1.9234285714285715, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 36977 + }, + { + "epoch": 1.9234887218045111, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 36978 + }, + { + "epoch": 1.9235488721804512, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 36979 + }, + { + "epoch": 1.9236090225563909, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 36980 + }, + { + "epoch": 1.923669172932331, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 36981 + }, + { + "epoch": 1.9237293233082706, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 36982 + }, + { + "epoch": 1.9237894736842105, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 36983 + }, + { + "epoch": 1.9238496240601504, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2288, + "step": 36984 + }, + { + "epoch": 1.9239097744360902, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 36985 + }, + { + "epoch": 1.92396992481203, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 36986 + }, + { + "epoch": 1.92403007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 36987 + }, + { + "epoch": 1.9240902255639099, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 36988 + }, + { + "epoch": 1.9241503759398495, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 36989 + }, + { + "epoch": 1.9242105263157896, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 36990 + }, + { + "epoch": 1.9242706766917292, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 36991 + }, + { + "epoch": 1.9243308270676693, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 36992 + }, + { + "epoch": 1.924390977443609, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 36993 + }, + { + "epoch": 1.9244511278195489, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 36994 + }, + { + "epoch": 1.9245112781954887, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 36995 + }, + { + "epoch": 1.9245714285714286, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 36996 + }, + { + "epoch": 1.9246315789473685, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 36997 + }, + { + "epoch": 1.9246917293233081, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 36998 + }, + { + "epoch": 1.9247518796992482, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 36999 + }, + { + "epoch": 1.9248120300751879, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 37000 + }, + { + "epoch": 1.924872180451128, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 37001 + }, + { + "epoch": 1.9249323308270676, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 37002 + }, + { + "epoch": 1.9249924812030075, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 37003 + }, + { + "epoch": 1.9250526315789473, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 37004 + }, + { + "epoch": 1.9251127819548872, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 37005 + }, + { + "epoch": 1.925172932330827, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 37006 + }, + { + "epoch": 1.925233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 37007 + }, + { + "epoch": 1.9252932330827068, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 37008 + }, + { + "epoch": 1.9253533834586465, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 37009 + }, + { + "epoch": 1.9254135338345866, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9527, + "step": 37010 + }, + { + "epoch": 1.9254736842105262, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 37011 + }, + { + "epoch": 1.9255338345864663, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 37012 + }, + { + "epoch": 1.925593984962406, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 37013 + }, + { + "epoch": 1.9256541353383458, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 37014 + }, + { + "epoch": 1.9257142857142857, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 37015 + }, + { + "epoch": 1.9257744360902256, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 37016 + }, + { + "epoch": 1.9258345864661655, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 37017 + }, + { + "epoch": 1.925894736842105, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 37018 + }, + { + "epoch": 1.9259548872180452, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 37019 + }, + { + "epoch": 1.9260150375939848, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 37020 + }, + { + "epoch": 1.926075187969925, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 37021 + }, + { + "epoch": 1.9261353383458646, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 37022 + }, + { + "epoch": 1.9261954887218045, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9313, + "step": 37023 + }, + { + "epoch": 1.9262556390977443, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 37024 + }, + { + "epoch": 1.9263157894736842, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 37025 + }, + { + "epoch": 1.926375939849624, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 37026 + }, + { + "epoch": 1.926436090225564, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 37027 + }, + { + "epoch": 1.9264962406015038, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 37028 + }, + { + "epoch": 1.9265563909774435, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 37029 + }, + { + "epoch": 1.9266165413533836, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 37030 + }, + { + "epoch": 1.9266766917293232, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 37031 + }, + { + "epoch": 1.9267368421052633, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 37032 + }, + { + "epoch": 1.926796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 37033 + }, + { + "epoch": 1.9268571428571428, + "grad_norm": 1.40625, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 37034 + }, + { + "epoch": 1.9269172932330827, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 37035 + }, + { + "epoch": 1.9269774436090226, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 37036 + }, + { + "epoch": 1.9270375939849624, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 37037 + }, + { + "epoch": 1.9270977443609023, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 37038 + }, + { + "epoch": 1.9271578947368422, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 37039 + }, + { + "epoch": 1.9272180451127818, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 37040 + }, + { + "epoch": 1.927278195488722, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 37041 + }, + { + "epoch": 1.9273383458646616, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 37042 + }, + { + "epoch": 1.9273984962406017, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 37043 + }, + { + "epoch": 1.9274586466165413, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 37044 + }, + { + "epoch": 1.9275187969924812, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 37045 + }, + { + "epoch": 1.927578947368421, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 37046 + }, + { + "epoch": 1.927639097744361, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 37047 + }, + { + "epoch": 1.9276992481203008, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 37048 + }, + { + "epoch": 1.9277593984962405, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 37049 + }, + { + "epoch": 1.9278195488721805, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.2925, + "step": 37050 + }, + { + "epoch": 1.9278796992481202, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 37051 + }, + { + "epoch": 1.9279398496240603, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 37052 + }, + { + "epoch": 1.928, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 37053 + }, + { + "epoch": 1.9280601503759398, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 37054 + }, + { + "epoch": 1.9281203007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 37055 + }, + { + "epoch": 1.9281804511278196, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 37056 + }, + { + "epoch": 1.9282406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 37057 + }, + { + "epoch": 1.9283007518796993, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 37058 + }, + { + "epoch": 1.9283609022556392, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2561, + "step": 37059 + }, + { + "epoch": 1.9284210526315788, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 37060 + }, + { + "epoch": 1.928481203007519, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 37061 + }, + { + "epoch": 1.9285413533834586, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 37062 + }, + { + "epoch": 1.9286015037593986, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 37063 + }, + { + "epoch": 1.9286616541353383, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 37064 + }, + { + "epoch": 1.9287218045112782, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 37065 + }, + { + "epoch": 1.928781954887218, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 37066 + }, + { + "epoch": 1.928842105263158, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 37067 + }, + { + "epoch": 1.9289022556390978, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 37068 + }, + { + "epoch": 1.9289624060150374, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 37069 + }, + { + "epoch": 1.9290225563909775, + "grad_norm": 1.59375, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 37070 + }, + { + "epoch": 1.9290827067669172, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 37071 + }, + { + "epoch": 1.9291428571428573, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 37072 + }, + { + "epoch": 1.929203007518797, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 37073 + }, + { + "epoch": 1.9292631578947368, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 37074 + }, + { + "epoch": 1.9293233082706767, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 37075 + }, + { + "epoch": 1.9293834586466165, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 37076 + }, + { + "epoch": 1.9294436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 37077 + }, + { + "epoch": 1.9295037593984963, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 37078 + }, + { + "epoch": 1.9295639097744361, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 37079 + }, + { + "epoch": 1.9296240601503758, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2087, + "step": 37080 + }, + { + "epoch": 1.929684210526316, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9488, + "step": 37081 + }, + { + "epoch": 1.9297443609022555, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 37082 + }, + { + "epoch": 1.9298045112781956, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 37083 + }, + { + "epoch": 1.9298646616541353, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 37084 + }, + { + "epoch": 1.9299248120300752, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 37085 + }, + { + "epoch": 1.929984962406015, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 37086 + }, + { + "epoch": 1.930045112781955, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 37087 + }, + { + "epoch": 1.9301052631578948, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 37088 + }, + { + "epoch": 1.9301654135338346, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 37089 + }, + { + "epoch": 1.9302255639097745, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 37090 + }, + { + "epoch": 1.9302857142857142, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 37091 + }, + { + "epoch": 1.9303458646616543, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 37092 + }, + { + "epoch": 1.930406015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 37093 + }, + { + "epoch": 1.930466165413534, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 37094 + }, + { + "epoch": 1.9305263157894736, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 37095 + }, + { + "epoch": 1.9305864661654135, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 37096 + }, + { + "epoch": 1.9306466165413534, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 37097 + }, + { + "epoch": 1.9307067669172933, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 37098 + }, + { + "epoch": 1.9307669172932331, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 37099 + }, + { + "epoch": 1.9308270676691728, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 37100 + }, + { + "epoch": 1.9308872180451129, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 37101 + }, + { + "epoch": 1.9309473684210525, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 37102 + }, + { + "epoch": 1.9310075187969926, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 37103 + }, + { + "epoch": 1.9310676691729323, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 37104 + }, + { + "epoch": 1.9311278195488721, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 37105 + }, + { + "epoch": 1.931187969924812, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 37106 + }, + { + "epoch": 1.9312481203007519, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 37107 + }, + { + "epoch": 1.9313082706766918, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 37108 + }, + { + "epoch": 1.9313684210526316, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 37109 + }, + { + "epoch": 1.9314285714285715, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 37110 + }, + { + "epoch": 1.9314887218045111, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1482, + "step": 37111 + }, + { + "epoch": 1.9315488721804512, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 37112 + }, + { + "epoch": 1.9316090225563909, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 37113 + }, + { + "epoch": 1.931669172932331, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 37114 + }, + { + "epoch": 1.9317293233082706, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 37115 + }, + { + "epoch": 1.9317894736842105, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 37116 + }, + { + "epoch": 1.9318496240601504, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 37117 + }, + { + "epoch": 1.9319097744360902, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 37118 + }, + { + "epoch": 1.9319699248120301, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 37119 + }, + { + "epoch": 1.9320300751879698, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1167, + "step": 37120 + }, + { + "epoch": 1.9320902255639099, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 37121 + }, + { + "epoch": 1.9321503759398495, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 37122 + }, + { + "epoch": 1.9322105263157896, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 37123 + }, + { + "epoch": 1.9322706766917292, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 37124 + }, + { + "epoch": 1.9323308270676691, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 37125 + }, + { + "epoch": 1.932390977443609, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 37126 + }, + { + "epoch": 1.9324511278195489, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 37127 + }, + { + "epoch": 1.9325112781954887, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 37128 + }, + { + "epoch": 1.9325714285714286, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 37129 + }, + { + "epoch": 1.9326315789473685, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 37130 + }, + { + "epoch": 1.9326917293233081, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 37131 + }, + { + "epoch": 1.9327518796992482, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 37132 + }, + { + "epoch": 1.9328120300751879, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 37133 + }, + { + "epoch": 1.932872180451128, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 37134 + }, + { + "epoch": 1.9329323308270676, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 37135 + }, + { + "epoch": 1.9329924812030075, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 37136 + }, + { + "epoch": 1.9330526315789474, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 37137 + }, + { + "epoch": 1.9331127819548872, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 37138 + }, + { + "epoch": 1.933172932330827, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 37139 + }, + { + "epoch": 1.933233082706767, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 37140 + }, + { + "epoch": 1.9332932330827068, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 37141 + }, + { + "epoch": 1.9333533834586465, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 37142 + }, + { + "epoch": 1.9334135338345866, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 37143 + }, + { + "epoch": 1.9334736842105262, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 37144 + }, + { + "epoch": 1.9335338345864663, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 37145 + }, + { + "epoch": 1.933593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 37146 + }, + { + "epoch": 1.9336541353383458, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 37147 + }, + { + "epoch": 1.9337142857142857, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 37148 + }, + { + "epoch": 1.9337744360902256, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 37149 + }, + { + "epoch": 1.9338345864661655, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 37150 + }, + { + "epoch": 1.933894736842105, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 37151 + }, + { + "epoch": 1.9339548872180452, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 37152 + }, + { + "epoch": 1.9340150375939849, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 37153 + }, + { + "epoch": 1.934075187969925, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 37154 + }, + { + "epoch": 1.9341353383458646, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 37155 + }, + { + "epoch": 1.9341954887218045, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 37156 + }, + { + "epoch": 1.9342556390977443, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 37157 + }, + { + "epoch": 1.9343157894736842, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 37158 + }, + { + "epoch": 1.934375939849624, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 37159 + }, + { + "epoch": 1.934436090225564, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 37160 + }, + { + "epoch": 1.9344962406015038, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 37161 + }, + { + "epoch": 1.9345563909774435, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 37162 + }, + { + "epoch": 1.9346165413533836, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 37163 + }, + { + "epoch": 1.9346766917293232, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 37164 + }, + { + "epoch": 1.9347368421052633, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 37165 + }, + { + "epoch": 1.934796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 37166 + }, + { + "epoch": 1.9348571428571428, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 37167 + }, + { + "epoch": 1.9349172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 37168 + }, + { + "epoch": 1.9349774436090226, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 37169 + }, + { + "epoch": 1.9350375939849624, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.14, + "step": 37170 + }, + { + "epoch": 1.935097744360902, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 37171 + }, + { + "epoch": 1.9351578947368422, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 37172 + }, + { + "epoch": 1.9352180451127818, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 37173 + }, + { + "epoch": 1.935278195488722, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 37174 + }, + { + "epoch": 1.9353383458646616, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 37175 + }, + { + "epoch": 1.9353984962406015, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 37176 + }, + { + "epoch": 1.9354586466165413, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 37177 + }, + { + "epoch": 1.9355187969924812, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 37178 + }, + { + "epoch": 1.935578947368421, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 37179 + }, + { + "epoch": 1.935639097744361, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 37180 + }, + { + "epoch": 1.9356992481203008, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 37181 + }, + { + "epoch": 1.9357593984962405, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 37182 + }, + { + "epoch": 1.9358195488721806, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 37183 + }, + { + "epoch": 1.9358796992481202, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.2001, + "step": 37184 + }, + { + "epoch": 1.9359398496240603, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 37185 + }, + { + "epoch": 1.936, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 37186 + }, + { + "epoch": 1.9360601503759398, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9361, + "step": 37187 + }, + { + "epoch": 1.9361203007518797, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 37188 + }, + { + "epoch": 1.9361804511278196, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 37189 + }, + { + "epoch": 1.9362406015037594, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 37190 + }, + { + "epoch": 1.9363007518796993, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 37191 + }, + { + "epoch": 1.9363609022556392, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 37192 + }, + { + "epoch": 1.9364210526315788, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 37193 + }, + { + "epoch": 1.936481203007519, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 37194 + }, + { + "epoch": 1.9365413533834586, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 37195 + }, + { + "epoch": 1.9366015037593987, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 37196 + }, + { + "epoch": 1.9366616541353383, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 37197 + }, + { + "epoch": 1.9367218045112782, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 37198 + }, + { + "epoch": 1.936781954887218, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 37199 + }, + { + "epoch": 1.936842105263158, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 37200 + }, + { + "epoch": 1.9369022556390978, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 37201 + }, + { + "epoch": 1.9369624060150374, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 37202 + }, + { + "epoch": 1.9370225563909775, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 37203 + }, + { + "epoch": 1.9370827067669172, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 37204 + }, + { + "epoch": 1.9371428571428573, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 37205 + }, + { + "epoch": 1.937203007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 37206 + }, + { + "epoch": 1.9372631578947368, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 37207 + }, + { + "epoch": 1.9373233082706767, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 37208 + }, + { + "epoch": 1.9373834586466165, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 37209 + }, + { + "epoch": 1.9374436090225564, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 37210 + }, + { + "epoch": 1.9375037593984963, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 37211 + }, + { + "epoch": 1.9375639097744362, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 37212 + }, + { + "epoch": 1.9376240601503758, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 37213 + }, + { + "epoch": 1.937684210526316, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 37214 + }, + { + "epoch": 1.9377443609022555, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 37215 + }, + { + "epoch": 1.9378045112781956, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1414, + "step": 37216 + }, + { + "epoch": 1.9378646616541353, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 37217 + }, + { + "epoch": 1.9379248120300752, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 37218 + }, + { + "epoch": 1.937984962406015, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 37219 + }, + { + "epoch": 1.938045112781955, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 37220 + }, + { + "epoch": 1.9381052631578948, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 37221 + }, + { + "epoch": 1.9381654135338344, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 37222 + }, + { + "epoch": 1.9382255639097745, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 37223 + }, + { + "epoch": 1.9382857142857142, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 37224 + }, + { + "epoch": 1.9383458646616543, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 37225 + }, + { + "epoch": 1.938406015037594, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 37226 + }, + { + "epoch": 1.9384661654135338, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 37227 + }, + { + "epoch": 1.9385263157894737, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 37228 + }, + { + "epoch": 1.9385864661654135, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 37229 + }, + { + "epoch": 1.9386466165413534, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 37230 + }, + { + "epoch": 1.9387067669172933, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 37231 + }, + { + "epoch": 1.9387669172932331, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 37232 + }, + { + "epoch": 1.9388270676691728, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9519, + "step": 37233 + }, + { + "epoch": 1.9388872180451129, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 37234 + }, + { + "epoch": 1.9389473684210525, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 37235 + }, + { + "epoch": 1.9390075187969926, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 37236 + }, + { + "epoch": 1.9390676691729323, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 37237 + }, + { + "epoch": 1.9391278195488721, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 37238 + }, + { + "epoch": 1.939187969924812, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 37239 + }, + { + "epoch": 1.9392481203007519, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 37240 + }, + { + "epoch": 1.9393082706766918, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 37241 + }, + { + "epoch": 1.9393684210526316, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 37242 + }, + { + "epoch": 1.9394285714285715, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 37243 + }, + { + "epoch": 1.9394887218045112, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 37244 + }, + { + "epoch": 1.9395488721804512, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 37245 + }, + { + "epoch": 1.939609022556391, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 37246 + }, + { + "epoch": 1.939669172932331, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 37247 + }, + { + "epoch": 1.9397293233082706, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 37248 + }, + { + "epoch": 1.9397894736842105, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 37249 + }, + { + "epoch": 1.9398496240601504, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 37250 + }, + { + "epoch": 1.9399097744360903, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 37251 + }, + { + "epoch": 1.9399699248120301, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 37252 + }, + { + "epoch": 1.9400300751879698, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 37253 + }, + { + "epoch": 1.9400902255639099, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 37254 + }, + { + "epoch": 1.9401503759398495, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 37255 + }, + { + "epoch": 1.9402105263157896, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 37256 + }, + { + "epoch": 1.9402706766917293, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 37257 + }, + { + "epoch": 1.9403308270676691, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 37258 + }, + { + "epoch": 1.940390977443609, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 37259 + }, + { + "epoch": 1.9404511278195489, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 37260 + }, + { + "epoch": 1.9405112781954887, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 37261 + }, + { + "epoch": 1.9405714285714286, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 37262 + }, + { + "epoch": 1.9406315789473685, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 37263 + }, + { + "epoch": 1.9406917293233081, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 37264 + }, + { + "epoch": 1.9407518796992482, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 37265 + }, + { + "epoch": 1.9408120300751879, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 37266 + }, + { + "epoch": 1.940872180451128, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 37267 + }, + { + "epoch": 1.9409323308270676, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 37268 + }, + { + "epoch": 1.9409924812030075, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 37269 + }, + { + "epoch": 1.9410526315789474, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 37270 + }, + { + "epoch": 1.9411127819548872, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 37271 + }, + { + "epoch": 1.941172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 37272 + }, + { + "epoch": 1.9412330827067668, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 37273 + }, + { + "epoch": 1.9412932330827068, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 37274 + }, + { + "epoch": 1.9413533834586465, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 37275 + }, + { + "epoch": 1.9414135338345866, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 37276 + }, + { + "epoch": 1.9414736842105262, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 37277 + }, + { + "epoch": 1.9415338345864661, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 37278 + }, + { + "epoch": 1.941593984962406, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 37279 + }, + { + "epoch": 1.9416541353383459, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 37280 + }, + { + "epoch": 1.9417142857142857, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 37281 + }, + { + "epoch": 1.9417744360902256, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 37282 + }, + { + "epoch": 1.9418345864661655, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 37283 + }, + { + "epoch": 1.9418947368421051, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.2551, + "step": 37284 + }, + { + "epoch": 1.9419548872180452, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 37285 + }, + { + "epoch": 1.9420150375939849, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 37286 + }, + { + "epoch": 1.942075187969925, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 37287 + }, + { + "epoch": 1.9421353383458646, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 37288 + }, + { + "epoch": 1.9421954887218045, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 37289 + }, + { + "epoch": 1.9422556390977443, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 37290 + }, + { + "epoch": 1.9423157894736842, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 37291 + }, + { + "epoch": 1.942375939849624, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 37292 + }, + { + "epoch": 1.942436090225564, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 37293 + }, + { + "epoch": 1.9424962406015038, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 37294 + }, + { + "epoch": 1.9425563909774435, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 37295 + }, + { + "epoch": 1.9426165413533836, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 37296 + }, + { + "epoch": 1.9426766917293232, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 37297 + }, + { + "epoch": 1.9427368421052633, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 37298 + }, + { + "epoch": 1.942796992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 37299 + }, + { + "epoch": 1.9428571428571428, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 37300 + }, + { + "epoch": 1.9429172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 37301 + }, + { + "epoch": 1.9429774436090226, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 37302 + }, + { + "epoch": 1.9430375939849625, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 37303 + }, + { + "epoch": 1.943097744360902, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 37304 + }, + { + "epoch": 1.9431578947368422, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 37305 + }, + { + "epoch": 1.9432180451127818, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 37306 + }, + { + "epoch": 1.943278195488722, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 37307 + }, + { + "epoch": 1.9433383458646616, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 37308 + }, + { + "epoch": 1.9433984962406015, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 37309 + }, + { + "epoch": 1.9434586466165413, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 37310 + }, + { + "epoch": 1.9435187969924812, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 37311 + }, + { + "epoch": 1.943578947368421, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 37312 + }, + { + "epoch": 1.943639097744361, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 37313 + }, + { + "epoch": 1.9436992481203008, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 37314 + }, + { + "epoch": 1.9437593984962405, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.939, + "step": 37315 + }, + { + "epoch": 1.9438195488721806, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 37316 + }, + { + "epoch": 1.9438796992481202, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 37317 + }, + { + "epoch": 1.9439398496240603, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9277, + "step": 37318 + }, + { + "epoch": 1.944, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 37319 + }, + { + "epoch": 1.9440601503759398, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 37320 + }, + { + "epoch": 1.9441203007518797, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 37321 + }, + { + "epoch": 1.9441804511278196, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 37322 + }, + { + "epoch": 1.9442406015037594, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 37323 + }, + { + "epoch": 1.944300751879699, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 37324 + }, + { + "epoch": 1.9443609022556392, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 37325 + }, + { + "epoch": 1.9444210526315788, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 37326 + }, + { + "epoch": 1.944481203007519, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 37327 + }, + { + "epoch": 1.9445413533834586, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 37328 + }, + { + "epoch": 1.9446015037593984, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 37329 + }, + { + "epoch": 1.9446616541353383, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 37330 + }, + { + "epoch": 1.9447218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 37331 + }, + { + "epoch": 1.944781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 37332 + }, + { + "epoch": 1.944842105263158, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 37333 + }, + { + "epoch": 1.9449022556390978, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 37334 + }, + { + "epoch": 1.9449624060150374, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 37335 + }, + { + "epoch": 1.9450225563909775, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 37336 + }, + { + "epoch": 1.9450827067669172, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 37337 + }, + { + "epoch": 1.9451428571428573, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 37338 + }, + { + "epoch": 1.945203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 37339 + }, + { + "epoch": 1.9452631578947368, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 37340 + }, + { + "epoch": 1.9453233082706767, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 37341 + }, + { + "epoch": 1.9453834586466165, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 37342 + }, + { + "epoch": 1.9454436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 37343 + }, + { + "epoch": 1.9455037593984963, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 37344 + }, + { + "epoch": 1.9455639097744362, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 37345 + }, + { + "epoch": 1.9456240601503758, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 37346 + }, + { + "epoch": 1.945684210526316, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1516, + "step": 37347 + }, + { + "epoch": 1.9457443609022556, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9504, + "step": 37348 + }, + { + "epoch": 1.9458045112781956, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 37349 + }, + { + "epoch": 1.9458646616541353, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 37350 + }, + { + "epoch": 1.9459248120300752, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 37351 + }, + { + "epoch": 1.945984962406015, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 37352 + }, + { + "epoch": 1.946045112781955, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 37353 + }, + { + "epoch": 1.9461052631578948, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 37354 + }, + { + "epoch": 1.9461654135338344, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 37355 + }, + { + "epoch": 1.9462255639097745, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 37356 + }, + { + "epoch": 1.9462857142857142, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 37357 + }, + { + "epoch": 1.9463458646616543, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 37358 + }, + { + "epoch": 1.946406015037594, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 37359 + }, + { + "epoch": 1.9464661654135338, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 37360 + }, + { + "epoch": 1.9465263157894737, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 37361 + }, + { + "epoch": 1.9465864661654135, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 37362 + }, + { + "epoch": 1.9466466165413534, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 37363 + }, + { + "epoch": 1.9467067669172933, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 37364 + }, + { + "epoch": 1.9467669172932331, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 37365 + }, + { + "epoch": 1.9468270676691728, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 37366 + }, + { + "epoch": 1.9468872180451129, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 37367 + }, + { + "epoch": 1.9469473684210525, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 37368 + }, + { + "epoch": 1.9470075187969926, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 37369 + }, + { + "epoch": 1.9470676691729323, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 37370 + }, + { + "epoch": 1.9471278195488722, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 37371 + }, + { + "epoch": 1.947187969924812, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 37372 + }, + { + "epoch": 1.947248120300752, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 37373 + }, + { + "epoch": 1.9473082706766918, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 37374 + }, + { + "epoch": 1.9473684210526314, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 37375 + }, + { + "epoch": 1.9474285714285715, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 37376 + }, + { + "epoch": 1.9474887218045112, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 37377 + }, + { + "epoch": 1.9475488721804513, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 37378 + }, + { + "epoch": 1.947609022556391, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 37379 + }, + { + "epoch": 1.9476691729323308, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 37380 + }, + { + "epoch": 1.9477293233082706, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 37381 + }, + { + "epoch": 1.9477894736842105, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 37382 + }, + { + "epoch": 1.9478496240601504, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9235, + "step": 37383 + }, + { + "epoch": 1.9479097744360903, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 37384 + }, + { + "epoch": 1.9479699248120301, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 37385 + }, + { + "epoch": 1.9480300751879698, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 37386 + }, + { + "epoch": 1.9480902255639099, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 37387 + }, + { + "epoch": 1.9481503759398495, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 37388 + }, + { + "epoch": 1.9482105263157896, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 37389 + }, + { + "epoch": 1.9482706766917293, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 37390 + }, + { + "epoch": 1.9483308270676691, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 37391 + }, + { + "epoch": 1.948390977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 37392 + }, + { + "epoch": 1.9484511278195489, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 37393 + }, + { + "epoch": 1.9485112781954887, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 37394 + }, + { + "epoch": 1.9485714285714286, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 37395 + }, + { + "epoch": 1.9486315789473685, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 37396 + }, + { + "epoch": 1.9486917293233081, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 37397 + }, + { + "epoch": 1.9487518796992482, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 37398 + }, + { + "epoch": 1.9488120300751879, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 37399 + }, + { + "epoch": 1.948872180451128, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 37400 + }, + { + "epoch": 1.9489323308270676, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 37401 + }, + { + "epoch": 1.9489924812030075, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 37402 + }, + { + "epoch": 1.9490526315789474, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 37403 + }, + { + "epoch": 1.9491127819548872, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 37404 + }, + { + "epoch": 1.9491729323308271, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 37405 + }, + { + "epoch": 1.9492330827067668, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 37406 + }, + { + "epoch": 1.9492932330827069, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 37407 + }, + { + "epoch": 1.9493533834586465, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 37408 + }, + { + "epoch": 1.9494135338345866, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9531, + "step": 37409 + }, + { + "epoch": 1.9494736842105262, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 37410 + }, + { + "epoch": 1.9495338345864661, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 37411 + }, + { + "epoch": 1.949593984962406, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 37412 + }, + { + "epoch": 1.9496541353383459, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 37413 + }, + { + "epoch": 1.9497142857142857, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 37414 + }, + { + "epoch": 1.9497744360902256, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 37415 + }, + { + "epoch": 1.9498345864661655, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 37416 + }, + { + "epoch": 1.9498947368421051, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 37417 + }, + { + "epoch": 1.9499548872180452, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 37418 + }, + { + "epoch": 1.9500150375939849, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 37419 + }, + { + "epoch": 1.950075187969925, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 37420 + }, + { + "epoch": 1.9501353383458646, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 37421 + }, + { + "epoch": 1.9501954887218045, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 37422 + }, + { + "epoch": 1.9502556390977444, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 37423 + }, + { + "epoch": 1.9503157894736842, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 37424 + }, + { + "epoch": 1.950375939849624, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 37425 + }, + { + "epoch": 1.9504360902255637, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 37426 + }, + { + "epoch": 1.9504962406015038, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 37427 + }, + { + "epoch": 1.9505563909774435, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 37428 + }, + { + "epoch": 1.9506165413533836, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 37429 + }, + { + "epoch": 1.9506766917293232, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 37430 + }, + { + "epoch": 1.950736842105263, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 37431 + }, + { + "epoch": 1.950796992481203, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 37432 + }, + { + "epoch": 1.9508571428571428, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 37433 + }, + { + "epoch": 1.9509172932330827, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 37434 + }, + { + "epoch": 1.9509774436090226, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 37435 + }, + { + "epoch": 1.9510375939849625, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 37436 + }, + { + "epoch": 1.951097744360902, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 37437 + }, + { + "epoch": 1.9511578947368422, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 37438 + }, + { + "epoch": 1.9512180451127819, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 37439 + }, + { + "epoch": 1.951278195488722, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 37440 + }, + { + "epoch": 1.9513383458646616, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 37441 + }, + { + "epoch": 1.9513984962406015, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 37442 + }, + { + "epoch": 1.9514586466165413, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 37443 + }, + { + "epoch": 1.9515187969924812, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 37444 + }, + { + "epoch": 1.951578947368421, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 37445 + }, + { + "epoch": 1.951639097744361, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 37446 + }, + { + "epoch": 1.9516992481203008, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 37447 + }, + { + "epoch": 1.9517593984962405, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 37448 + }, + { + "epoch": 1.9518195488721806, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 37449 + }, + { + "epoch": 1.9518796992481202, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 37450 + }, + { + "epoch": 1.9519398496240603, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 37451 + }, + { + "epoch": 1.952, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 37452 + }, + { + "epoch": 1.9520601503759398, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1408, + "step": 37453 + }, + { + "epoch": 1.9521203007518797, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 37454 + }, + { + "epoch": 1.9521804511278196, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 37455 + }, + { + "epoch": 1.9522406015037594, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 37456 + }, + { + "epoch": 1.952300751879699, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 37457 + }, + { + "epoch": 1.9523609022556392, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 37458 + }, + { + "epoch": 1.9524210526315788, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 37459 + }, + { + "epoch": 1.952481203007519, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 37460 + }, + { + "epoch": 1.9525413533834586, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 37461 + }, + { + "epoch": 1.9526015037593984, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 37462 + }, + { + "epoch": 1.9526616541353383, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 37463 + }, + { + "epoch": 1.9527218045112782, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 37464 + }, + { + "epoch": 1.952781954887218, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 37465 + }, + { + "epoch": 1.952842105263158, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2317, + "step": 37466 + }, + { + "epoch": 1.9529022556390978, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 37467 + }, + { + "epoch": 1.9529624060150375, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 37468 + }, + { + "epoch": 1.9530225563909775, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 37469 + }, + { + "epoch": 1.9530827067669172, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 37470 + }, + { + "epoch": 1.9531428571428573, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 37471 + }, + { + "epoch": 1.953203007518797, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 37472 + }, + { + "epoch": 1.9532631578947368, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 37473 + }, + { + "epoch": 1.9533233082706767, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 37474 + }, + { + "epoch": 1.9533834586466166, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 37475 + }, + { + "epoch": 1.9534436090225564, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.2406, + "step": 37476 + }, + { + "epoch": 1.953503759398496, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 37477 + }, + { + "epoch": 1.9535639097744362, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 37478 + }, + { + "epoch": 1.9536240601503758, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 37479 + }, + { + "epoch": 1.953684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 37480 + }, + { + "epoch": 1.9537443609022556, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 37481 + }, + { + "epoch": 1.9538045112781957, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 37482 + }, + { + "epoch": 1.9538646616541353, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 37483 + }, + { + "epoch": 1.9539248120300752, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 37484 + }, + { + "epoch": 1.953984962406015, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 37485 + }, + { + "epoch": 1.954045112781955, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 37486 + }, + { + "epoch": 1.9541052631578948, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 37487 + }, + { + "epoch": 1.9541654135338344, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 37488 + }, + { + "epoch": 1.9542255639097745, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 37489 + }, + { + "epoch": 1.9542857142857142, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 37490 + }, + { + "epoch": 1.9543458646616543, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 37491 + }, + { + "epoch": 1.954406015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 37492 + }, + { + "epoch": 1.9544661654135338, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 37493 + }, + { + "epoch": 1.9545263157894737, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 37494 + }, + { + "epoch": 1.9545864661654135, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 37495 + }, + { + "epoch": 1.9546466165413534, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 37496 + }, + { + "epoch": 1.9547067669172933, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 37497 + }, + { + "epoch": 1.9547669172932332, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 37498 + }, + { + "epoch": 1.9548270676691728, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 37499 + }, + { + "epoch": 1.954887218045113, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 37500 + }, + { + "epoch": 1.9549473684210525, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 37501 + }, + { + "epoch": 1.9550075187969926, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 37502 + }, + { + "epoch": 1.9550676691729323, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 37503 + }, + { + "epoch": 1.9551278195488722, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 37504 + }, + { + "epoch": 1.955187969924812, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 37505 + }, + { + "epoch": 1.955248120300752, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 37506 + }, + { + "epoch": 1.9553082706766918, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 37507 + }, + { + "epoch": 1.9553684210526314, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 37508 + }, + { + "epoch": 1.9554285714285715, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 37509 + }, + { + "epoch": 1.9554887218045112, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 37510 + }, + { + "epoch": 1.9555488721804513, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 37511 + }, + { + "epoch": 1.955609022556391, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.2248, + "step": 37512 + }, + { + "epoch": 1.9556691729323308, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 37513 + }, + { + "epoch": 1.9557293233082707, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 37514 + }, + { + "epoch": 1.9557894736842105, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 37515 + }, + { + "epoch": 1.9558496240601504, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 37516 + }, + { + "epoch": 1.9559097744360903, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 37517 + }, + { + "epoch": 1.9559699248120301, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 37518 + }, + { + "epoch": 1.9560300751879698, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 37519 + }, + { + "epoch": 1.9560902255639099, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 37520 + }, + { + "epoch": 1.9561503759398495, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 37521 + }, + { + "epoch": 1.9562105263157896, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 37522 + }, + { + "epoch": 1.9562706766917293, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 37523 + }, + { + "epoch": 1.9563308270676691, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 37524 + }, + { + "epoch": 1.956390977443609, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 37525 + }, + { + "epoch": 1.9564511278195489, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9331, + "step": 37526 + }, + { + "epoch": 1.9565112781954888, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 37527 + }, + { + "epoch": 1.9565714285714284, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 37528 + }, + { + "epoch": 1.9566315789473685, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 37529 + }, + { + "epoch": 1.9566917293233081, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 37530 + }, + { + "epoch": 1.9567518796992482, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 37531 + }, + { + "epoch": 1.956812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 37532 + }, + { + "epoch": 1.956872180451128, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 37533 + }, + { + "epoch": 1.9569323308270676, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 37534 + }, + { + "epoch": 1.9569924812030075, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 37535 + }, + { + "epoch": 1.9570526315789474, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 37536 + }, + { + "epoch": 1.9571127819548872, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 37537 + }, + { + "epoch": 1.9571729323308271, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 37538 + }, + { + "epoch": 1.9572330827067668, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 37539 + }, + { + "epoch": 1.9572932330827069, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 37540 + }, + { + "epoch": 1.9573533834586465, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 37541 + }, + { + "epoch": 1.9574135338345866, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 37542 + }, + { + "epoch": 1.9574736842105263, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 37543 + }, + { + "epoch": 1.9575338345864661, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 37544 + }, + { + "epoch": 1.957593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 37545 + }, + { + "epoch": 1.9576541353383459, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 37546 + }, + { + "epoch": 1.9577142857142857, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 37547 + }, + { + "epoch": 1.9577744360902256, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 37548 + }, + { + "epoch": 1.9578345864661655, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 37549 + }, + { + "epoch": 1.9578947368421051, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 37550 + }, + { + "epoch": 1.9579548872180452, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 37551 + }, + { + "epoch": 1.9580150375939849, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 37552 + }, + { + "epoch": 1.958075187969925, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 37553 + }, + { + "epoch": 1.9581353383458646, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 37554 + }, + { + "epoch": 1.9581954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 37555 + }, + { + "epoch": 1.9582556390977444, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 37556 + }, + { + "epoch": 1.9583157894736842, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 37557 + }, + { + "epoch": 1.958375939849624, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 37558 + }, + { + "epoch": 1.9584360902255638, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 37559 + }, + { + "epoch": 1.9584962406015038, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 37560 + }, + { + "epoch": 1.9585563909774435, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 37561 + }, + { + "epoch": 1.9586165413533836, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 37562 + }, + { + "epoch": 1.9586766917293232, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 37563 + }, + { + "epoch": 1.958736842105263, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 37564 + }, + { + "epoch": 1.958796992481203, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 37565 + }, + { + "epoch": 1.9588571428571429, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 37566 + }, + { + "epoch": 1.9589172932330827, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 37567 + }, + { + "epoch": 1.9589774436090226, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 37568 + }, + { + "epoch": 1.9590375939849625, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 37569 + }, + { + "epoch": 1.9590977443609021, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 37570 + }, + { + "epoch": 1.9591578947368422, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 37571 + }, + { + "epoch": 1.9592180451127819, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 37572 + }, + { + "epoch": 1.959278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 37573 + }, + { + "epoch": 1.9593383458646616, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 37574 + }, + { + "epoch": 1.9593984962406015, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 37575 + }, + { + "epoch": 1.9594586466165413, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 37576 + }, + { + "epoch": 1.9595187969924812, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 37577 + }, + { + "epoch": 1.959578947368421, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 37578 + }, + { + "epoch": 1.9596390977443607, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 37579 + }, + { + "epoch": 1.9596992481203008, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 37580 + }, + { + "epoch": 1.9597593984962405, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 37581 + }, + { + "epoch": 1.9598195488721806, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 37582 + }, + { + "epoch": 1.9598796992481202, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 37583 + }, + { + "epoch": 1.9599398496240603, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 37584 + }, + { + "epoch": 1.96, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 37585 + }, + { + "epoch": 1.9600601503759398, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 37586 + }, + { + "epoch": 1.9601203007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 37587 + }, + { + "epoch": 1.9601804511278196, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 37588 + }, + { + "epoch": 1.9602406015037594, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 37589 + }, + { + "epoch": 1.960300751879699, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9417, + "step": 37590 + }, + { + "epoch": 1.9603609022556392, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9043, + "step": 37591 + }, + { + "epoch": 1.9604210526315788, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 37592 + }, + { + "epoch": 1.960481203007519, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 37593 + }, + { + "epoch": 1.9605413533834586, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 37594 + }, + { + "epoch": 1.9606015037593985, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 37595 + }, + { + "epoch": 1.9606616541353383, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 37596 + }, + { + "epoch": 1.9607218045112782, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 37597 + }, + { + "epoch": 1.960781954887218, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 37598 + }, + { + "epoch": 1.960842105263158, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 37599 + }, + { + "epoch": 1.9609022556390978, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 37600 + }, + { + "epoch": 1.9609624060150375, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 37601 + }, + { + "epoch": 1.9610225563909776, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 37602 + }, + { + "epoch": 1.9610827067669172, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 37603 + }, + { + "epoch": 1.9611428571428573, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 37604 + }, + { + "epoch": 1.961203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 37605 + }, + { + "epoch": 1.9612631578947368, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 37606 + }, + { + "epoch": 1.9613233082706767, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 37607 + }, + { + "epoch": 1.9613834586466166, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 37608 + }, + { + "epoch": 1.9614436090225564, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 37609 + }, + { + "epoch": 1.961503759398496, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 37610 + }, + { + "epoch": 1.9615639097744362, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 37611 + }, + { + "epoch": 1.9616240601503758, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 37612 + }, + { + "epoch": 1.961684210526316, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 37613 + }, + { + "epoch": 1.9617443609022556, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 37614 + }, + { + "epoch": 1.9618045112781954, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 37615 + }, + { + "epoch": 1.9618646616541353, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 37616 + }, + { + "epoch": 1.9619248120300752, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 37617 + }, + { + "epoch": 1.961984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 37618 + }, + { + "epoch": 1.962045112781955, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 37619 + }, + { + "epoch": 1.9621052631578948, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 37620 + }, + { + "epoch": 1.9621654135338344, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 37621 + }, + { + "epoch": 1.9622255639097745, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 37622 + }, + { + "epoch": 1.9622857142857142, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 37623 + }, + { + "epoch": 1.9623458646616543, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 37624 + }, + { + "epoch": 1.962406015037594, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 37625 + }, + { + "epoch": 1.9624661654135338, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.146, + "step": 37626 + }, + { + "epoch": 1.9625263157894737, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 37627 + }, + { + "epoch": 1.9625864661654135, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 37628 + }, + { + "epoch": 1.9626466165413534, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 37629 + }, + { + "epoch": 1.9627067669172933, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 37630 + }, + { + "epoch": 1.9627669172932332, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 37631 + }, + { + "epoch": 1.9628270676691728, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 37632 + }, + { + "epoch": 1.962887218045113, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 37633 + }, + { + "epoch": 1.9629473684210526, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 37634 + }, + { + "epoch": 1.9630075187969926, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 37635 + }, + { + "epoch": 1.9630676691729323, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 37636 + }, + { + "epoch": 1.9631278195488722, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 37637 + }, + { + "epoch": 1.963187969924812, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 37638 + }, + { + "epoch": 1.963248120300752, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 37639 + }, + { + "epoch": 1.9633082706766918, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 37640 + }, + { + "epoch": 1.9633684210526314, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 37641 + }, + { + "epoch": 1.9634285714285715, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 37642 + }, + { + "epoch": 1.9634887218045112, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 37643 + }, + { + "epoch": 1.9635488721804513, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 37644 + }, + { + "epoch": 1.963609022556391, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9186, + "step": 37645 + }, + { + "epoch": 1.9636691729323308, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 37646 + }, + { + "epoch": 1.9637293233082707, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 37647 + }, + { + "epoch": 1.9637894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 37648 + }, + { + "epoch": 1.9638496240601504, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 37649 + }, + { + "epoch": 1.9639097744360903, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 37650 + }, + { + "epoch": 1.9639699248120301, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 37651 + }, + { + "epoch": 1.9640300751879698, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 37652 + }, + { + "epoch": 1.9640902255639099, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 37653 + }, + { + "epoch": 1.9641503759398495, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 37654 + }, + { + "epoch": 1.9642105263157896, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 37655 + }, + { + "epoch": 1.9642706766917293, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.168, + "step": 37656 + }, + { + "epoch": 1.9643308270676691, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 37657 + }, + { + "epoch": 1.964390977443609, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 37658 + }, + { + "epoch": 1.964451127819549, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 37659 + }, + { + "epoch": 1.9645112781954888, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 37660 + }, + { + "epoch": 1.9645714285714284, + "grad_norm": 1.8515625, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 37661 + }, + { + "epoch": 1.9646315789473685, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 37662 + }, + { + "epoch": 1.9646917293233082, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 37663 + }, + { + "epoch": 1.9647518796992482, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 37664 + }, + { + "epoch": 1.964812030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 37665 + }, + { + "epoch": 1.9648721804511278, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 37666 + }, + { + "epoch": 1.9649323308270676, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 37667 + }, + { + "epoch": 1.9649924812030075, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 37668 + }, + { + "epoch": 1.9650526315789474, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 37669 + }, + { + "epoch": 1.9651127819548873, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 37670 + }, + { + "epoch": 1.9651729323308271, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 37671 + }, + { + "epoch": 1.9652330827067668, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 37672 + }, + { + "epoch": 1.9652932330827069, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 37673 + }, + { + "epoch": 1.9653533834586465, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 37674 + }, + { + "epoch": 1.9654135338345866, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 37675 + }, + { + "epoch": 1.9654736842105263, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 37676 + }, + { + "epoch": 1.9655338345864661, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 37677 + }, + { + "epoch": 1.965593984962406, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 37678 + }, + { + "epoch": 1.9656541353383459, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 37679 + }, + { + "epoch": 1.9657142857142857, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 37680 + }, + { + "epoch": 1.9657744360902256, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 37681 + }, + { + "epoch": 1.9658345864661655, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 37682 + }, + { + "epoch": 1.9658947368421051, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 37683 + }, + { + "epoch": 1.9659548872180452, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 37684 + }, + { + "epoch": 1.9660150375939849, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 37685 + }, + { + "epoch": 1.966075187969925, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 37686 + }, + { + "epoch": 1.9661353383458646, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 37687 + }, + { + "epoch": 1.9661954887218045, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 37688 + }, + { + "epoch": 1.9662556390977444, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 37689 + }, + { + "epoch": 1.9663157894736842, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1694, + "step": 37690 + }, + { + "epoch": 1.966375939849624, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 37691 + }, + { + "epoch": 1.9664360902255638, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 37692 + }, + { + "epoch": 1.9664962406015039, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 37693 + }, + { + "epoch": 1.9665563909774435, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 37694 + }, + { + "epoch": 1.9666165413533836, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 37695 + }, + { + "epoch": 1.9666766917293232, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 37696 + }, + { + "epoch": 1.9667368421052631, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 37697 + }, + { + "epoch": 1.966796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1329, + "step": 37698 + }, + { + "epoch": 1.9668571428571429, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 37699 + }, + { + "epoch": 1.9669172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 37700 + }, + { + "epoch": 1.9669774436090226, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9501, + "step": 37701 + }, + { + "epoch": 1.9670375939849625, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 37702 + }, + { + "epoch": 1.9670977443609021, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 37703 + }, + { + "epoch": 1.9671578947368422, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 37704 + }, + { + "epoch": 1.9672180451127819, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 37705 + }, + { + "epoch": 1.967278195488722, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 37706 + }, + { + "epoch": 1.9673383458646616, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 37707 + }, + { + "epoch": 1.9673984962406015, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 37708 + }, + { + "epoch": 1.9674586466165414, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 37709 + }, + { + "epoch": 1.9675187969924812, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 37710 + }, + { + "epoch": 1.967578947368421, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 37711 + }, + { + "epoch": 1.9676390977443607, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 37712 + }, + { + "epoch": 1.9676992481203008, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 37713 + }, + { + "epoch": 1.9677593984962405, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 37714 + }, + { + "epoch": 1.9678195488721806, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 37715 + }, + { + "epoch": 1.9678796992481202, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 37716 + }, + { + "epoch": 1.96793984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 37717 + }, + { + "epoch": 1.968, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 37718 + }, + { + "epoch": 1.9680601503759398, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 37719 + }, + { + "epoch": 1.9681203007518797, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 37720 + }, + { + "epoch": 1.9681804511278196, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 37721 + }, + { + "epoch": 1.9682406015037595, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 37722 + }, + { + "epoch": 1.968300751879699, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 37723 + }, + { + "epoch": 1.9683609022556392, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.2487, + "step": 37724 + }, + { + "epoch": 1.9684210526315788, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 37725 + }, + { + "epoch": 1.968481203007519, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 37726 + }, + { + "epoch": 1.9685413533834586, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 37727 + }, + { + "epoch": 1.9686015037593985, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 37728 + }, + { + "epoch": 1.9686616541353383, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 37729 + }, + { + "epoch": 1.9687218045112782, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 37730 + }, + { + "epoch": 1.968781954887218, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 37731 + }, + { + "epoch": 1.968842105263158, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 37732 + }, + { + "epoch": 1.9689022556390978, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 37733 + }, + { + "epoch": 1.9689624060150375, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 37734 + }, + { + "epoch": 1.9690225563909776, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 37735 + }, + { + "epoch": 1.9690827067669172, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 37736 + }, + { + "epoch": 1.9691428571428573, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 37737 + }, + { + "epoch": 1.969203007518797, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 37738 + }, + { + "epoch": 1.9692631578947368, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 37739 + }, + { + "epoch": 1.9693233082706767, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 37740 + }, + { + "epoch": 1.9693834586466166, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 37741 + }, + { + "epoch": 1.9694436090225564, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 37742 + }, + { + "epoch": 1.969503759398496, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 37743 + }, + { + "epoch": 1.9695639097744362, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1662, + "step": 37744 + }, + { + "epoch": 1.9696240601503758, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 37745 + }, + { + "epoch": 1.969684210526316, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 37746 + }, + { + "epoch": 1.9697443609022556, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 37747 + }, + { + "epoch": 1.9698045112781954, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 37748 + }, + { + "epoch": 1.9698646616541353, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 37749 + }, + { + "epoch": 1.9699248120300752, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 37750 + }, + { + "epoch": 1.969984962406015, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 37751 + }, + { + "epoch": 1.970045112781955, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 37752 + }, + { + "epoch": 1.9701052631578948, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 37753 + }, + { + "epoch": 1.9701654135338345, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1303, + "step": 37754 + }, + { + "epoch": 1.9702255639097745, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 37755 + }, + { + "epoch": 1.9702857142857142, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 37756 + }, + { + "epoch": 1.9703458646616543, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 37757 + }, + { + "epoch": 1.970406015037594, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 37758 + }, + { + "epoch": 1.9704661654135338, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 37759 + }, + { + "epoch": 1.9705263157894737, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.3764, + "step": 37760 + }, + { + "epoch": 1.9705864661654136, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 37761 + }, + { + "epoch": 1.9706466165413534, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 37762 + }, + { + "epoch": 1.970706766917293, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 37763 + }, + { + "epoch": 1.9707669172932332, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 37764 + }, + { + "epoch": 1.9708270676691728, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 37765 + }, + { + "epoch": 1.970887218045113, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 37766 + }, + { + "epoch": 1.9709473684210526, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 37767 + }, + { + "epoch": 1.9710075187969924, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 37768 + }, + { + "epoch": 1.9710676691729323, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 37769 + }, + { + "epoch": 1.9711278195488722, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 37770 + }, + { + "epoch": 1.971187969924812, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2118, + "step": 37771 + }, + { + "epoch": 1.971248120300752, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 37772 + }, + { + "epoch": 1.9713082706766918, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 37773 + }, + { + "epoch": 1.9713684210526314, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 37774 + }, + { + "epoch": 1.9714285714285715, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 37775 + }, + { + "epoch": 1.9714887218045112, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 37776 + }, + { + "epoch": 1.9715488721804513, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 37777 + }, + { + "epoch": 1.971609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 37778 + }, + { + "epoch": 1.9716691729323308, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 37779 + }, + { + "epoch": 1.9717293233082707, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 37780 + }, + { + "epoch": 1.9717894736842105, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 37781 + }, + { + "epoch": 1.9718496240601504, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 37782 + }, + { + "epoch": 1.9719097744360903, + "grad_norm": 2.0625, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 37783 + }, + { + "epoch": 1.9719699248120302, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 37784 + }, + { + "epoch": 1.9720300751879698, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 37785 + }, + { + "epoch": 1.97209022556391, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 37786 + }, + { + "epoch": 1.9721503759398495, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 37787 + }, + { + "epoch": 1.9722105263157896, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 37788 + }, + { + "epoch": 1.9722706766917293, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 37789 + }, + { + "epoch": 1.9723308270676692, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 37790 + }, + { + "epoch": 1.972390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 37791 + }, + { + "epoch": 1.972451127819549, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 37792 + }, + { + "epoch": 1.9725112781954888, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 37793 + }, + { + "epoch": 1.9725714285714284, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 37794 + }, + { + "epoch": 1.9726315789473685, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 37795 + }, + { + "epoch": 1.9726917293233082, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9442, + "step": 37796 + }, + { + "epoch": 1.9727518796992483, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 37797 + }, + { + "epoch": 1.972812030075188, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 37798 + }, + { + "epoch": 1.9728721804511278, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 37799 + }, + { + "epoch": 1.9729323308270676, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 37800 + }, + { + "epoch": 1.9729924812030075, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9414, + "step": 37801 + }, + { + "epoch": 1.9730526315789474, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 37802 + }, + { + "epoch": 1.9731127819548873, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 37803 + }, + { + "epoch": 1.9731729323308271, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 37804 + }, + { + "epoch": 1.9732330827067668, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 37805 + }, + { + "epoch": 1.9732932330827069, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 37806 + }, + { + "epoch": 1.9733533834586465, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 37807 + }, + { + "epoch": 1.9734135338345866, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 37808 + }, + { + "epoch": 1.9734736842105263, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 37809 + }, + { + "epoch": 1.9735338345864661, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 37810 + }, + { + "epoch": 1.973593984962406, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 37811 + }, + { + "epoch": 1.9736541353383459, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 37812 + }, + { + "epoch": 1.9737142857142858, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 37813 + }, + { + "epoch": 1.9737744360902254, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 37814 + }, + { + "epoch": 1.9738345864661655, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 37815 + }, + { + "epoch": 1.9738947368421051, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 37816 + }, + { + "epoch": 1.9739548872180452, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 37817 + }, + { + "epoch": 1.974015037593985, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 37818 + }, + { + "epoch": 1.9740751879699248, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 37819 + }, + { + "epoch": 1.9741353383458646, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 37820 + }, + { + "epoch": 1.9741954887218045, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 37821 + }, + { + "epoch": 1.9742556390977444, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 37822 + }, + { + "epoch": 1.9743157894736842, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 37823 + }, + { + "epoch": 1.9743759398496241, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 37824 + }, + { + "epoch": 1.9744360902255638, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9505, + "step": 37825 + }, + { + "epoch": 1.9744962406015039, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 37826 + }, + { + "epoch": 1.9745563909774435, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 37827 + }, + { + "epoch": 1.9746165413533836, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 37828 + }, + { + "epoch": 1.9746766917293233, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 37829 + }, + { + "epoch": 1.9747368421052631, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 37830 + }, + { + "epoch": 1.974796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 37831 + }, + { + "epoch": 1.9748571428571429, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 37832 + }, + { + "epoch": 1.9749172932330827, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 37833 + }, + { + "epoch": 1.9749774436090226, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 37834 + }, + { + "epoch": 1.9750375939849625, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 37835 + }, + { + "epoch": 1.9750977443609021, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 37836 + }, + { + "epoch": 1.9751578947368422, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 37837 + }, + { + "epoch": 1.9752180451127819, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 37838 + }, + { + "epoch": 1.975278195488722, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 37839 + }, + { + "epoch": 1.9753383458646616, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 37840 + }, + { + "epoch": 1.9753984962406015, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 37841 + }, + { + "epoch": 1.9754586466165414, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 37842 + }, + { + "epoch": 1.9755187969924812, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1494, + "step": 37843 + }, + { + "epoch": 1.975578947368421, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 37844 + }, + { + "epoch": 1.9756390977443608, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 37845 + }, + { + "epoch": 1.9756992481203008, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 37846 + }, + { + "epoch": 1.9757593984962405, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 37847 + }, + { + "epoch": 1.9758195488721806, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 37848 + }, + { + "epoch": 1.9758796992481202, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 37849 + }, + { + "epoch": 1.97593984962406, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 37850 + }, + { + "epoch": 1.976, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 37851 + }, + { + "epoch": 1.9760601503759399, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 37852 + }, + { + "epoch": 1.9761203007518797, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 37853 + }, + { + "epoch": 1.9761804511278196, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 37854 + }, + { + "epoch": 1.9762406015037595, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 37855 + }, + { + "epoch": 1.9763007518796991, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 37856 + }, + { + "epoch": 1.9763609022556392, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 37857 + }, + { + "epoch": 1.9764210526315789, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 37858 + }, + { + "epoch": 1.976481203007519, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 37859 + }, + { + "epoch": 1.9765413533834586, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 37860 + }, + { + "epoch": 1.9766015037593985, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 37861 + }, + { + "epoch": 1.9766616541353383, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 37862 + }, + { + "epoch": 1.9767218045112782, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 37863 + }, + { + "epoch": 1.976781954887218, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 37864 + }, + { + "epoch": 1.9768421052631577, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 37865 + }, + { + "epoch": 1.9769022556390978, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 37866 + }, + { + "epoch": 1.9769624060150375, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 37867 + }, + { + "epoch": 1.9770225563909776, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 37868 + }, + { + "epoch": 1.9770827067669172, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9372, + "step": 37869 + }, + { + "epoch": 1.977142857142857, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 37870 + }, + { + "epoch": 1.977203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 37871 + }, + { + "epoch": 1.9772631578947368, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 37872 + }, + { + "epoch": 1.9773233082706767, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 37873 + }, + { + "epoch": 1.9773834586466166, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 37874 + }, + { + "epoch": 1.9774436090225564, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 37875 + }, + { + "epoch": 1.977503759398496, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 37876 + }, + { + "epoch": 1.9775639097744362, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 37877 + }, + { + "epoch": 1.9776240601503758, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.2647, + "step": 37878 + }, + { + "epoch": 1.977684210526316, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 37879 + }, + { + "epoch": 1.9777443609022556, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 37880 + }, + { + "epoch": 1.9778045112781955, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 37881 + }, + { + "epoch": 1.9778646616541353, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 37882 + }, + { + "epoch": 1.9779248120300752, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 37883 + }, + { + "epoch": 1.977984962406015, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 37884 + }, + { + "epoch": 1.978045112781955, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 37885 + }, + { + "epoch": 1.9781052631578948, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 37886 + }, + { + "epoch": 1.9781654135338345, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 37887 + }, + { + "epoch": 1.9782255639097746, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 37888 + }, + { + "epoch": 1.9782857142857142, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 37889 + }, + { + "epoch": 1.9783458646616543, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 37890 + }, + { + "epoch": 1.978406015037594, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 37891 + }, + { + "epoch": 1.9784661654135338, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 37892 + }, + { + "epoch": 1.9785263157894737, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 37893 + }, + { + "epoch": 1.9785864661654136, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 37894 + }, + { + "epoch": 1.9786466165413534, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 37895 + }, + { + "epoch": 1.978706766917293, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.8829, + "step": 37896 + }, + { + "epoch": 1.9787669172932332, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 37897 + }, + { + "epoch": 1.9788270676691728, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 37898 + }, + { + "epoch": 1.978887218045113, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 37899 + }, + { + "epoch": 1.9789473684210526, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 37900 + }, + { + "epoch": 1.9790075187969924, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 37901 + }, + { + "epoch": 1.9790676691729323, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 37902 + }, + { + "epoch": 1.9791278195488722, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 37903 + }, + { + "epoch": 1.979187969924812, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 37904 + }, + { + "epoch": 1.979248120300752, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 37905 + }, + { + "epoch": 1.9793082706766918, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 37906 + }, + { + "epoch": 1.9793684210526314, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 37907 + }, + { + "epoch": 1.9794285714285715, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 37908 + }, + { + "epoch": 1.9794887218045112, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 37909 + }, + { + "epoch": 1.9795488721804513, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 37910 + }, + { + "epoch": 1.979609022556391, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 37911 + }, + { + "epoch": 1.9796691729323308, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 37912 + }, + { + "epoch": 1.9797293233082707, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 37913 + }, + { + "epoch": 1.9797894736842105, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 37914 + }, + { + "epoch": 1.9798496240601504, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 37915 + }, + { + "epoch": 1.97990977443609, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 37916 + }, + { + "epoch": 1.9799699248120302, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 37917 + }, + { + "epoch": 1.9800300751879698, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 37918 + }, + { + "epoch": 1.98009022556391, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 37919 + }, + { + "epoch": 1.9801503759398495, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 37920 + }, + { + "epoch": 1.9802105263157894, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 37921 + }, + { + "epoch": 1.9802706766917293, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 37922 + }, + { + "epoch": 1.9803308270676692, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 37923 + }, + { + "epoch": 1.980390977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 37924 + }, + { + "epoch": 1.980451127819549, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 37925 + }, + { + "epoch": 1.9805112781954888, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 37926 + }, + { + "epoch": 1.9805714285714284, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 37927 + }, + { + "epoch": 1.9806315789473685, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 37928 + }, + { + "epoch": 1.9806917293233082, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 37929 + }, + { + "epoch": 1.9807518796992483, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 37930 + }, + { + "epoch": 1.980812030075188, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 37931 + }, + { + "epoch": 1.9808721804511278, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 37932 + }, + { + "epoch": 1.9809323308270677, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 37933 + }, + { + "epoch": 1.9809924812030075, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 37934 + }, + { + "epoch": 1.9810526315789474, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 37935 + }, + { + "epoch": 1.9811127819548873, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 37936 + }, + { + "epoch": 1.9811729323308271, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 37937 + }, + { + "epoch": 1.9812330827067668, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 37938 + }, + { + "epoch": 1.9812932330827069, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 37939 + }, + { + "epoch": 1.9813533834586465, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 37940 + }, + { + "epoch": 1.9814135338345866, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 37941 + }, + { + "epoch": 1.9814736842105263, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 37942 + }, + { + "epoch": 1.9815338345864661, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 37943 + }, + { + "epoch": 1.981593984962406, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 37944 + }, + { + "epoch": 1.981654135338346, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 37945 + }, + { + "epoch": 1.9817142857142858, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 37946 + }, + { + "epoch": 1.9817744360902254, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 37947 + }, + { + "epoch": 1.9818345864661655, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 37948 + }, + { + "epoch": 1.9818947368421052, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9273, + "step": 37949 + }, + { + "epoch": 1.9819548872180452, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 37950 + }, + { + "epoch": 1.982015037593985, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 37951 + }, + { + "epoch": 1.9820751879699248, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 37952 + }, + { + "epoch": 1.9821353383458646, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 37953 + }, + { + "epoch": 1.9821954887218045, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 37954 + }, + { + "epoch": 1.9822556390977444, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 37955 + }, + { + "epoch": 1.9823157894736843, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 37956 + }, + { + "epoch": 1.9823759398496241, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 37957 + }, + { + "epoch": 1.9824360902255638, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1382, + "step": 37958 + }, + { + "epoch": 1.9824962406015039, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.155, + "step": 37959 + }, + { + "epoch": 1.9825563909774435, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 37960 + }, + { + "epoch": 1.9826165413533836, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 37961 + }, + { + "epoch": 1.9826766917293233, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 37962 + }, + { + "epoch": 1.9827368421052631, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 37963 + }, + { + "epoch": 1.982796992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 37964 + }, + { + "epoch": 1.9828571428571429, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 37965 + }, + { + "epoch": 1.9829172932330827, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 37966 + }, + { + "epoch": 1.9829774436090224, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 37967 + }, + { + "epoch": 1.9830375939849625, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 37968 + }, + { + "epoch": 1.9830977443609021, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 37969 + }, + { + "epoch": 1.9831578947368422, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 37970 + }, + { + "epoch": 1.9832180451127819, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 37971 + }, + { + "epoch": 1.9832781954887218, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 37972 + }, + { + "epoch": 1.9833383458646616, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9181, + "step": 37973 + }, + { + "epoch": 1.9833984962406015, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 37974 + }, + { + "epoch": 1.9834586466165414, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 37975 + }, + { + "epoch": 1.9835187969924812, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 37976 + }, + { + "epoch": 1.983578947368421, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 37977 + }, + { + "epoch": 1.9836390977443608, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 37978 + }, + { + "epoch": 1.9836992481203009, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 37979 + }, + { + "epoch": 1.9837593984962405, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 37980 + }, + { + "epoch": 1.9838195488721806, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 37981 + }, + { + "epoch": 1.9838796992481202, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 37982 + }, + { + "epoch": 1.9839398496240601, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 37983 + }, + { + "epoch": 1.984, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 37984 + }, + { + "epoch": 1.9840601503759399, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 37985 + }, + { + "epoch": 1.9841203007518797, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 37986 + }, + { + "epoch": 1.9841804511278196, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 37987 + }, + { + "epoch": 1.9842406015037595, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 37988 + }, + { + "epoch": 1.9843007518796991, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9052, + "step": 37989 + }, + { + "epoch": 1.9843609022556392, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 37990 + }, + { + "epoch": 1.9844210526315789, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 37991 + }, + { + "epoch": 1.984481203007519, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 37992 + }, + { + "epoch": 1.9845413533834586, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 37993 + }, + { + "epoch": 1.9846015037593985, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 37994 + }, + { + "epoch": 1.9846616541353383, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.136, + "step": 37995 + }, + { + "epoch": 1.9847218045112782, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 37996 + }, + { + "epoch": 1.984781954887218, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 37997 + }, + { + "epoch": 1.9848421052631577, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 37998 + }, + { + "epoch": 1.9849022556390978, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9504, + "step": 37999 + }, + { + "epoch": 1.9849624060150375, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 38000 + }, + { + "epoch": 1.9850225563909776, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 38001 + }, + { + "epoch": 1.9850827067669172, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 38002 + }, + { + "epoch": 1.985142857142857, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 38003 + }, + { + "epoch": 1.985203007518797, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 38004 + }, + { + "epoch": 1.9852631578947368, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 38005 + }, + { + "epoch": 1.9853233082706767, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 38006 + }, + { + "epoch": 1.9853834586466166, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 38007 + }, + { + "epoch": 1.9854436090225565, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 38008 + }, + { + "epoch": 1.985503759398496, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 38009 + }, + { + "epoch": 1.9855639097744362, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 38010 + }, + { + "epoch": 1.9856240601503758, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 38011 + }, + { + "epoch": 1.985684210526316, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 38012 + }, + { + "epoch": 1.9857443609022556, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 38013 + }, + { + "epoch": 1.9858045112781955, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 38014 + }, + { + "epoch": 1.9858646616541353, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 38015 + }, + { + "epoch": 1.9859248120300752, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 38016 + }, + { + "epoch": 1.985984962406015, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 38017 + }, + { + "epoch": 1.9860451127819547, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 38018 + }, + { + "epoch": 1.9861052631578948, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 38019 + }, + { + "epoch": 1.9861654135338345, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 38020 + }, + { + "epoch": 1.9862255639097746, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 38021 + }, + { + "epoch": 1.9862857142857142, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 38022 + }, + { + "epoch": 1.986345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 38023 + }, + { + "epoch": 1.986406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 38024 + }, + { + "epoch": 1.9864661654135338, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 38025 + }, + { + "epoch": 1.9865263157894737, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 38026 + }, + { + "epoch": 1.9865864661654136, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 38027 + }, + { + "epoch": 1.9866466165413534, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 38028 + }, + { + "epoch": 1.986706766917293, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1827, + "step": 38029 + }, + { + "epoch": 1.9867669172932332, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 38030 + }, + { + "epoch": 1.9868270676691728, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 38031 + }, + { + "epoch": 1.986887218045113, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 38032 + }, + { + "epoch": 1.9869473684210526, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 38033 + }, + { + "epoch": 1.9870075187969924, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 38034 + }, + { + "epoch": 1.9870676691729323, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 38035 + }, + { + "epoch": 1.9871278195488722, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 38036 + }, + { + "epoch": 1.987187969924812, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 38037 + }, + { + "epoch": 1.987248120300752, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 38038 + }, + { + "epoch": 1.9873082706766918, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 38039 + }, + { + "epoch": 1.9873684210526315, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 38040 + }, + { + "epoch": 1.9874285714285715, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 38041 + }, + { + "epoch": 1.9874887218045112, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 38042 + }, + { + "epoch": 1.9875488721804513, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 38043 + }, + { + "epoch": 1.987609022556391, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 38044 + }, + { + "epoch": 1.9876691729323308, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 38045 + }, + { + "epoch": 1.9877293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 38046 + }, + { + "epoch": 1.9877894736842106, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 38047 + }, + { + "epoch": 1.9878496240601504, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 38048 + }, + { + "epoch": 1.98790977443609, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 38049 + }, + { + "epoch": 1.9879699248120302, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 38050 + }, + { + "epoch": 1.9880300751879698, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 38051 + }, + { + "epoch": 1.98809022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 38052 + }, + { + "epoch": 1.9881503759398496, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 38053 + }, + { + "epoch": 1.9882105263157894, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 38054 + }, + { + "epoch": 1.9882706766917293, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 38055 + }, + { + "epoch": 1.9883308270676692, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 38056 + }, + { + "epoch": 1.988390977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 38057 + }, + { + "epoch": 1.988451127819549, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 38058 + }, + { + "epoch": 1.9885112781954888, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 38059 + }, + { + "epoch": 1.9885714285714284, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 38060 + }, + { + "epoch": 1.9886315789473685, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 38061 + }, + { + "epoch": 1.9886917293233082, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 38062 + }, + { + "epoch": 1.9887518796992483, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 38063 + }, + { + "epoch": 1.988812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 38064 + }, + { + "epoch": 1.9888721804511278, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 38065 + }, + { + "epoch": 1.9889323308270677, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 38066 + }, + { + "epoch": 1.9889924812030075, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 38067 + }, + { + "epoch": 1.9890526315789474, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 38068 + }, + { + "epoch": 1.989112781954887, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 38069 + }, + { + "epoch": 1.9891729323308271, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 38070 + }, + { + "epoch": 1.9892330827067668, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 38071 + }, + { + "epoch": 1.989293233082707, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 38072 + }, + { + "epoch": 1.9893533834586465, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 38073 + }, + { + "epoch": 1.9894135338345864, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 38074 + }, + { + "epoch": 1.9894736842105263, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 38075 + }, + { + "epoch": 1.9895338345864662, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 38076 + }, + { + "epoch": 1.989593984962406, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 38077 + }, + { + "epoch": 1.989654135338346, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 38078 + }, + { + "epoch": 1.9897142857142858, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 38079 + }, + { + "epoch": 1.9897744360902254, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 38080 + }, + { + "epoch": 1.9898345864661655, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 38081 + }, + { + "epoch": 1.9898947368421052, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 38082 + }, + { + "epoch": 1.9899548872180453, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 38083 + }, + { + "epoch": 1.990015037593985, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 38084 + }, + { + "epoch": 1.9900751879699248, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 38085 + }, + { + "epoch": 1.9901353383458646, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 38086 + }, + { + "epoch": 1.9901954887218045, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 38087 + }, + { + "epoch": 1.9902556390977444, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 38088 + }, + { + "epoch": 1.9903157894736843, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 38089 + }, + { + "epoch": 1.9903759398496241, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 38090 + }, + { + "epoch": 1.9904360902255638, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 38091 + }, + { + "epoch": 1.9904962406015039, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 38092 + }, + { + "epoch": 1.9905563909774435, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 38093 + }, + { + "epoch": 1.9906165413533836, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 38094 + }, + { + "epoch": 1.9906766917293233, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 38095 + }, + { + "epoch": 1.9907368421052631, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 38096 + }, + { + "epoch": 1.990796992481203, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 38097 + }, + { + "epoch": 1.9908571428571429, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 38098 + }, + { + "epoch": 1.9909172932330828, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 38099 + }, + { + "epoch": 1.9909774436090224, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 38100 + }, + { + "epoch": 1.9910375939849625, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2116, + "step": 38101 + }, + { + "epoch": 1.9910977443609021, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 38102 + }, + { + "epoch": 1.9911578947368422, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2271, + "step": 38103 + }, + { + "epoch": 1.9912180451127819, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 38104 + }, + { + "epoch": 1.9912781954887218, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 38105 + }, + { + "epoch": 1.9913383458646616, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 38106 + }, + { + "epoch": 1.9913984962406015, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 38107 + }, + { + "epoch": 1.9914586466165414, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 38108 + }, + { + "epoch": 1.9915187969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 38109 + }, + { + "epoch": 1.9915789473684211, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 38110 + }, + { + "epoch": 1.9916390977443608, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 38111 + }, + { + "epoch": 1.9916992481203009, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 38112 + }, + { + "epoch": 1.9917593984962405, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 38113 + }, + { + "epoch": 1.9918195488721806, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 38114 + }, + { + "epoch": 1.9918796992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 38115 + }, + { + "epoch": 1.9919398496240601, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 38116 + }, + { + "epoch": 1.992, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 38117 + }, + { + "epoch": 1.9920601503759399, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 38118 + }, + { + "epoch": 1.9921203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 38119 + }, + { + "epoch": 1.9921804511278194, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 38120 + }, + { + "epoch": 1.9922406015037595, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 38121 + }, + { + "epoch": 1.9923007518796991, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 38122 + }, + { + "epoch": 1.9923609022556392, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 38123 + }, + { + "epoch": 1.9924210526315789, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 38124 + }, + { + "epoch": 1.9924812030075187, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 38125 + }, + { + "epoch": 1.9925413533834586, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 38126 + }, + { + "epoch": 1.9926015037593985, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 38127 + }, + { + "epoch": 1.9926616541353384, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1411, + "step": 38128 + }, + { + "epoch": 1.9927218045112782, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 38129 + }, + { + "epoch": 1.992781954887218, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 38130 + }, + { + "epoch": 1.9928421052631577, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 38131 + }, + { + "epoch": 1.9929022556390978, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 38132 + }, + { + "epoch": 1.9929624060150375, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 38133 + }, + { + "epoch": 1.9930225563909776, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 38134 + }, + { + "epoch": 1.9930827067669172, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 38135 + }, + { + "epoch": 1.993142857142857, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 38136 + }, + { + "epoch": 1.993203007518797, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 38137 + }, + { + "epoch": 1.9932631578947368, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2202, + "step": 38138 + }, + { + "epoch": 1.9933233082706767, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 38139 + }, + { + "epoch": 1.9933834586466166, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 38140 + }, + { + "epoch": 1.9934436090225565, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 38141 + }, + { + "epoch": 1.9935037593984961, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 38142 + }, + { + "epoch": 1.9935639097744362, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 38143 + }, + { + "epoch": 1.9936240601503759, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9387, + "step": 38144 + }, + { + "epoch": 1.993684210526316, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 38145 + }, + { + "epoch": 1.9937443609022556, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9487, + "step": 38146 + }, + { + "epoch": 1.9938045112781955, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 38147 + }, + { + "epoch": 1.9938646616541353, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1177, + "step": 38148 + }, + { + "epoch": 1.9939248120300752, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 38149 + }, + { + "epoch": 1.993984962406015, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 38150 + }, + { + "epoch": 1.9940451127819547, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 38151 + }, + { + "epoch": 1.9941052631578948, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 38152 + }, + { + "epoch": 1.9941654135338345, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 38153 + }, + { + "epoch": 1.9942255639097746, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 38154 + }, + { + "epoch": 1.9942857142857142, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 38155 + }, + { + "epoch": 1.994345864661654, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 38156 + }, + { + "epoch": 1.994406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 38157 + }, + { + "epoch": 1.9944661654135338, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 38158 + }, + { + "epoch": 1.9945263157894737, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 38159 + }, + { + "epoch": 1.9945864661654136, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 38160 + }, + { + "epoch": 1.9946466165413534, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 38161 + }, + { + "epoch": 1.994706766917293, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 38162 + }, + { + "epoch": 1.9947669172932332, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 38163 + }, + { + "epoch": 1.9948270676691728, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 38164 + }, + { + "epoch": 1.994887218045113, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 38165 + }, + { + "epoch": 1.9949473684210526, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 38166 + }, + { + "epoch": 1.9950075187969925, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 38167 + }, + { + "epoch": 1.9950676691729323, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 38168 + }, + { + "epoch": 1.9951278195488722, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 38169 + }, + { + "epoch": 1.995187969924812, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 38170 + }, + { + "epoch": 1.9952481203007517, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 38171 + }, + { + "epoch": 1.9953082706766918, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.935, + "step": 38172 + }, + { + "epoch": 1.9953684210526315, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 38173 + }, + { + "epoch": 1.9954285714285716, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 38174 + }, + { + "epoch": 1.9954887218045112, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 38175 + }, + { + "epoch": 1.9955488721804513, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 38176 + }, + { + "epoch": 1.995609022556391, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 38177 + }, + { + "epoch": 1.9956691729323308, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 38178 + }, + { + "epoch": 1.9957293233082707, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9374, + "step": 38179 + }, + { + "epoch": 1.9957894736842106, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 38180 + }, + { + "epoch": 1.9958496240601504, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 38181 + }, + { + "epoch": 1.99590977443609, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 38182 + }, + { + "epoch": 1.9959699248120302, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 38183 + }, + { + "epoch": 1.9960300751879698, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 38184 + }, + { + "epoch": 1.99609022556391, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1399, + "step": 38185 + }, + { + "epoch": 1.9961503759398496, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 38186 + }, + { + "epoch": 1.9962105263157894, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 38187 + }, + { + "epoch": 1.9962706766917293, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 38188 + }, + { + "epoch": 1.9963308270676692, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9429, + "step": 38189 + }, + { + "epoch": 1.996390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 38190 + }, + { + "epoch": 1.996451127819549, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 38191 + }, + { + "epoch": 1.9965112781954888, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 38192 + }, + { + "epoch": 1.9965714285714284, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 38193 + }, + { + "epoch": 1.9966315789473685, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 38194 + }, + { + "epoch": 1.9966917293233082, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 38195 + }, + { + "epoch": 1.9967518796992483, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 38196 + }, + { + "epoch": 1.996812030075188, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 38197 + }, + { + "epoch": 1.9968721804511278, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 38198 + }, + { + "epoch": 1.9969323308270677, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 38199 + }, + { + "epoch": 1.9969924812030075, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 38200 + }, + { + "epoch": 1.9970526315789474, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 38201 + }, + { + "epoch": 1.997112781954887, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 38202 + }, + { + "epoch": 1.9971729323308272, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 38203 + }, + { + "epoch": 1.9972330827067668, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 38204 + }, + { + "epoch": 1.997293233082707, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 38205 + }, + { + "epoch": 1.9973533834586465, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 38206 + }, + { + "epoch": 1.9974135338345864, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 38207 + }, + { + "epoch": 1.9974736842105263, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 38208 + }, + { + "epoch": 1.9975338345864662, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 38209 + }, + { + "epoch": 1.997593984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 38210 + }, + { + "epoch": 1.997654135338346, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 38211 + }, + { + "epoch": 1.9977142857142858, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 38212 + }, + { + "epoch": 1.9977744360902254, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 38213 + }, + { + "epoch": 1.9978345864661655, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 38214 + }, + { + "epoch": 1.9978947368421052, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 38215 + }, + { + "epoch": 1.9979548872180453, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 38216 + }, + { + "epoch": 1.998015037593985, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 38217 + }, + { + "epoch": 1.9980751879699248, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 38218 + }, + { + "epoch": 1.9981353383458647, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 38219 + }, + { + "epoch": 1.9981954887218045, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 38220 + }, + { + "epoch": 1.9982556390977444, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 38221 + }, + { + "epoch": 1.998315789473684, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 38222 + }, + { + "epoch": 1.9983759398496241, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 38223 + }, + { + "epoch": 1.9984360902255638, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 38224 + }, + { + "epoch": 1.9984962406015039, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 38225 + }, + { + "epoch": 1.9985563909774435, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 38226 + }, + { + "epoch": 1.9986165413533836, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 38227 + }, + { + "epoch": 1.9986766917293233, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 38228 + }, + { + "epoch": 1.9987368421052631, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 38229 + }, + { + "epoch": 1.998796992481203, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 38230 + }, + { + "epoch": 1.9988571428571429, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 38231 + }, + { + "epoch": 1.9989172932330828, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 38232 + }, + { + "epoch": 1.9989774436090224, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 38233 + }, + { + "epoch": 1.9990375939849625, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 38234 + }, + { + "epoch": 1.9990977443609022, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 38235 + }, + { + "epoch": 1.9991578947368422, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 38236 + }, + { + "epoch": 1.999218045112782, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 38237 + }, + { + "epoch": 1.9992781954887218, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 38238 + }, + { + "epoch": 1.9993383458646616, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 38239 + }, + { + "epoch": 1.9993984962406015, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.145, + "step": 38240 + }, + { + "epoch": 1.9994586466165414, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 38241 + }, + { + "epoch": 1.9995187969924813, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 38242 + }, + { + "epoch": 1.9995789473684211, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 38243 + }, + { + "epoch": 1.9996390977443608, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 38244 + }, + { + "epoch": 1.9996992481203009, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 38245 + }, + { + "epoch": 1.9997593984962405, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 38246 + }, + { + "epoch": 1.9998195488721806, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 38247 + }, + { + "epoch": 1.9998796992481203, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 38248 + }, + { + "epoch": 1.9999398496240601, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 38249 + }, + { + "epoch": 2.0, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1518, + "step": 38250 + }, + { + "epoch": 2.0000601503759396, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 38251 + }, + { + "epoch": 2.0001203007518797, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 38252 + }, + { + "epoch": 2.0001804511278194, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9529, + "step": 38253 + }, + { + "epoch": 2.0002406015037595, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 38254 + }, + { + "epoch": 2.000300751879699, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 38255 + }, + { + "epoch": 2.0003609022556392, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 38256 + }, + { + "epoch": 2.000421052631579, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 38257 + }, + { + "epoch": 2.000481203007519, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 38258 + }, + { + "epoch": 2.0005413533834586, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 38259 + }, + { + "epoch": 2.0006015037593987, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 38260 + }, + { + "epoch": 2.0006616541353384, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.879, + "step": 38261 + }, + { + "epoch": 2.000721804511278, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1654, + "step": 38262 + }, + { + "epoch": 2.000781954887218, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 38263 + }, + { + "epoch": 2.0008421052631578, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 38264 + }, + { + "epoch": 2.000902255639098, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 38265 + }, + { + "epoch": 2.0009624060150375, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1963, + "step": 38266 + }, + { + "epoch": 2.0010225563909776, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 38267 + }, + { + "epoch": 2.0010827067669172, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 38268 + }, + { + "epoch": 2.0011428571428573, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9354, + "step": 38269 + }, + { + "epoch": 2.001203007518797, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 38270 + }, + { + "epoch": 2.0012631578947366, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 38271 + }, + { + "epoch": 2.0013233082706767, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 38272 + }, + { + "epoch": 2.0013834586466164, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 38273 + }, + { + "epoch": 2.0014436090225565, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 38274 + }, + { + "epoch": 2.001503759398496, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 38275 + }, + { + "epoch": 2.001563909774436, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 38276 + }, + { + "epoch": 2.001624060150376, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 38277 + }, + { + "epoch": 2.001684210526316, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 38278 + }, + { + "epoch": 2.0017443609022556, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 38279 + }, + { + "epoch": 2.0018045112781957, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 38280 + }, + { + "epoch": 2.0018646616541353, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 38281 + }, + { + "epoch": 2.001924812030075, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 38282 + }, + { + "epoch": 2.001984962406015, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 38283 + }, + { + "epoch": 2.0020451127819547, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 38284 + }, + { + "epoch": 2.002105263157895, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 38285 + }, + { + "epoch": 2.0021654135338345, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 38286 + }, + { + "epoch": 2.0022255639097746, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 38287 + }, + { + "epoch": 2.0022857142857142, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 38288 + }, + { + "epoch": 2.0023458646616543, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 38289 + }, + { + "epoch": 2.002406015037594, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 38290 + }, + { + "epoch": 2.0024661654135336, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 38291 + }, + { + "epoch": 2.0025263157894737, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 38292 + }, + { + "epoch": 2.0025864661654134, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 38293 + }, + { + "epoch": 2.0026466165413535, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 38294 + }, + { + "epoch": 2.002706766917293, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 38295 + }, + { + "epoch": 2.002766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 38296 + }, + { + "epoch": 2.002827067669173, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1364, + "step": 38297 + }, + { + "epoch": 2.002887218045113, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 38298 + }, + { + "epoch": 2.0029473684210526, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 38299 + }, + { + "epoch": 2.0030075187969927, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 38300 + }, + { + "epoch": 2.0030676691729323, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 38301 + }, + { + "epoch": 2.003127819548872, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 38302 + }, + { + "epoch": 2.003187969924812, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 38303 + }, + { + "epoch": 2.0032481203007517, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 38304 + }, + { + "epoch": 2.003308270676692, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 38305 + }, + { + "epoch": 2.0033684210526315, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 38306 + }, + { + "epoch": 2.0034285714285716, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 38307 + }, + { + "epoch": 2.003488721804511, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 38308 + }, + { + "epoch": 2.0035488721804513, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 38309 + }, + { + "epoch": 2.003609022556391, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 38310 + }, + { + "epoch": 2.003669172932331, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 38311 + }, + { + "epoch": 2.0037293233082707, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 38312 + }, + { + "epoch": 2.0037894736842103, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 38313 + }, + { + "epoch": 2.0038496240601504, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 38314 + }, + { + "epoch": 2.00390977443609, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 38315 + }, + { + "epoch": 2.00396992481203, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1515, + "step": 38316 + }, + { + "epoch": 2.00403007518797, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 38317 + }, + { + "epoch": 2.00409022556391, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 38318 + }, + { + "epoch": 2.0041503759398496, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 38319 + }, + { + "epoch": 2.0042105263157897, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 38320 + }, + { + "epoch": 2.0042706766917293, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9434, + "step": 38321 + }, + { + "epoch": 2.004330827067669, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 38322 + }, + { + "epoch": 2.004390977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1658, + "step": 38323 + }, + { + "epoch": 2.0044511278195487, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 38324 + }, + { + "epoch": 2.004511278195489, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 38325 + }, + { + "epoch": 2.0045714285714284, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 38326 + }, + { + "epoch": 2.0046315789473685, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 38327 + }, + { + "epoch": 2.004691729323308, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 38328 + }, + { + "epoch": 2.0047518796992483, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 38329 + }, + { + "epoch": 2.004812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 38330 + }, + { + "epoch": 2.004872180451128, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 38331 + }, + { + "epoch": 2.0049323308270677, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 38332 + }, + { + "epoch": 2.0049924812030073, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 38333 + }, + { + "epoch": 2.0050526315789474, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 38334 + }, + { + "epoch": 2.005112781954887, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 38335 + }, + { + "epoch": 2.005172932330827, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 38336 + }, + { + "epoch": 2.005233082706767, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 38337 + }, + { + "epoch": 2.005293233082707, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 38338 + }, + { + "epoch": 2.0053533834586466, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 38339 + }, + { + "epoch": 2.0054135338345866, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 38340 + }, + { + "epoch": 2.0054736842105263, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 38341 + }, + { + "epoch": 2.005533834586466, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 38342 + }, + { + "epoch": 2.005593984962406, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 38343 + }, + { + "epoch": 2.0056541353383457, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1366, + "step": 38344 + }, + { + "epoch": 2.005714285714286, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 38345 + }, + { + "epoch": 2.0057744360902254, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 38346 + }, + { + "epoch": 2.0058345864661655, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 38347 + }, + { + "epoch": 2.005894736842105, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 38348 + }, + { + "epoch": 2.0059548872180453, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 38349 + }, + { + "epoch": 2.006015037593985, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 38350 + }, + { + "epoch": 2.006075187969925, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 38351 + }, + { + "epoch": 2.0061353383458647, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 38352 + }, + { + "epoch": 2.0061954887218043, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9443, + "step": 38353 + }, + { + "epoch": 2.0062556390977444, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 38354 + }, + { + "epoch": 2.006315789473684, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 38355 + }, + { + "epoch": 2.006375939849624, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 38356 + }, + { + "epoch": 2.006436090225564, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 38357 + }, + { + "epoch": 2.006496240601504, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 38358 + }, + { + "epoch": 2.0065563909774435, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 38359 + }, + { + "epoch": 2.0066165413533836, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 38360 + }, + { + "epoch": 2.0066766917293233, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 38361 + }, + { + "epoch": 2.0067368421052634, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 38362 + }, + { + "epoch": 2.006796992481203, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 38363 + }, + { + "epoch": 2.0068571428571427, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 38364 + }, + { + "epoch": 2.0069172932330828, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 38365 + }, + { + "epoch": 2.0069774436090224, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 38366 + }, + { + "epoch": 2.0070375939849625, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1752, + "step": 38367 + }, + { + "epoch": 2.007097744360902, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 38368 + }, + { + "epoch": 2.0071578947368423, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 38369 + }, + { + "epoch": 2.007218045112782, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 38370 + }, + { + "epoch": 2.007278195488722, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 38371 + }, + { + "epoch": 2.0073383458646616, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 38372 + }, + { + "epoch": 2.0073984962406013, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 38373 + }, + { + "epoch": 2.0074586466165414, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 38374 + }, + { + "epoch": 2.007518796992481, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 38375 + }, + { + "epoch": 2.007578947368421, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 38376 + }, + { + "epoch": 2.007639097744361, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 38377 + }, + { + "epoch": 2.007699248120301, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 38378 + }, + { + "epoch": 2.0077593984962405, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 38379 + }, + { + "epoch": 2.0078195488721806, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 38380 + }, + { + "epoch": 2.0078796992481203, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 38381 + }, + { + "epoch": 2.0079398496240604, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 38382 + }, + { + "epoch": 2.008, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 38383 + }, + { + "epoch": 2.0080601503759397, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 38384 + }, + { + "epoch": 2.0081203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 38385 + }, + { + "epoch": 2.0081804511278194, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 38386 + }, + { + "epoch": 2.0082406015037595, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 38387 + }, + { + "epoch": 2.008300751879699, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 38388 + }, + { + "epoch": 2.0083609022556392, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 38389 + }, + { + "epoch": 2.008421052631579, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 38390 + }, + { + "epoch": 2.008481203007519, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 38391 + }, + { + "epoch": 2.0085413533834586, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 38392 + }, + { + "epoch": 2.0086015037593983, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1339, + "step": 38393 + }, + { + "epoch": 2.0086616541353384, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 38394 + }, + { + "epoch": 2.008721804511278, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 38395 + }, + { + "epoch": 2.008781954887218, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 38396 + }, + { + "epoch": 2.0088421052631578, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 38397 + }, + { + "epoch": 2.008902255639098, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 38398 + }, + { + "epoch": 2.0089624060150375, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 38399 + }, + { + "epoch": 2.0090225563909776, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 38400 + }, + { + "epoch": 2.0090827067669172, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 38401 + }, + { + "epoch": 2.0091428571428573, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 38402 + }, + { + "epoch": 2.009203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 38403 + }, + { + "epoch": 2.0092631578947366, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 38404 + }, + { + "epoch": 2.0093233082706767, + "grad_norm": 1.6328125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 38405 + }, + { + "epoch": 2.0093834586466164, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 38406 + }, + { + "epoch": 2.0094436090225565, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 38407 + }, + { + "epoch": 2.009503759398496, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 38408 + }, + { + "epoch": 2.009563909774436, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 38409 + }, + { + "epoch": 2.009624060150376, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 38410 + }, + { + "epoch": 2.009684210526316, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 38411 + }, + { + "epoch": 2.0097443609022556, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1339, + "step": 38412 + }, + { + "epoch": 2.0098045112781957, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 38413 + }, + { + "epoch": 2.0098646616541354, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 38414 + }, + { + "epoch": 2.009924812030075, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 38415 + }, + { + "epoch": 2.009984962406015, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 38416 + }, + { + "epoch": 2.0100451127819547, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 38417 + }, + { + "epoch": 2.010105263157895, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 38418 + }, + { + "epoch": 2.0101654135338345, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 38419 + }, + { + "epoch": 2.0102255639097746, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 38420 + }, + { + "epoch": 2.0102857142857142, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 38421 + }, + { + "epoch": 2.0103458646616543, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 38422 + }, + { + "epoch": 2.010406015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.94, + "step": 38423 + }, + { + "epoch": 2.0104661654135336, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 38424 + }, + { + "epoch": 2.0105263157894737, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 38425 + }, + { + "epoch": 2.0105864661654134, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 38426 + }, + { + "epoch": 2.0106466165413535, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 38427 + }, + { + "epoch": 2.010706766917293, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 38428 + }, + { + "epoch": 2.010766917293233, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 38429 + }, + { + "epoch": 2.010827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 38430 + }, + { + "epoch": 2.010887218045113, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 38431 + }, + { + "epoch": 2.0109473684210526, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 38432 + }, + { + "epoch": 2.0110075187969927, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 38433 + }, + { + "epoch": 2.0110676691729323, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 38434 + }, + { + "epoch": 2.011127819548872, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 38435 + }, + { + "epoch": 2.011187969924812, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 38436 + }, + { + "epoch": 2.0112481203007517, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 38437 + }, + { + "epoch": 2.011308270676692, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 38438 + }, + { + "epoch": 2.0113684210526315, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 38439 + }, + { + "epoch": 2.0114285714285716, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 38440 + }, + { + "epoch": 2.011488721804511, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 38441 + }, + { + "epoch": 2.0115488721804513, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 38442 + }, + { + "epoch": 2.011609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 38443 + }, + { + "epoch": 2.0116691729323306, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 38444 + }, + { + "epoch": 2.0117293233082707, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 38445 + }, + { + "epoch": 2.0117894736842103, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 38446 + }, + { + "epoch": 2.0118496240601504, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 38447 + }, + { + "epoch": 2.01190977443609, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 38448 + }, + { + "epoch": 2.01196992481203, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 38449 + }, + { + "epoch": 2.01203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9497, + "step": 38450 + }, + { + "epoch": 2.01209022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 38451 + }, + { + "epoch": 2.0121503759398496, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 38452 + }, + { + "epoch": 2.0122105263157897, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 38453 + }, + { + "epoch": 2.0122706766917293, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 38454 + }, + { + "epoch": 2.012330827067669, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 38455 + }, + { + "epoch": 2.012390977443609, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 38456 + }, + { + "epoch": 2.0124511278195487, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 38457 + }, + { + "epoch": 2.012511278195489, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 38458 + }, + { + "epoch": 2.0125714285714285, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 38459 + }, + { + "epoch": 2.0126315789473685, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 38460 + }, + { + "epoch": 2.012691729323308, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 38461 + }, + { + "epoch": 2.0127518796992483, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 38462 + }, + { + "epoch": 2.012812030075188, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 38463 + }, + { + "epoch": 2.012872180451128, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 38464 + }, + { + "epoch": 2.0129323308270677, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 38465 + }, + { + "epoch": 2.0129924812030073, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1612, + "step": 38466 + }, + { + "epoch": 2.0130526315789474, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 38467 + }, + { + "epoch": 2.013112781954887, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 38468 + }, + { + "epoch": 2.013172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 38469 + }, + { + "epoch": 2.013233082706767, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 38470 + }, + { + "epoch": 2.013293233082707, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 38471 + }, + { + "epoch": 2.0133533834586466, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 38472 + }, + { + "epoch": 2.0134135338345867, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 38473 + }, + { + "epoch": 2.0134736842105263, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 38474 + }, + { + "epoch": 2.013533834586466, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 38475 + }, + { + "epoch": 2.013593984962406, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 38476 + }, + { + "epoch": 2.0136541353383457, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 38477 + }, + { + "epoch": 2.013714285714286, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 38478 + }, + { + "epoch": 2.0137744360902254, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 38479 + }, + { + "epoch": 2.0138345864661655, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 38480 + }, + { + "epoch": 2.013894736842105, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 38481 + }, + { + "epoch": 2.0139548872180453, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 38482 + }, + { + "epoch": 2.014015037593985, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 38483 + }, + { + "epoch": 2.014075187969925, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 38484 + }, + { + "epoch": 2.0141353383458647, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 38485 + }, + { + "epoch": 2.0141954887218043, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 38486 + }, + { + "epoch": 2.0142556390977444, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 38487 + }, + { + "epoch": 2.014315789473684, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 38488 + }, + { + "epoch": 2.014375939849624, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 38489 + }, + { + "epoch": 2.014436090225564, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 38490 + }, + { + "epoch": 2.014496240601504, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 38491 + }, + { + "epoch": 2.0145563909774435, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 38492 + }, + { + "epoch": 2.0146165413533836, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 38493 + }, + { + "epoch": 2.0146766917293233, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 38494 + }, + { + "epoch": 2.014736842105263, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 38495 + }, + { + "epoch": 2.014796992481203, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 38496 + }, + { + "epoch": 2.0148571428571427, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 38497 + }, + { + "epoch": 2.0149172932330828, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 38498 + }, + { + "epoch": 2.0149774436090224, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 38499 + }, + { + "epoch": 2.0150375939849625, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9384, + "step": 38500 + }, + { + "epoch": 2.015097744360902, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 38501 + }, + { + "epoch": 2.0151578947368423, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 38502 + }, + { + "epoch": 2.015218045112782, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 38503 + }, + { + "epoch": 2.015278195488722, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 38504 + }, + { + "epoch": 2.0153383458646617, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 38505 + }, + { + "epoch": 2.0153984962406013, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 38506 + }, + { + "epoch": 2.0154586466165414, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 38507 + }, + { + "epoch": 2.015518796992481, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 38508 + }, + { + "epoch": 2.015578947368421, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 38509 + }, + { + "epoch": 2.015639097744361, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 38510 + }, + { + "epoch": 2.015699248120301, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 38511 + }, + { + "epoch": 2.0157593984962405, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 38512 + }, + { + "epoch": 2.0158195488721806, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 38513 + }, + { + "epoch": 2.0158796992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 38514 + }, + { + "epoch": 2.0159398496240604, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 38515 + }, + { + "epoch": 2.016, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 38516 + }, + { + "epoch": 2.0160601503759397, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 38517 + }, + { + "epoch": 2.0161203007518798, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 38518 + }, + { + "epoch": 2.0161804511278194, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 38519 + }, + { + "epoch": 2.0162406015037595, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 38520 + }, + { + "epoch": 2.016300751879699, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 38521 + }, + { + "epoch": 2.0163609022556392, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 38522 + }, + { + "epoch": 2.016421052631579, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 38523 + }, + { + "epoch": 2.016481203007519, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 38524 + }, + { + "epoch": 2.0165413533834586, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 38525 + }, + { + "epoch": 2.0166015037593983, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 38526 + }, + { + "epoch": 2.0166616541353384, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 38527 + }, + { + "epoch": 2.016721804511278, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 38528 + }, + { + "epoch": 2.016781954887218, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 38529 + }, + { + "epoch": 2.0168421052631578, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 38530 + }, + { + "epoch": 2.016902255639098, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 38531 + }, + { + "epoch": 2.0169624060150375, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 38532 + }, + { + "epoch": 2.0170225563909776, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 38533 + }, + { + "epoch": 2.0170827067669173, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9432, + "step": 38534 + }, + { + "epoch": 2.0171428571428573, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 38535 + }, + { + "epoch": 2.017203007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 38536 + }, + { + "epoch": 2.0172631578947366, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 38537 + }, + { + "epoch": 2.0173233082706767, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 38538 + }, + { + "epoch": 2.0173834586466164, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 38539 + }, + { + "epoch": 2.0174436090225565, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 38540 + }, + { + "epoch": 2.017503759398496, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 38541 + }, + { + "epoch": 2.0175639097744362, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 38542 + }, + { + "epoch": 2.017624060150376, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 38543 + }, + { + "epoch": 2.017684210526316, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9406, + "step": 38544 + }, + { + "epoch": 2.0177443609022556, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 38545 + }, + { + "epoch": 2.0178045112781957, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 38546 + }, + { + "epoch": 2.0178646616541354, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 38547 + }, + { + "epoch": 2.017924812030075, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 38548 + }, + { + "epoch": 2.017984962406015, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 38549 + }, + { + "epoch": 2.0180451127819548, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 38550 + }, + { + "epoch": 2.018105263157895, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 38551 + }, + { + "epoch": 2.0181654135338345, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 38552 + }, + { + "epoch": 2.0182255639097746, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 38553 + }, + { + "epoch": 2.0182857142857142, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 38554 + }, + { + "epoch": 2.0183458646616543, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 38555 + }, + { + "epoch": 2.018406015037594, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 38556 + }, + { + "epoch": 2.0184661654135336, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 38557 + }, + { + "epoch": 2.0185263157894737, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 38558 + }, + { + "epoch": 2.0185864661654134, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 38559 + }, + { + "epoch": 2.0186466165413535, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 38560 + }, + { + "epoch": 2.018706766917293, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 38561 + }, + { + "epoch": 2.018766917293233, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 38562 + }, + { + "epoch": 2.018827067669173, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 38563 + }, + { + "epoch": 2.018887218045113, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 38564 + }, + { + "epoch": 2.0189473684210526, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 38565 + }, + { + "epoch": 2.0190075187969927, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 38566 + }, + { + "epoch": 2.0190676691729323, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 38567 + }, + { + "epoch": 2.019127819548872, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 38568 + }, + { + "epoch": 2.019187969924812, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 38569 + }, + { + "epoch": 2.0192481203007517, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 38570 + }, + { + "epoch": 2.019308270676692, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 38571 + }, + { + "epoch": 2.0193684210526315, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 38572 + }, + { + "epoch": 2.0194285714285716, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 38573 + }, + { + "epoch": 2.0194887218045112, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 38574 + }, + { + "epoch": 2.0195488721804513, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 38575 + }, + { + "epoch": 2.019609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 38576 + }, + { + "epoch": 2.0196691729323306, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 38577 + }, + { + "epoch": 2.0197293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9298, + "step": 38578 + }, + { + "epoch": 2.0197894736842104, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 38579 + }, + { + "epoch": 2.0198496240601505, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 38580 + }, + { + "epoch": 2.01990977443609, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 38581 + }, + { + "epoch": 2.01996992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 38582 + }, + { + "epoch": 2.02003007518797, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 38583 + }, + { + "epoch": 2.02009022556391, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 38584 + }, + { + "epoch": 2.0201503759398496, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 38585 + }, + { + "epoch": 2.0202105263157897, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 38586 + }, + { + "epoch": 2.0202706766917293, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 38587 + }, + { + "epoch": 2.020330827067669, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 38588 + }, + { + "epoch": 2.020390977443609, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 38589 + }, + { + "epoch": 2.0204511278195487, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 38590 + }, + { + "epoch": 2.020511278195489, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 38591 + }, + { + "epoch": 2.0205714285714285, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 38592 + }, + { + "epoch": 2.0206315789473686, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1414, + "step": 38593 + }, + { + "epoch": 2.020691729323308, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 38594 + }, + { + "epoch": 2.0207518796992483, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 38595 + }, + { + "epoch": 2.020812030075188, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 38596 + }, + { + "epoch": 2.0208721804511276, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 38597 + }, + { + "epoch": 2.0209323308270677, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 38598 + }, + { + "epoch": 2.0209924812030073, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 38599 + }, + { + "epoch": 2.0210526315789474, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 38600 + }, + { + "epoch": 2.021112781954887, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 38601 + }, + { + "epoch": 2.021172932330827, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 38602 + }, + { + "epoch": 2.021233082706767, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 38603 + }, + { + "epoch": 2.021293233082707, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 38604 + }, + { + "epoch": 2.0213533834586466, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9533, + "step": 38605 + }, + { + "epoch": 2.0214135338345867, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 38606 + }, + { + "epoch": 2.0214736842105263, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 38607 + }, + { + "epoch": 2.021533834586466, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 38608 + }, + { + "epoch": 2.021593984962406, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 38609 + }, + { + "epoch": 2.0216541353383457, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 38610 + }, + { + "epoch": 2.021714285714286, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 38611 + }, + { + "epoch": 2.0217744360902254, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 38612 + }, + { + "epoch": 2.0218345864661655, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 38613 + }, + { + "epoch": 2.021894736842105, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 38614 + }, + { + "epoch": 2.0219548872180453, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 38615 + }, + { + "epoch": 2.022015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 38616 + }, + { + "epoch": 2.022075187969925, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 38617 + }, + { + "epoch": 2.0221353383458647, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 38618 + }, + { + "epoch": 2.0221954887218043, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 38619 + }, + { + "epoch": 2.0222556390977444, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 38620 + }, + { + "epoch": 2.022315789473684, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9151, + "step": 38621 + }, + { + "epoch": 2.022375939849624, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 38622 + }, + { + "epoch": 2.022436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 38623 + }, + { + "epoch": 2.022496240601504, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 38624 + }, + { + "epoch": 2.0225563909774436, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 38625 + }, + { + "epoch": 2.0226165413533836, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 38626 + }, + { + "epoch": 2.0226766917293233, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 38627 + }, + { + "epoch": 2.022736842105263, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 38628 + }, + { + "epoch": 2.022796992481203, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 38629 + }, + { + "epoch": 2.0228571428571427, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 38630 + }, + { + "epoch": 2.022917293233083, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 38631 + }, + { + "epoch": 2.0229774436090224, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 38632 + }, + { + "epoch": 2.0230375939849625, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 38633 + }, + { + "epoch": 2.023097744360902, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 38634 + }, + { + "epoch": 2.0231578947368423, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 38635 + }, + { + "epoch": 2.023218045112782, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 38636 + }, + { + "epoch": 2.023278195488722, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 38637 + }, + { + "epoch": 2.0233383458646617, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 38638 + }, + { + "epoch": 2.0233984962406013, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 38639 + }, + { + "epoch": 2.0234586466165414, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 38640 + }, + { + "epoch": 2.023518796992481, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 38641 + }, + { + "epoch": 2.023578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 38642 + }, + { + "epoch": 2.023639097744361, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 38643 + }, + { + "epoch": 2.023699248120301, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 38644 + }, + { + "epoch": 2.0237593984962405, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 38645 + }, + { + "epoch": 2.0238195488721806, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 38646 + }, + { + "epoch": 2.0238796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 38647 + }, + { + "epoch": 2.0239398496240604, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 38648 + }, + { + "epoch": 2.024, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 38649 + }, + { + "epoch": 2.0240601503759397, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 38650 + }, + { + "epoch": 2.0241203007518798, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 38651 + }, + { + "epoch": 2.0241804511278194, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 38652 + }, + { + "epoch": 2.0242406015037595, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 38653 + }, + { + "epoch": 2.024300751879699, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 38654 + }, + { + "epoch": 2.0243609022556392, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 38655 + }, + { + "epoch": 2.024421052631579, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 38656 + }, + { + "epoch": 2.024481203007519, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 38657 + }, + { + "epoch": 2.0245413533834586, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 38658 + }, + { + "epoch": 2.0246015037593983, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 38659 + }, + { + "epoch": 2.0246616541353384, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 38660 + }, + { + "epoch": 2.024721804511278, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 38661 + }, + { + "epoch": 2.024781954887218, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 38662 + }, + { + "epoch": 2.0248421052631578, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 38663 + }, + { + "epoch": 2.024902255639098, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 38664 + }, + { + "epoch": 2.0249624060150375, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1665, + "step": 38665 + }, + { + "epoch": 2.0250225563909776, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 38666 + }, + { + "epoch": 2.0250827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 38667 + }, + { + "epoch": 2.0251428571428574, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 38668 + }, + { + "epoch": 2.025203007518797, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 38669 + }, + { + "epoch": 2.0252631578947367, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 38670 + }, + { + "epoch": 2.0253233082706767, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.92, + "step": 38671 + }, + { + "epoch": 2.0253834586466164, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 38672 + }, + { + "epoch": 2.0254436090225565, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 38673 + }, + { + "epoch": 2.025503759398496, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 38674 + }, + { + "epoch": 2.0255639097744362, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 38675 + }, + { + "epoch": 2.025624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 38676 + }, + { + "epoch": 2.025684210526316, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 38677 + }, + { + "epoch": 2.0257443609022556, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 38678 + }, + { + "epoch": 2.0258045112781953, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 38679 + }, + { + "epoch": 2.0258646616541354, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 38680 + }, + { + "epoch": 2.025924812030075, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9346, + "step": 38681 + }, + { + "epoch": 2.025984962406015, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 38682 + }, + { + "epoch": 2.0260451127819548, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 38683 + }, + { + "epoch": 2.026105263157895, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 38684 + }, + { + "epoch": 2.0261654135338345, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 38685 + }, + { + "epoch": 2.0262255639097746, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 38686 + }, + { + "epoch": 2.0262857142857142, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 38687 + }, + { + "epoch": 2.0263458646616543, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 38688 + }, + { + "epoch": 2.026406015037594, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 38689 + }, + { + "epoch": 2.0264661654135336, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 38690 + }, + { + "epoch": 2.0265263157894737, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 38691 + }, + { + "epoch": 2.0265864661654134, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 38692 + }, + { + "epoch": 2.0266466165413535, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 38693 + }, + { + "epoch": 2.026706766917293, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 38694 + }, + { + "epoch": 2.026766917293233, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 38695 + }, + { + "epoch": 2.026827067669173, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 38696 + }, + { + "epoch": 2.026887218045113, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 38697 + }, + { + "epoch": 2.0269473684210526, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 38698 + }, + { + "epoch": 2.0270075187969923, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9431, + "step": 38699 + }, + { + "epoch": 2.0270676691729324, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 38700 + }, + { + "epoch": 2.027127819548872, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 38701 + }, + { + "epoch": 2.027187969924812, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 38702 + }, + { + "epoch": 2.0272481203007517, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 38703 + }, + { + "epoch": 2.027308270676692, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 38704 + }, + { + "epoch": 2.0273684210526315, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 38705 + }, + { + "epoch": 2.0274285714285716, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 38706 + }, + { + "epoch": 2.0274887218045112, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 38707 + }, + { + "epoch": 2.0275488721804513, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 38708 + }, + { + "epoch": 2.027609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 38709 + }, + { + "epoch": 2.0276691729323306, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 38710 + }, + { + "epoch": 2.0277293233082707, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 38711 + }, + { + "epoch": 2.0277894736842104, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 38712 + }, + { + "epoch": 2.0278496240601505, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 38713 + }, + { + "epoch": 2.02790977443609, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 38714 + }, + { + "epoch": 2.02796992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9434, + "step": 38715 + }, + { + "epoch": 2.02803007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.147, + "step": 38716 + }, + { + "epoch": 2.02809022556391, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 38717 + }, + { + "epoch": 2.0281503759398496, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 38718 + }, + { + "epoch": 2.0282105263157897, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 38719 + }, + { + "epoch": 2.0282706766917293, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 38720 + }, + { + "epoch": 2.028330827067669, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 38721 + }, + { + "epoch": 2.028390977443609, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 38722 + }, + { + "epoch": 2.0284511278195487, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 38723 + }, + { + "epoch": 2.028511278195489, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 38724 + }, + { + "epoch": 2.0285714285714285, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 38725 + }, + { + "epoch": 2.0286315789473686, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 38726 + }, + { + "epoch": 2.028691729323308, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 38727 + }, + { + "epoch": 2.0287518796992483, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 38728 + }, + { + "epoch": 2.028812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 38729 + }, + { + "epoch": 2.0288721804511276, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 38730 + }, + { + "epoch": 2.0289323308270677, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 38731 + }, + { + "epoch": 2.0289924812030073, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 38732 + }, + { + "epoch": 2.0290526315789474, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 38733 + }, + { + "epoch": 2.029112781954887, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 38734 + }, + { + "epoch": 2.029172932330827, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 38735 + }, + { + "epoch": 2.029233082706767, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 38736 + }, + { + "epoch": 2.029293233082707, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 38737 + }, + { + "epoch": 2.0293533834586466, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1841, + "step": 38738 + }, + { + "epoch": 2.0294135338345867, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 38739 + }, + { + "epoch": 2.0294736842105263, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 38740 + }, + { + "epoch": 2.029533834586466, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 38741 + }, + { + "epoch": 2.029593984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 38742 + }, + { + "epoch": 2.0296541353383457, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 38743 + }, + { + "epoch": 2.029714285714286, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 38744 + }, + { + "epoch": 2.0297744360902255, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 38745 + }, + { + "epoch": 2.0298345864661655, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 38746 + }, + { + "epoch": 2.029894736842105, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 38747 + }, + { + "epoch": 2.0299548872180453, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 38748 + }, + { + "epoch": 2.030015037593985, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 38749 + }, + { + "epoch": 2.030075187969925, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 38750 + }, + { + "epoch": 2.0301353383458647, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 38751 + }, + { + "epoch": 2.0301954887218043, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 38752 + }, + { + "epoch": 2.0302556390977444, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9501, + "step": 38753 + }, + { + "epoch": 2.030315789473684, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 38754 + }, + { + "epoch": 2.030375939849624, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 38755 + }, + { + "epoch": 2.030436090225564, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 38756 + }, + { + "epoch": 2.030496240601504, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 38757 + }, + { + "epoch": 2.0305563909774436, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 38758 + }, + { + "epoch": 2.0306165413533837, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 38759 + }, + { + "epoch": 2.0306766917293233, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1531, + "step": 38760 + }, + { + "epoch": 2.030736842105263, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 38761 + }, + { + "epoch": 2.030796992481203, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 38762 + }, + { + "epoch": 2.0308571428571427, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 38763 + }, + { + "epoch": 2.030917293233083, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 38764 + }, + { + "epoch": 2.0309774436090224, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 38765 + }, + { + "epoch": 2.0310375939849625, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 38766 + }, + { + "epoch": 2.031097744360902, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 38767 + }, + { + "epoch": 2.0311578947368423, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 38768 + }, + { + "epoch": 2.031218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 38769 + }, + { + "epoch": 2.031278195488722, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 38770 + }, + { + "epoch": 2.0313383458646617, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 38771 + }, + { + "epoch": 2.0313984962406013, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 38772 + }, + { + "epoch": 2.0314586466165414, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 38773 + }, + { + "epoch": 2.031518796992481, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 38774 + }, + { + "epoch": 2.031578947368421, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 38775 + }, + { + "epoch": 2.031639097744361, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 38776 + }, + { + "epoch": 2.031699248120301, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 38777 + }, + { + "epoch": 2.0317593984962405, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 38778 + }, + { + "epoch": 2.0318195488721806, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 38779 + }, + { + "epoch": 2.0318796992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 38780 + }, + { + "epoch": 2.03193984962406, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 38781 + }, + { + "epoch": 2.032, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 38782 + }, + { + "epoch": 2.0320601503759397, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 38783 + }, + { + "epoch": 2.0321203007518798, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 38784 + }, + { + "epoch": 2.0321804511278194, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 38785 + }, + { + "epoch": 2.0322406015037595, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 38786 + }, + { + "epoch": 2.032300751879699, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 38787 + }, + { + "epoch": 2.0323609022556393, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9032, + "step": 38788 + }, + { + "epoch": 2.032421052631579, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 38789 + }, + { + "epoch": 2.032481203007519, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 38790 + }, + { + "epoch": 2.0325413533834586, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 38791 + }, + { + "epoch": 2.0326015037593983, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 38792 + }, + { + "epoch": 2.0326616541353384, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 38793 + }, + { + "epoch": 2.032721804511278, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 38794 + }, + { + "epoch": 2.032781954887218, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 38795 + }, + { + "epoch": 2.032842105263158, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 38796 + }, + { + "epoch": 2.032902255639098, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 38797 + }, + { + "epoch": 2.0329624060150375, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 38798 + }, + { + "epoch": 2.0330225563909776, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 38799 + }, + { + "epoch": 2.0330827067669173, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 38800 + }, + { + "epoch": 2.033142857142857, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 38801 + }, + { + "epoch": 2.033203007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 38802 + }, + { + "epoch": 2.0332631578947367, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 38803 + }, + { + "epoch": 2.0333233082706768, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 38804 + }, + { + "epoch": 2.0333834586466164, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 38805 + }, + { + "epoch": 2.0334436090225565, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 38806 + }, + { + "epoch": 2.033503759398496, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 38807 + }, + { + "epoch": 2.0335639097744362, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 38808 + }, + { + "epoch": 2.033624060150376, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 38809 + }, + { + "epoch": 2.033684210526316, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 38810 + }, + { + "epoch": 2.0337443609022556, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 38811 + }, + { + "epoch": 2.0338045112781953, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 38812 + }, + { + "epoch": 2.0338646616541354, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 38813 + }, + { + "epoch": 2.033924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 38814 + }, + { + "epoch": 2.033984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 38815 + }, + { + "epoch": 2.0340451127819548, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 38816 + }, + { + "epoch": 2.034105263157895, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 38817 + }, + { + "epoch": 2.0341654135338345, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 38818 + }, + { + "epoch": 2.0342255639097746, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 38819 + }, + { + "epoch": 2.0342857142857143, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 38820 + }, + { + "epoch": 2.0343458646616543, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 38821 + }, + { + "epoch": 2.034406015037594, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 38822 + }, + { + "epoch": 2.0344661654135336, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 38823 + }, + { + "epoch": 2.0345263157894737, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 38824 + }, + { + "epoch": 2.0345864661654134, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 38825 + }, + { + "epoch": 2.0346466165413535, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 38826 + }, + { + "epoch": 2.034706766917293, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 38827 + }, + { + "epoch": 2.0347669172932332, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 38828 + }, + { + "epoch": 2.034827067669173, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 38829 + }, + { + "epoch": 2.034887218045113, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1073, + "step": 38830 + }, + { + "epoch": 2.0349473684210526, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 38831 + }, + { + "epoch": 2.0350075187969923, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 38832 + }, + { + "epoch": 2.0350676691729324, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9202, + "step": 38833 + }, + { + "epoch": 2.035127819548872, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 38834 + }, + { + "epoch": 2.035187969924812, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 38835 + }, + { + "epoch": 2.0352481203007518, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 38836 + }, + { + "epoch": 2.035308270676692, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 38837 + }, + { + "epoch": 2.0353684210526315, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 38838 + }, + { + "epoch": 2.0354285714285716, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 38839 + }, + { + "epoch": 2.0354887218045112, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 38840 + }, + { + "epoch": 2.0355488721804513, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 38841 + }, + { + "epoch": 2.035609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 38842 + }, + { + "epoch": 2.0356691729323306, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 38843 + }, + { + "epoch": 2.0357293233082707, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 38844 + }, + { + "epoch": 2.0357894736842104, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 38845 + }, + { + "epoch": 2.0358496240601505, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 38846 + }, + { + "epoch": 2.03590977443609, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 38847 + }, + { + "epoch": 2.03596992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 38848 + }, + { + "epoch": 2.03603007518797, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 38849 + }, + { + "epoch": 2.03609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 38850 + }, + { + "epoch": 2.0361503759398496, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 38851 + }, + { + "epoch": 2.0362105263157897, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 38852 + }, + { + "epoch": 2.0362706766917293, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 38853 + }, + { + "epoch": 2.036330827067669, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 38854 + }, + { + "epoch": 2.036390977443609, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 38855 + }, + { + "epoch": 2.0364511278195487, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 38856 + }, + { + "epoch": 2.036511278195489, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 38857 + }, + { + "epoch": 2.0365714285714285, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 38858 + }, + { + "epoch": 2.0366315789473686, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 38859 + }, + { + "epoch": 2.036691729323308, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 38860 + }, + { + "epoch": 2.0367518796992483, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 38861 + }, + { + "epoch": 2.036812030075188, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 38862 + }, + { + "epoch": 2.0368721804511276, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 38863 + }, + { + "epoch": 2.0369323308270677, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 38864 + }, + { + "epoch": 2.0369924812030074, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 38865 + }, + { + "epoch": 2.0370526315789474, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 38866 + }, + { + "epoch": 2.037112781954887, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 38867 + }, + { + "epoch": 2.037172932330827, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 38868 + }, + { + "epoch": 2.037233082706767, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 38869 + }, + { + "epoch": 2.037293233082707, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9281, + "step": 38870 + }, + { + "epoch": 2.0373533834586466, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 38871 + }, + { + "epoch": 2.0374135338345867, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 38872 + }, + { + "epoch": 2.0374736842105263, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 38873 + }, + { + "epoch": 2.037533834586466, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 38874 + }, + { + "epoch": 2.037593984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.2019, + "step": 38875 + }, + { + "epoch": 2.0376541353383457, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 38876 + }, + { + "epoch": 2.037714285714286, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 38877 + }, + { + "epoch": 2.0377744360902255, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 38878 + }, + { + "epoch": 2.0378345864661656, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 38879 + }, + { + "epoch": 2.037894736842105, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 38880 + }, + { + "epoch": 2.0379548872180453, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 38881 + }, + { + "epoch": 2.038015037593985, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 38882 + }, + { + "epoch": 2.0380751879699246, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 38883 + }, + { + "epoch": 2.0381353383458647, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 38884 + }, + { + "epoch": 2.0381954887218043, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 38885 + }, + { + "epoch": 2.0382556390977444, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 38886 + }, + { + "epoch": 2.038315789473684, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 38887 + }, + { + "epoch": 2.038375939849624, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 38888 + }, + { + "epoch": 2.038436090225564, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 38889 + }, + { + "epoch": 2.038496240601504, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 38890 + }, + { + "epoch": 2.0385563909774436, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 38891 + }, + { + "epoch": 2.0386165413533837, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 38892 + }, + { + "epoch": 2.0386766917293233, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 38893 + }, + { + "epoch": 2.038736842105263, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 38894 + }, + { + "epoch": 2.038796992481203, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 38895 + }, + { + "epoch": 2.0388571428571427, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 38896 + }, + { + "epoch": 2.038917293233083, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 38897 + }, + { + "epoch": 2.0389774436090224, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 38898 + }, + { + "epoch": 2.0390375939849625, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 38899 + }, + { + "epoch": 2.039097744360902, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 38900 + }, + { + "epoch": 2.0391578947368423, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 38901 + }, + { + "epoch": 2.039218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 38902 + }, + { + "epoch": 2.039278195488722, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 38903 + }, + { + "epoch": 2.0393383458646617, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 38904 + }, + { + "epoch": 2.0393984962406013, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 38905 + }, + { + "epoch": 2.0394586466165414, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 38906 + }, + { + "epoch": 2.039518796992481, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 38907 + }, + { + "epoch": 2.039578947368421, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 38908 + }, + { + "epoch": 2.039639097744361, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 38909 + }, + { + "epoch": 2.039699248120301, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 38910 + }, + { + "epoch": 2.0397593984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 38911 + }, + { + "epoch": 2.0398195488721806, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 38912 + }, + { + "epoch": 2.0398796992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 38913 + }, + { + "epoch": 2.03993984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 38914 + }, + { + "epoch": 2.04, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 38915 + }, + { + "epoch": 2.0400601503759397, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1593, + "step": 38916 + }, + { + "epoch": 2.04012030075188, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 38917 + }, + { + "epoch": 2.0401804511278194, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 38918 + }, + { + "epoch": 2.0402406015037595, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 38919 + }, + { + "epoch": 2.040300751879699, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 38920 + }, + { + "epoch": 2.0403609022556393, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 38921 + }, + { + "epoch": 2.040421052631579, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 38922 + }, + { + "epoch": 2.040481203007519, + "grad_norm": 1.3984375, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 38923 + }, + { + "epoch": 2.0405413533834587, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 38924 + }, + { + "epoch": 2.0406015037593983, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 38925 + }, + { + "epoch": 2.0406616541353384, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 38926 + }, + { + "epoch": 2.040721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 38927 + }, + { + "epoch": 2.040781954887218, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 38928 + }, + { + "epoch": 2.040842105263158, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 38929 + }, + { + "epoch": 2.040902255639098, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 38930 + }, + { + "epoch": 2.0409624060150375, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 38931 + }, + { + "epoch": 2.0410225563909776, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 38932 + }, + { + "epoch": 2.0410827067669173, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 38933 + }, + { + "epoch": 2.041142857142857, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 38934 + }, + { + "epoch": 2.041203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 38935 + }, + { + "epoch": 2.0412631578947367, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 38936 + }, + { + "epoch": 2.0413233082706768, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 38937 + }, + { + "epoch": 2.0413834586466164, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 38938 + }, + { + "epoch": 2.0414436090225565, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 38939 + }, + { + "epoch": 2.041503759398496, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9284, + "step": 38940 + }, + { + "epoch": 2.0415639097744362, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 38941 + }, + { + "epoch": 2.041624060150376, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 38942 + }, + { + "epoch": 2.041684210526316, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 38943 + }, + { + "epoch": 2.0417443609022556, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 38944 + }, + { + "epoch": 2.0418045112781953, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1662, + "step": 38945 + }, + { + "epoch": 2.0418646616541354, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 38946 + }, + { + "epoch": 2.041924812030075, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 38947 + }, + { + "epoch": 2.041984962406015, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 38948 + }, + { + "epoch": 2.0420451127819548, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 38949 + }, + { + "epoch": 2.042105263157895, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 38950 + }, + { + "epoch": 2.0421654135338345, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 38951 + }, + { + "epoch": 2.0422255639097746, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 38952 + }, + { + "epoch": 2.0422857142857143, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 38953 + }, + { + "epoch": 2.0423458646616544, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 38954 + }, + { + "epoch": 2.042406015037594, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 38955 + }, + { + "epoch": 2.0424661654135337, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 38956 + }, + { + "epoch": 2.0425263157894737, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 38957 + }, + { + "epoch": 2.0425864661654134, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 38958 + }, + { + "epoch": 2.0426466165413535, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 38959 + }, + { + "epoch": 2.042706766917293, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 38960 + }, + { + "epoch": 2.0427669172932332, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 38961 + }, + { + "epoch": 2.042827067669173, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 38962 + }, + { + "epoch": 2.042887218045113, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 38963 + }, + { + "epoch": 2.0429473684210526, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 38964 + }, + { + "epoch": 2.0430075187969923, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 38965 + }, + { + "epoch": 2.0430676691729324, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 38966 + }, + { + "epoch": 2.043127819548872, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 38967 + }, + { + "epoch": 2.043187969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 38968 + }, + { + "epoch": 2.0432481203007518, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 38969 + }, + { + "epoch": 2.043308270676692, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 38970 + }, + { + "epoch": 2.0433684210526315, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 38971 + }, + { + "epoch": 2.0434285714285716, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 38972 + }, + { + "epoch": 2.0434887218045112, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 38973 + }, + { + "epoch": 2.0435488721804513, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 38974 + }, + { + "epoch": 2.043609022556391, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9362, + "step": 38975 + }, + { + "epoch": 2.0436691729323306, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 38976 + }, + { + "epoch": 2.0437293233082707, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 38977 + }, + { + "epoch": 2.0437894736842104, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 38978 + }, + { + "epoch": 2.0438496240601505, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 38979 + }, + { + "epoch": 2.04390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 38980 + }, + { + "epoch": 2.04396992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 38981 + }, + { + "epoch": 2.04403007518797, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 38982 + }, + { + "epoch": 2.04409022556391, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 38983 + }, + { + "epoch": 2.0441503759398496, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 1.1634, + "step": 38984 + }, + { + "epoch": 2.0442105263157893, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.8977, + "step": 38985 + }, + { + "epoch": 2.0442706766917293, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 38986 + }, + { + "epoch": 2.044330827067669, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 38987 + }, + { + "epoch": 2.044390977443609, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 38988 + }, + { + "epoch": 2.0444511278195487, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 38989 + }, + { + "epoch": 2.044511278195489, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 38990 + }, + { + "epoch": 2.0445714285714285, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 38991 + }, + { + "epoch": 2.0446315789473686, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 38992 + }, + { + "epoch": 2.0446917293233082, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 38993 + }, + { + "epoch": 2.0447518796992483, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.926, + "step": 38994 + }, + { + "epoch": 2.044812030075188, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9297, + "step": 38995 + }, + { + "epoch": 2.0448721804511276, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9036, + "step": 38996 + }, + { + "epoch": 2.0449323308270677, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 38997 + }, + { + "epoch": 2.0449924812030074, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 38998 + }, + { + "epoch": 2.0450526315789475, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 38999 + }, + { + "epoch": 2.045112781954887, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 39000 + }, + { + "epoch": 2.045172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9201, + "step": 39001 + }, + { + "epoch": 2.045233082706767, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 39002 + }, + { + "epoch": 2.045293233082707, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 39003 + }, + { + "epoch": 2.0453533834586466, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.2506, + "step": 39004 + }, + { + "epoch": 2.0454135338345867, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 39005 + }, + { + "epoch": 2.0454736842105263, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9396, + "step": 39006 + }, + { + "epoch": 2.045533834586466, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 39007 + }, + { + "epoch": 2.045593984962406, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 39008 + }, + { + "epoch": 2.0456541353383457, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 39009 + }, + { + "epoch": 2.045714285714286, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 39010 + }, + { + "epoch": 2.0457744360902255, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 39011 + }, + { + "epoch": 2.0458345864661656, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 39012 + }, + { + "epoch": 2.045894736842105, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 39013 + }, + { + "epoch": 2.0459548872180453, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 39014 + }, + { + "epoch": 2.046015037593985, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 39015 + }, + { + "epoch": 2.0460751879699246, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 39016 + }, + { + "epoch": 2.0461353383458647, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 39017 + }, + { + "epoch": 2.0461954887218043, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 39018 + }, + { + "epoch": 2.0462556390977444, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 39019 + }, + { + "epoch": 2.046315789473684, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 39020 + }, + { + "epoch": 2.046375939849624, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 39021 + }, + { + "epoch": 2.046436090225564, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 39022 + }, + { + "epoch": 2.046496240601504, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 39023 + }, + { + "epoch": 2.0465563909774436, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 39024 + }, + { + "epoch": 2.0466165413533837, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 39025 + }, + { + "epoch": 2.0466766917293233, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 39026 + }, + { + "epoch": 2.046736842105263, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 39027 + }, + { + "epoch": 2.046796992481203, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 39028 + }, + { + "epoch": 2.0468571428571427, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 39029 + }, + { + "epoch": 2.046917293233083, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 39030 + }, + { + "epoch": 2.0469774436090225, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 39031 + }, + { + "epoch": 2.0470375939849625, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 39032 + }, + { + "epoch": 2.047097744360902, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 39033 + }, + { + "epoch": 2.0471578947368423, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 39034 + }, + { + "epoch": 2.047218045112782, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 39035 + }, + { + "epoch": 2.0472781954887216, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 39036 + }, + { + "epoch": 2.0473383458646617, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 39037 + }, + { + "epoch": 2.0473984962406013, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 39038 + }, + { + "epoch": 2.0474586466165414, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 39039 + }, + { + "epoch": 2.047518796992481, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 39040 + }, + { + "epoch": 2.047578947368421, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 39041 + }, + { + "epoch": 2.047639097744361, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1677, + "step": 39042 + }, + { + "epoch": 2.047699248120301, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 39043 + }, + { + "epoch": 2.0477593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 39044 + }, + { + "epoch": 2.0478195488721807, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 39045 + }, + { + "epoch": 2.0478796992481203, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 39046 + }, + { + "epoch": 2.04793984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 39047 + }, + { + "epoch": 2.048, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 39048 + }, + { + "epoch": 2.0480601503759397, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 39049 + }, + { + "epoch": 2.04812030075188, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 39050 + }, + { + "epoch": 2.0481804511278194, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 39051 + }, + { + "epoch": 2.0482406015037595, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 39052 + }, + { + "epoch": 2.048300751879699, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 39053 + }, + { + "epoch": 2.0483609022556393, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 39054 + }, + { + "epoch": 2.048421052631579, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 39055 + }, + { + "epoch": 2.048481203007519, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9257, + "step": 39056 + }, + { + "epoch": 2.0485413533834587, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 39057 + }, + { + "epoch": 2.0486015037593983, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 39058 + }, + { + "epoch": 2.0486616541353384, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 39059 + }, + { + "epoch": 2.048721804511278, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 39060 + }, + { + "epoch": 2.048781954887218, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 39061 + }, + { + "epoch": 2.048842105263158, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 39062 + }, + { + "epoch": 2.048902255639098, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1349, + "step": 39063 + }, + { + "epoch": 2.0489624060150375, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 39064 + }, + { + "epoch": 2.0490225563909776, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 39065 + }, + { + "epoch": 2.0490827067669173, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 39066 + }, + { + "epoch": 2.049142857142857, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 39067 + }, + { + "epoch": 2.049203007518797, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 39068 + }, + { + "epoch": 2.0492631578947367, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 39069 + }, + { + "epoch": 2.0493233082706768, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 39070 + }, + { + "epoch": 2.0493834586466164, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 39071 + }, + { + "epoch": 2.0494436090225565, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 39072 + }, + { + "epoch": 2.049503759398496, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9341, + "step": 39073 + }, + { + "epoch": 2.0495639097744363, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 39074 + }, + { + "epoch": 2.049624060150376, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 39075 + }, + { + "epoch": 2.049684210526316, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 39076 + }, + { + "epoch": 2.0497443609022556, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 39077 + }, + { + "epoch": 2.0498045112781953, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 39078 + }, + { + "epoch": 2.0498646616541354, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 39079 + }, + { + "epoch": 2.049924812030075, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 39080 + }, + { + "epoch": 2.049984962406015, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 39081 + }, + { + "epoch": 2.050045112781955, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 39082 + }, + { + "epoch": 2.050105263157895, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 39083 + }, + { + "epoch": 2.0501654135338345, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 39084 + }, + { + "epoch": 2.0502255639097746, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 39085 + }, + { + "epoch": 2.0502857142857143, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 39086 + }, + { + "epoch": 2.0503458646616544, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 39087 + }, + { + "epoch": 2.050406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 39088 + }, + { + "epoch": 2.0504661654135337, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 39089 + }, + { + "epoch": 2.0505263157894738, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1491, + "step": 39090 + }, + { + "epoch": 2.0505864661654134, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 39091 + }, + { + "epoch": 2.0506466165413535, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 39092 + }, + { + "epoch": 2.050706766917293, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 39093 + }, + { + "epoch": 2.0507669172932332, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 39094 + }, + { + "epoch": 2.050827067669173, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 39095 + }, + { + "epoch": 2.050887218045113, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 39096 + }, + { + "epoch": 2.0509473684210526, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 39097 + }, + { + "epoch": 2.0510075187969923, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 39098 + }, + { + "epoch": 2.0510676691729324, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 39099 + }, + { + "epoch": 2.051127819548872, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 39100 + }, + { + "epoch": 2.051187969924812, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 39101 + }, + { + "epoch": 2.0512481203007518, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 39102 + }, + { + "epoch": 2.051308270676692, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 39103 + }, + { + "epoch": 2.0513684210526315, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 39104 + }, + { + "epoch": 2.0514285714285716, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 39105 + }, + { + "epoch": 2.0514887218045113, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 39106 + }, + { + "epoch": 2.0515488721804513, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9219, + "step": 39107 + }, + { + "epoch": 2.051609022556391, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 39108 + }, + { + "epoch": 2.0516691729323306, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 39109 + }, + { + "epoch": 2.0517293233082707, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 39110 + }, + { + "epoch": 2.0517894736842104, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 39111 + }, + { + "epoch": 2.0518496240601505, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 39112 + }, + { + "epoch": 2.05190977443609, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 39113 + }, + { + "epoch": 2.05196992481203, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 39114 + }, + { + "epoch": 2.05203007518797, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 39115 + }, + { + "epoch": 2.05209022556391, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 39116 + }, + { + "epoch": 2.0521503759398496, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 39117 + }, + { + "epoch": 2.0522105263157893, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 39118 + }, + { + "epoch": 2.0522706766917294, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 39119 + }, + { + "epoch": 2.052330827067669, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 39120 + }, + { + "epoch": 2.052390977443609, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 39121 + }, + { + "epoch": 2.0524511278195487, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 39122 + }, + { + "epoch": 2.052511278195489, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 39123 + }, + { + "epoch": 2.0525714285714285, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 39124 + }, + { + "epoch": 2.0526315789473686, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 39125 + }, + { + "epoch": 2.0526917293233082, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.209, + "step": 39126 + }, + { + "epoch": 2.0527518796992483, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 39127 + }, + { + "epoch": 2.052812030075188, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 39128 + }, + { + "epoch": 2.0528721804511276, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 39129 + }, + { + "epoch": 2.0529323308270677, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 39130 + }, + { + "epoch": 2.0529924812030074, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 39131 + }, + { + "epoch": 2.0530526315789475, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 39132 + }, + { + "epoch": 2.053112781954887, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 39133 + }, + { + "epoch": 2.053172932330827, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 39134 + }, + { + "epoch": 2.053233082706767, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 39135 + }, + { + "epoch": 2.053293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 39136 + }, + { + "epoch": 2.0533533834586466, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 39137 + }, + { + "epoch": 2.0534135338345862, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1911, + "step": 39138 + }, + { + "epoch": 2.0534736842105263, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 39139 + }, + { + "epoch": 2.053533834586466, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 39140 + }, + { + "epoch": 2.053593984962406, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 39141 + }, + { + "epoch": 2.0536541353383457, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 39142 + }, + { + "epoch": 2.053714285714286, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 39143 + }, + { + "epoch": 2.0537744360902255, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 39144 + }, + { + "epoch": 2.0538345864661656, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 39145 + }, + { + "epoch": 2.053894736842105, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 39146 + }, + { + "epoch": 2.0539548872180453, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 39147 + }, + { + "epoch": 2.054015037593985, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 39148 + }, + { + "epoch": 2.0540751879699246, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 39149 + }, + { + "epoch": 2.0541353383458647, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 39150 + }, + { + "epoch": 2.0541954887218044, + "grad_norm": 0.466796875, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 39151 + }, + { + "epoch": 2.0542556390977444, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 39152 + }, + { + "epoch": 2.054315789473684, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 39153 + }, + { + "epoch": 2.054375939849624, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 39154 + }, + { + "epoch": 2.054436090225564, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 39155 + }, + { + "epoch": 2.054496240601504, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 39156 + }, + { + "epoch": 2.0545563909774436, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 39157 + }, + { + "epoch": 2.0546165413533837, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 39158 + }, + { + "epoch": 2.0546766917293233, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 39159 + }, + { + "epoch": 2.054736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 39160 + }, + { + "epoch": 2.054796992481203, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 39161 + }, + { + "epoch": 2.0548571428571427, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 39162 + }, + { + "epoch": 2.054917293233083, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 39163 + }, + { + "epoch": 2.0549774436090225, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 39164 + }, + { + "epoch": 2.0550375939849626, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 39165 + }, + { + "epoch": 2.055097744360902, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 39166 + }, + { + "epoch": 2.0551578947368423, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9233, + "step": 39167 + }, + { + "epoch": 2.055218045112782, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 39168 + }, + { + "epoch": 2.0552781954887216, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 39169 + }, + { + "epoch": 2.0553383458646617, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 39170 + }, + { + "epoch": 2.0553984962406013, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 39171 + }, + { + "epoch": 2.0554586466165414, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 39172 + }, + { + "epoch": 2.055518796992481, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 39173 + }, + { + "epoch": 2.055578947368421, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 39174 + }, + { + "epoch": 2.055639097744361, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 39175 + }, + { + "epoch": 2.055699248120301, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 39176 + }, + { + "epoch": 2.0557593984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 39177 + }, + { + "epoch": 2.0558195488721807, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 39178 + }, + { + "epoch": 2.0558796992481203, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 39179 + }, + { + "epoch": 2.05593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 39180 + }, + { + "epoch": 2.056, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 39181 + }, + { + "epoch": 2.0560601503759397, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 39182 + }, + { + "epoch": 2.05612030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 39183 + }, + { + "epoch": 2.0561804511278194, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 39184 + }, + { + "epoch": 2.0562406015037595, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 39185 + }, + { + "epoch": 2.056300751879699, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 39186 + }, + { + "epoch": 2.0563609022556393, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 39187 + }, + { + "epoch": 2.056421052631579, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 39188 + }, + { + "epoch": 2.056481203007519, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 39189 + }, + { + "epoch": 2.0565413533834587, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 39190 + }, + { + "epoch": 2.0566015037593983, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9395, + "step": 39191 + }, + { + "epoch": 2.0566616541353384, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 39192 + }, + { + "epoch": 2.056721804511278, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 39193 + }, + { + "epoch": 2.056781954887218, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 39194 + }, + { + "epoch": 2.056842105263158, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 39195 + }, + { + "epoch": 2.056902255639098, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 39196 + }, + { + "epoch": 2.0569624060150375, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1663, + "step": 39197 + }, + { + "epoch": 2.0570225563909776, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 39198 + }, + { + "epoch": 2.0570827067669173, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 39199 + }, + { + "epoch": 2.057142857142857, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 39200 + }, + { + "epoch": 2.057203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 39201 + }, + { + "epoch": 2.0572631578947367, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 39202 + }, + { + "epoch": 2.0573233082706768, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 39203 + }, + { + "epoch": 2.0573834586466164, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 39204 + }, + { + "epoch": 2.0574436090225565, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 39205 + }, + { + "epoch": 2.057503759398496, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 39206 + }, + { + "epoch": 2.0575639097744363, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 39207 + }, + { + "epoch": 2.057624060150376, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 39208 + }, + { + "epoch": 2.057684210526316, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 39209 + }, + { + "epoch": 2.0577443609022557, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 39210 + }, + { + "epoch": 2.0578045112781953, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 39211 + }, + { + "epoch": 2.0578646616541354, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 39212 + }, + { + "epoch": 2.057924812030075, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 39213 + }, + { + "epoch": 2.057984962406015, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 39214 + }, + { + "epoch": 2.058045112781955, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 39215 + }, + { + "epoch": 2.058105263157895, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 39216 + }, + { + "epoch": 2.0581654135338345, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 39217 + }, + { + "epoch": 2.0582255639097746, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 39218 + }, + { + "epoch": 2.0582857142857143, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 39219 + }, + { + "epoch": 2.058345864661654, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 39220 + }, + { + "epoch": 2.058406015037594, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 39221 + }, + { + "epoch": 2.0584661654135337, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 39222 + }, + { + "epoch": 2.0585263157894738, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 39223 + }, + { + "epoch": 2.0585864661654134, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 39224 + }, + { + "epoch": 2.0586466165413535, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 39225 + }, + { + "epoch": 2.058706766917293, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 39226 + }, + { + "epoch": 2.0587669172932332, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 39227 + }, + { + "epoch": 2.058827067669173, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 39228 + }, + { + "epoch": 2.058887218045113, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 39229 + }, + { + "epoch": 2.0589473684210526, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 39230 + }, + { + "epoch": 2.0590075187969923, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 39231 + }, + { + "epoch": 2.0590676691729324, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 39232 + }, + { + "epoch": 2.059127819548872, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 39233 + }, + { + "epoch": 2.059187969924812, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 39234 + }, + { + "epoch": 2.0592481203007518, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 39235 + }, + { + "epoch": 2.059308270676692, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 39236 + }, + { + "epoch": 2.0593684210526315, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 39237 + }, + { + "epoch": 2.0594285714285716, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 39238 + }, + { + "epoch": 2.0594887218045113, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 39239 + }, + { + "epoch": 2.059548872180451, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 39240 + }, + { + "epoch": 2.059609022556391, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 39241 + }, + { + "epoch": 2.0596691729323306, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 39242 + }, + { + "epoch": 2.0597293233082707, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 39243 + }, + { + "epoch": 2.0597894736842104, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 39244 + }, + { + "epoch": 2.0598496240601505, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 39245 + }, + { + "epoch": 2.05990977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 39246 + }, + { + "epoch": 2.0599699248120302, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 39247 + }, + { + "epoch": 2.06003007518797, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 39248 + }, + { + "epoch": 2.06009022556391, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 39249 + }, + { + "epoch": 2.0601503759398496, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 39250 + }, + { + "epoch": 2.0602105263157893, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 39251 + }, + { + "epoch": 2.0602706766917294, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9382, + "step": 39252 + }, + { + "epoch": 2.060330827067669, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 39253 + }, + { + "epoch": 2.060390977443609, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 39254 + }, + { + "epoch": 2.0604511278195488, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 39255 + }, + { + "epoch": 2.060511278195489, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9667, + "step": 39256 + }, + { + "epoch": 2.0605714285714285, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 39257 + }, + { + "epoch": 2.0606315789473686, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 39258 + }, + { + "epoch": 2.0606917293233082, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 39259 + }, + { + "epoch": 2.0607518796992483, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 39260 + }, + { + "epoch": 2.060812030075188, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 39261 + }, + { + "epoch": 2.0608721804511276, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 39262 + }, + { + "epoch": 2.0609323308270677, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 39263 + }, + { + "epoch": 2.0609924812030074, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 39264 + }, + { + "epoch": 2.0610526315789475, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 39265 + }, + { + "epoch": 2.061112781954887, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 39266 + }, + { + "epoch": 2.061172932330827, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 39267 + }, + { + "epoch": 2.061233082706767, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 39268 + }, + { + "epoch": 2.061293233082707, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 39269 + }, + { + "epoch": 2.0613533834586466, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 39270 + }, + { + "epoch": 2.0614135338345863, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 39271 + }, + { + "epoch": 2.0614736842105263, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 39272 + }, + { + "epoch": 2.061533834586466, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 39273 + }, + { + "epoch": 2.061593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 39274 + }, + { + "epoch": 2.0616541353383457, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9237, + "step": 39275 + }, + { + "epoch": 2.061714285714286, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 39276 + }, + { + "epoch": 2.0617744360902255, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 39277 + }, + { + "epoch": 2.0618345864661656, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 39278 + }, + { + "epoch": 2.0618947368421052, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 39279 + }, + { + "epoch": 2.0619548872180453, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 39280 + }, + { + "epoch": 2.062015037593985, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 39281 + }, + { + "epoch": 2.0620751879699246, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 39282 + }, + { + "epoch": 2.0621353383458647, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 39283 + }, + { + "epoch": 2.0621954887218044, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 39284 + }, + { + "epoch": 2.0622556390977445, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 39285 + }, + { + "epoch": 2.062315789473684, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 39286 + }, + { + "epoch": 2.062375939849624, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 39287 + }, + { + "epoch": 2.062436090225564, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 39288 + }, + { + "epoch": 2.062496240601504, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 39289 + }, + { + "epoch": 2.0625563909774436, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 39290 + }, + { + "epoch": 2.0626165413533837, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 39291 + }, + { + "epoch": 2.0626766917293233, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 39292 + }, + { + "epoch": 2.062736842105263, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 39293 + }, + { + "epoch": 2.062796992481203, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 39294 + }, + { + "epoch": 2.0628571428571427, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 39295 + }, + { + "epoch": 2.062917293233083, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 39296 + }, + { + "epoch": 2.0629774436090225, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 39297 + }, + { + "epoch": 2.0630375939849626, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 39298 + }, + { + "epoch": 2.063097744360902, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 39299 + }, + { + "epoch": 2.0631578947368423, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 39300 + }, + { + "epoch": 2.063218045112782, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 39301 + }, + { + "epoch": 2.0632781954887216, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 39302 + }, + { + "epoch": 2.0633383458646617, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 39303 + }, + { + "epoch": 2.0633984962406013, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 39304 + }, + { + "epoch": 2.0634586466165414, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 39305 + }, + { + "epoch": 2.063518796992481, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 39306 + }, + { + "epoch": 2.063578947368421, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 39307 + }, + { + "epoch": 2.063639097744361, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 39308 + }, + { + "epoch": 2.063699248120301, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 39309 + }, + { + "epoch": 2.0637593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 39310 + }, + { + "epoch": 2.0638195488721807, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 39311 + }, + { + "epoch": 2.0638796992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 39312 + }, + { + "epoch": 2.06393984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 39313 + }, + { + "epoch": 2.064, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 39314 + }, + { + "epoch": 2.0640601503759397, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 39315 + }, + { + "epoch": 2.06412030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 39316 + }, + { + "epoch": 2.0641804511278194, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 39317 + }, + { + "epoch": 2.0642406015037595, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 39318 + }, + { + "epoch": 2.064300751879699, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 39319 + }, + { + "epoch": 2.0643609022556393, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 39320 + }, + { + "epoch": 2.064421052631579, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 39321 + }, + { + "epoch": 2.0644812030075186, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 39322 + }, + { + "epoch": 2.0645413533834587, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 39323 + }, + { + "epoch": 2.0646015037593983, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 39324 + }, + { + "epoch": 2.0646616541353384, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1725, + "step": 39325 + }, + { + "epoch": 2.064721804511278, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 39326 + }, + { + "epoch": 2.064781954887218, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.917, + "step": 39327 + }, + { + "epoch": 2.064842105263158, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 39328 + }, + { + "epoch": 2.064902255639098, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 39329 + }, + { + "epoch": 2.0649624060150376, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 39330 + }, + { + "epoch": 2.0650225563909776, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.948, + "step": 39331 + }, + { + "epoch": 2.0650827067669173, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 39332 + }, + { + "epoch": 2.065142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 39333 + }, + { + "epoch": 2.065203007518797, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 39334 + }, + { + "epoch": 2.0652631578947367, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 39335 + }, + { + "epoch": 2.065323308270677, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 39336 + }, + { + "epoch": 2.0653834586466164, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 39337 + }, + { + "epoch": 2.0654436090225565, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 39338 + }, + { + "epoch": 2.065503759398496, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 39339 + }, + { + "epoch": 2.0655639097744363, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 39340 + }, + { + "epoch": 2.065624060150376, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 39341 + }, + { + "epoch": 2.0656842105263156, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 39342 + }, + { + "epoch": 2.0657443609022557, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 39343 + }, + { + "epoch": 2.0658045112781953, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 39344 + }, + { + "epoch": 2.0658646616541354, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 39345 + }, + { + "epoch": 2.065924812030075, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 39346 + }, + { + "epoch": 2.065984962406015, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 39347 + }, + { + "epoch": 2.066045112781955, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 39348 + }, + { + "epoch": 2.066105263157895, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 39349 + }, + { + "epoch": 2.0661654135338345, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 39350 + }, + { + "epoch": 2.0662255639097746, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 39351 + }, + { + "epoch": 2.0662857142857143, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 39352 + }, + { + "epoch": 2.066345864661654, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 39353 + }, + { + "epoch": 2.066406015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 39354 + }, + { + "epoch": 2.0664661654135337, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 39355 + }, + { + "epoch": 2.0665263157894738, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 39356 + }, + { + "epoch": 2.0665864661654134, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 39357 + }, + { + "epoch": 2.0666466165413535, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 39358 + }, + { + "epoch": 2.066706766917293, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 39359 + }, + { + "epoch": 2.0667669172932333, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 39360 + }, + { + "epoch": 2.066827067669173, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 39361 + }, + { + "epoch": 2.066887218045113, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9329, + "step": 39362 + }, + { + "epoch": 2.0669473684210526, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9058, + "step": 39363 + }, + { + "epoch": 2.0670075187969923, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 39364 + }, + { + "epoch": 2.0670676691729324, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 39365 + }, + { + "epoch": 2.067127819548872, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 39366 + }, + { + "epoch": 2.067187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 39367 + }, + { + "epoch": 2.067248120300752, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 39368 + }, + { + "epoch": 2.067308270676692, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 39369 + }, + { + "epoch": 2.0673684210526315, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 39370 + }, + { + "epoch": 2.0674285714285716, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 39371 + }, + { + "epoch": 2.0674887218045113, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.9511, + "step": 39372 + }, + { + "epoch": 2.067548872180451, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 39373 + }, + { + "epoch": 2.067609022556391, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 39374 + }, + { + "epoch": 2.0676691729323307, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 39375 + }, + { + "epoch": 2.0677293233082708, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1289, + "step": 39376 + }, + { + "epoch": 2.0677894736842104, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 39377 + }, + { + "epoch": 2.0678496240601505, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 39378 + }, + { + "epoch": 2.06790977443609, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 39379 + }, + { + "epoch": 2.0679699248120302, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 39380 + }, + { + "epoch": 2.06803007518797, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.2076, + "step": 39381 + }, + { + "epoch": 2.06809022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 39382 + }, + { + "epoch": 2.0681503759398496, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 39383 + }, + { + "epoch": 2.0682105263157893, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 39384 + }, + { + "epoch": 2.0682706766917294, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 39385 + }, + { + "epoch": 2.068330827067669, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 39386 + }, + { + "epoch": 2.068390977443609, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 39387 + }, + { + "epoch": 2.0684511278195488, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 39388 + }, + { + "epoch": 2.068511278195489, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 39389 + }, + { + "epoch": 2.0685714285714285, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 39390 + }, + { + "epoch": 2.0686315789473686, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 39391 + }, + { + "epoch": 2.0686917293233082, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 39392 + }, + { + "epoch": 2.0687518796992483, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 39393 + }, + { + "epoch": 2.068812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 39394 + }, + { + "epoch": 2.0688721804511276, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 39395 + }, + { + "epoch": 2.0689323308270677, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 39396 + }, + { + "epoch": 2.0689924812030074, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 39397 + }, + { + "epoch": 2.0690526315789475, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 39398 + }, + { + "epoch": 2.069112781954887, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 39399 + }, + { + "epoch": 2.069172932330827, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 39400 + }, + { + "epoch": 2.069233082706767, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 39401 + }, + { + "epoch": 2.069293233082707, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 39402 + }, + { + "epoch": 2.0693533834586466, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 39403 + }, + { + "epoch": 2.0694135338345863, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 39404 + }, + { + "epoch": 2.0694736842105264, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 39405 + }, + { + "epoch": 2.069533834586466, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 39406 + }, + { + "epoch": 2.069593984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 39407 + }, + { + "epoch": 2.0696541353383457, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 39408 + }, + { + "epoch": 2.069714285714286, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 39409 + }, + { + "epoch": 2.0697744360902255, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 39410 + }, + { + "epoch": 2.0698345864661656, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 39411 + }, + { + "epoch": 2.0698947368421052, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 39412 + }, + { + "epoch": 2.0699548872180453, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 39413 + }, + { + "epoch": 2.070015037593985, + "grad_norm": 1.5390625, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 39414 + }, + { + "epoch": 2.0700751879699246, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 39415 + }, + { + "epoch": 2.0701353383458647, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 39416 + }, + { + "epoch": 2.0701954887218044, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 39417 + }, + { + "epoch": 2.0702556390977445, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 39418 + }, + { + "epoch": 2.070315789473684, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 39419 + }, + { + "epoch": 2.070375939849624, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 39420 + }, + { + "epoch": 2.070436090225564, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9348, + "step": 39421 + }, + { + "epoch": 2.070496240601504, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9364, + "step": 39422 + }, + { + "epoch": 2.0705563909774436, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 39423 + }, + { + "epoch": 2.0706165413533832, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 39424 + }, + { + "epoch": 2.0706766917293233, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 39425 + }, + { + "epoch": 2.070736842105263, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 39426 + }, + { + "epoch": 2.070796992481203, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 39427 + }, + { + "epoch": 2.0708571428571427, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 39428 + }, + { + "epoch": 2.070917293233083, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 39429 + }, + { + "epoch": 2.0709774436090225, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9414, + "step": 39430 + }, + { + "epoch": 2.0710375939849626, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 39431 + }, + { + "epoch": 2.071097744360902, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 39432 + }, + { + "epoch": 2.0711578947368423, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 39433 + }, + { + "epoch": 2.071218045112782, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9376, + "step": 39434 + }, + { + "epoch": 2.0712781954887216, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 39435 + }, + { + "epoch": 2.0713383458646617, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 39436 + }, + { + "epoch": 2.0713984962406014, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 39437 + }, + { + "epoch": 2.0714586466165414, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 39438 + }, + { + "epoch": 2.071518796992481, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 39439 + }, + { + "epoch": 2.071578947368421, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 39440 + }, + { + "epoch": 2.071639097744361, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 39441 + }, + { + "epoch": 2.071699248120301, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 39442 + }, + { + "epoch": 2.0717593984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 39443 + }, + { + "epoch": 2.0718195488721802, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 39444 + }, + { + "epoch": 2.0718796992481203, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 39445 + }, + { + "epoch": 2.07193984962406, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 39446 + }, + { + "epoch": 2.072, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 39447 + }, + { + "epoch": 2.0720601503759397, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 39448 + }, + { + "epoch": 2.07212030075188, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 39449 + }, + { + "epoch": 2.0721804511278195, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 39450 + }, + { + "epoch": 2.0722406015037595, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 39451 + }, + { + "epoch": 2.072300751879699, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 39452 + }, + { + "epoch": 2.0723609022556393, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 39453 + }, + { + "epoch": 2.072421052631579, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 39454 + }, + { + "epoch": 2.0724812030075186, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 39455 + }, + { + "epoch": 2.0725413533834587, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 39456 + }, + { + "epoch": 2.0726015037593983, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 39457 + }, + { + "epoch": 2.0726616541353384, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 39458 + }, + { + "epoch": 2.072721804511278, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1115, + "step": 39459 + }, + { + "epoch": 2.072781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 39460 + }, + { + "epoch": 2.072842105263158, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 39461 + }, + { + "epoch": 2.072902255639098, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 39462 + }, + { + "epoch": 2.0729624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1167, + "step": 39463 + }, + { + "epoch": 2.0730225563909777, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 39464 + }, + { + "epoch": 2.0730827067669173, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 39465 + }, + { + "epoch": 2.073142857142857, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 39466 + }, + { + "epoch": 2.073203007518797, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 39467 + }, + { + "epoch": 2.0732631578947367, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 39468 + }, + { + "epoch": 2.073323308270677, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 39469 + }, + { + "epoch": 2.0733834586466164, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9298, + "step": 39470 + }, + { + "epoch": 2.0734436090225565, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 39471 + }, + { + "epoch": 2.073503759398496, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 39472 + }, + { + "epoch": 2.0735639097744363, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 39473 + }, + { + "epoch": 2.073624060150376, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 39474 + }, + { + "epoch": 2.0736842105263156, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 39475 + }, + { + "epoch": 2.0737443609022557, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 39476 + }, + { + "epoch": 2.0738045112781953, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 39477 + }, + { + "epoch": 2.0738646616541354, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 39478 + }, + { + "epoch": 2.073924812030075, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 39479 + }, + { + "epoch": 2.073984962406015, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 39480 + }, + { + "epoch": 2.074045112781955, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 39481 + }, + { + "epoch": 2.074105263157895, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 39482 + }, + { + "epoch": 2.0741654135338345, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 39483 + }, + { + "epoch": 2.0742255639097746, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 39484 + }, + { + "epoch": 2.0742857142857143, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 39485 + }, + { + "epoch": 2.074345864661654, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 39486 + }, + { + "epoch": 2.074406015037594, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 39487 + }, + { + "epoch": 2.0744661654135337, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 39488 + }, + { + "epoch": 2.0745263157894738, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9344, + "step": 39489 + }, + { + "epoch": 2.0745864661654134, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 39490 + }, + { + "epoch": 2.0746466165413535, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 39491 + }, + { + "epoch": 2.074706766917293, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 39492 + }, + { + "epoch": 2.0747669172932333, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9303, + "step": 39493 + }, + { + "epoch": 2.074827067669173, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 39494 + }, + { + "epoch": 2.074887218045113, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 39495 + }, + { + "epoch": 2.0749473684210527, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 39496 + }, + { + "epoch": 2.0750075187969923, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 39497 + }, + { + "epoch": 2.0750676691729324, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 39498 + }, + { + "epoch": 2.075127819548872, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 39499 + }, + { + "epoch": 2.075187969924812, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1548, + "step": 39500 + }, + { + "epoch": 2.075248120300752, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 39501 + }, + { + "epoch": 2.075308270676692, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 39502 + }, + { + "epoch": 2.0753684210526315, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 39503 + }, + { + "epoch": 2.0754285714285716, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 39504 + }, + { + "epoch": 2.0754887218045113, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 39505 + }, + { + "epoch": 2.075548872180451, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 39506 + }, + { + "epoch": 2.075609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 39507 + }, + { + "epoch": 2.0756691729323307, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 39508 + }, + { + "epoch": 2.0757293233082708, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 39509 + }, + { + "epoch": 2.0757894736842104, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 39510 + }, + { + "epoch": 2.0758496240601505, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 39511 + }, + { + "epoch": 2.07590977443609, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 39512 + }, + { + "epoch": 2.0759699248120302, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9323, + "step": 39513 + }, + { + "epoch": 2.07603007518797, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 39514 + }, + { + "epoch": 2.07609022556391, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 39515 + }, + { + "epoch": 2.0761503759398496, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9438, + "step": 39516 + }, + { + "epoch": 2.0762105263157893, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 39517 + }, + { + "epoch": 2.0762706766917294, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 39518 + }, + { + "epoch": 2.076330827067669, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 39519 + }, + { + "epoch": 2.076390977443609, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 39520 + }, + { + "epoch": 2.0764511278195488, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 39521 + }, + { + "epoch": 2.076511278195489, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1501, + "step": 39522 + }, + { + "epoch": 2.0765714285714285, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9343, + "step": 39523 + }, + { + "epoch": 2.0766315789473686, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 39524 + }, + { + "epoch": 2.0766917293233083, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 39525 + }, + { + "epoch": 2.0767518796992483, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 39526 + }, + { + "epoch": 2.076812030075188, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 39527 + }, + { + "epoch": 2.0768721804511276, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 39528 + }, + { + "epoch": 2.0769323308270677, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 39529 + }, + { + "epoch": 2.0769924812030074, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 39530 + }, + { + "epoch": 2.0770526315789475, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9427, + "step": 39531 + }, + { + "epoch": 2.077112781954887, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 39532 + }, + { + "epoch": 2.0771729323308272, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 39533 + }, + { + "epoch": 2.077233082706767, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 39534 + }, + { + "epoch": 2.077293233082707, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 39535 + }, + { + "epoch": 2.0773533834586466, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 39536 + }, + { + "epoch": 2.0774135338345863, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 39537 + }, + { + "epoch": 2.0774736842105264, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 39538 + }, + { + "epoch": 2.077533834586466, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 39539 + }, + { + "epoch": 2.077593984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 39540 + }, + { + "epoch": 2.0776541353383458, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 39541 + }, + { + "epoch": 2.077714285714286, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 39542 + }, + { + "epoch": 2.0777744360902255, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9499, + "step": 39543 + }, + { + "epoch": 2.0778345864661656, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 39544 + }, + { + "epoch": 2.0778947368421052, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 39545 + }, + { + "epoch": 2.077954887218045, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 39546 + }, + { + "epoch": 2.078015037593985, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 39547 + }, + { + "epoch": 2.0780751879699246, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 39548 + }, + { + "epoch": 2.0781353383458647, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 39549 + }, + { + "epoch": 2.0781954887218044, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 39550 + }, + { + "epoch": 2.0782556390977445, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 39551 + }, + { + "epoch": 2.078315789473684, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 39552 + }, + { + "epoch": 2.078375939849624, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 39553 + }, + { + "epoch": 2.078436090225564, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 39554 + }, + { + "epoch": 2.078496240601504, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 39555 + }, + { + "epoch": 2.0785563909774436, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 39556 + }, + { + "epoch": 2.0786165413533833, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 39557 + }, + { + "epoch": 2.0786766917293233, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 39558 + }, + { + "epoch": 2.078736842105263, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 39559 + }, + { + "epoch": 2.078796992481203, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 39560 + }, + { + "epoch": 2.0788571428571427, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 39561 + }, + { + "epoch": 2.078917293233083, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 39562 + }, + { + "epoch": 2.0789774436090225, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 39563 + }, + { + "epoch": 2.0790375939849626, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 39564 + }, + { + "epoch": 2.0790977443609022, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 39565 + }, + { + "epoch": 2.0791578947368423, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 39566 + }, + { + "epoch": 2.079218045112782, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 39567 + }, + { + "epoch": 2.0792781954887216, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 39568 + }, + { + "epoch": 2.0793383458646617, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 39569 + }, + { + "epoch": 2.0793984962406014, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 39570 + }, + { + "epoch": 2.0794586466165415, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 39571 + }, + { + "epoch": 2.079518796992481, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 39572 + }, + { + "epoch": 2.079578947368421, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 39573 + }, + { + "epoch": 2.079639097744361, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 39574 + }, + { + "epoch": 2.079699248120301, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 39575 + }, + { + "epoch": 2.0797593984962406, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 39576 + }, + { + "epoch": 2.0798195488721802, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 39577 + }, + { + "epoch": 2.0798796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 39578 + }, + { + "epoch": 2.07993984962406, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 39579 + }, + { + "epoch": 2.08, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 39580 + }, + { + "epoch": 2.0800601503759397, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 39581 + }, + { + "epoch": 2.08012030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 39582 + }, + { + "epoch": 2.0801804511278195, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 39583 + }, + { + "epoch": 2.0802406015037596, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 39584 + }, + { + "epoch": 2.080300751879699, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 39585 + }, + { + "epoch": 2.0803609022556393, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 39586 + }, + { + "epoch": 2.080421052631579, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 39587 + }, + { + "epoch": 2.0804812030075186, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 39588 + }, + { + "epoch": 2.0805413533834587, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 39589 + }, + { + "epoch": 2.0806015037593983, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 39590 + }, + { + "epoch": 2.0806616541353384, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 39591 + }, + { + "epoch": 2.080721804511278, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 39592 + }, + { + "epoch": 2.080781954887218, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 39593 + }, + { + "epoch": 2.080842105263158, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 39594 + }, + { + "epoch": 2.080902255639098, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 39595 + }, + { + "epoch": 2.0809624060150376, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 39596 + }, + { + "epoch": 2.0810225563909777, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9386, + "step": 39597 + }, + { + "epoch": 2.0810827067669173, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 39598 + }, + { + "epoch": 2.081142857142857, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 39599 + }, + { + "epoch": 2.081203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 39600 + }, + { + "epoch": 2.0812631578947367, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 39601 + }, + { + "epoch": 2.081323308270677, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 39602 + }, + { + "epoch": 2.0813834586466164, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 39603 + }, + { + "epoch": 2.0814436090225565, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 39604 + }, + { + "epoch": 2.081503759398496, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 39605 + }, + { + "epoch": 2.0815639097744363, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 39606 + }, + { + "epoch": 2.081624060150376, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 39607 + }, + { + "epoch": 2.0816842105263156, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 39608 + }, + { + "epoch": 2.0817443609022557, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 39609 + }, + { + "epoch": 2.0818045112781953, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 39610 + }, + { + "epoch": 2.0818646616541354, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 39611 + }, + { + "epoch": 2.081924812030075, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 39612 + }, + { + "epoch": 2.081984962406015, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 39613 + }, + { + "epoch": 2.082045112781955, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 39614 + }, + { + "epoch": 2.082105263157895, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 39615 + }, + { + "epoch": 2.0821654135338346, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 39616 + }, + { + "epoch": 2.0822255639097746, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 39617 + }, + { + "epoch": 2.0822857142857143, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.307, + "step": 39618 + }, + { + "epoch": 2.082345864661654, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 39619 + }, + { + "epoch": 2.082406015037594, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 39620 + }, + { + "epoch": 2.0824661654135337, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 39621 + }, + { + "epoch": 2.082526315789474, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 39622 + }, + { + "epoch": 2.0825864661654134, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 39623 + }, + { + "epoch": 2.0826466165413535, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 39624 + }, + { + "epoch": 2.082706766917293, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 39625 + }, + { + "epoch": 2.0827669172932333, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 39626 + }, + { + "epoch": 2.082827067669173, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 39627 + }, + { + "epoch": 2.082887218045113, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 39628 + }, + { + "epoch": 2.0829473684210527, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 39629 + }, + { + "epoch": 2.0830075187969923, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 39630 + }, + { + "epoch": 2.0830676691729324, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 39631 + }, + { + "epoch": 2.083127819548872, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 39632 + }, + { + "epoch": 2.083187969924812, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 39633 + }, + { + "epoch": 2.083248120300752, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 39634 + }, + { + "epoch": 2.083308270676692, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 39635 + }, + { + "epoch": 2.0833684210526315, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 39636 + }, + { + "epoch": 2.0834285714285716, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 39637 + }, + { + "epoch": 2.0834887218045113, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 39638 + }, + { + "epoch": 2.083548872180451, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 39639 + }, + { + "epoch": 2.083609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 39640 + }, + { + "epoch": 2.0836691729323307, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 39641 + }, + { + "epoch": 2.0837293233082708, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 39642 + }, + { + "epoch": 2.0837894736842104, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 39643 + }, + { + "epoch": 2.0838496240601505, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 39644 + }, + { + "epoch": 2.08390977443609, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 39645 + }, + { + "epoch": 2.0839699248120302, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 39646 + }, + { + "epoch": 2.08403007518797, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 39647 + }, + { + "epoch": 2.0840902255639095, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 39648 + }, + { + "epoch": 2.0841503759398496, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 39649 + }, + { + "epoch": 2.0842105263157893, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 39650 + }, + { + "epoch": 2.0842706766917294, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 39651 + }, + { + "epoch": 2.084330827067669, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 39652 + }, + { + "epoch": 2.084390977443609, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1773, + "step": 39653 + }, + { + "epoch": 2.0844511278195488, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 39654 + }, + { + "epoch": 2.084511278195489, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 39655 + }, + { + "epoch": 2.0845714285714285, + "grad_norm": 1.8359375, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 39656 + }, + { + "epoch": 2.0846315789473686, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 39657 + }, + { + "epoch": 2.0846917293233083, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 39658 + }, + { + "epoch": 2.084751879699248, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 39659 + }, + { + "epoch": 2.084812030075188, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 39660 + }, + { + "epoch": 2.0848721804511277, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 39661 + }, + { + "epoch": 2.0849323308270677, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 39662 + }, + { + "epoch": 2.0849924812030074, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 39663 + }, + { + "epoch": 2.0850526315789475, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 39664 + }, + { + "epoch": 2.085112781954887, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 39665 + }, + { + "epoch": 2.0851729323308272, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 39666 + }, + { + "epoch": 2.085233082706767, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 39667 + }, + { + "epoch": 2.085293233082707, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 39668 + }, + { + "epoch": 2.0853533834586466, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 39669 + }, + { + "epoch": 2.0854135338345863, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 39670 + }, + { + "epoch": 2.0854736842105264, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 39671 + }, + { + "epoch": 2.085533834586466, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 39672 + }, + { + "epoch": 2.085593984962406, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.923, + "step": 39673 + }, + { + "epoch": 2.0856541353383458, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 39674 + }, + { + "epoch": 2.085714285714286, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 39675 + }, + { + "epoch": 2.0857744360902255, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 39676 + }, + { + "epoch": 2.0858345864661656, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 39677 + }, + { + "epoch": 2.0858947368421052, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 39678 + }, + { + "epoch": 2.085954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 39679 + }, + { + "epoch": 2.086015037593985, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 39680 + }, + { + "epoch": 2.0860751879699246, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 39681 + }, + { + "epoch": 2.0861353383458647, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 39682 + }, + { + "epoch": 2.0861954887218044, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1449, + "step": 39683 + }, + { + "epoch": 2.0862556390977445, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 39684 + }, + { + "epoch": 2.086315789473684, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 39685 + }, + { + "epoch": 2.086375939849624, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 39686 + }, + { + "epoch": 2.086436090225564, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 39687 + }, + { + "epoch": 2.086496240601504, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 39688 + }, + { + "epoch": 2.0865563909774436, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9514, + "step": 39689 + }, + { + "epoch": 2.0866165413533833, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 39690 + }, + { + "epoch": 2.0866766917293234, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 39691 + }, + { + "epoch": 2.086736842105263, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 39692 + }, + { + "epoch": 2.086796992481203, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 39693 + }, + { + "epoch": 2.0868571428571427, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 39694 + }, + { + "epoch": 2.086917293233083, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 39695 + }, + { + "epoch": 2.0869774436090225, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 39696 + }, + { + "epoch": 2.0870375939849626, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 39697 + }, + { + "epoch": 2.0870977443609022, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 39698 + }, + { + "epoch": 2.0871578947368423, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 39699 + }, + { + "epoch": 2.087218045112782, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 39700 + }, + { + "epoch": 2.0872781954887216, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 39701 + }, + { + "epoch": 2.0873383458646617, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1497, + "step": 39702 + }, + { + "epoch": 2.0873984962406014, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 39703 + }, + { + "epoch": 2.0874586466165415, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 39704 + }, + { + "epoch": 2.087518796992481, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 39705 + }, + { + "epoch": 2.087578947368421, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 39706 + }, + { + "epoch": 2.087639097744361, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 39707 + }, + { + "epoch": 2.087699248120301, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 39708 + }, + { + "epoch": 2.0877593984962406, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 39709 + }, + { + "epoch": 2.0878195488721802, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 39710 + }, + { + "epoch": 2.0878796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 39711 + }, + { + "epoch": 2.08793984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 39712 + }, + { + "epoch": 2.088, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 39713 + }, + { + "epoch": 2.0880601503759397, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 39714 + }, + { + "epoch": 2.08812030075188, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 39715 + }, + { + "epoch": 2.0881804511278195, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 39716 + }, + { + "epoch": 2.0882406015037596, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 39717 + }, + { + "epoch": 2.088300751879699, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 39718 + }, + { + "epoch": 2.0883609022556393, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 39719 + }, + { + "epoch": 2.088421052631579, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 39720 + }, + { + "epoch": 2.0884812030075186, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 39721 + }, + { + "epoch": 2.0885413533834587, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 39722 + }, + { + "epoch": 2.0886015037593983, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 39723 + }, + { + "epoch": 2.0886616541353384, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 39724 + }, + { + "epoch": 2.088721804511278, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 39725 + }, + { + "epoch": 2.088781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 39726 + }, + { + "epoch": 2.088842105263158, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 39727 + }, + { + "epoch": 2.088902255639098, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 39728 + }, + { + "epoch": 2.0889624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 39729 + }, + { + "epoch": 2.0890225563909777, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 39730 + }, + { + "epoch": 2.0890827067669173, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9527, + "step": 39731 + }, + { + "epoch": 2.089142857142857, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 39732 + }, + { + "epoch": 2.089203007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 39733 + }, + { + "epoch": 2.0892631578947367, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 39734 + }, + { + "epoch": 2.089323308270677, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 39735 + }, + { + "epoch": 2.0893834586466165, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 39736 + }, + { + "epoch": 2.0894436090225565, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 39737 + }, + { + "epoch": 2.089503759398496, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9211, + "step": 39738 + }, + { + "epoch": 2.0895639097744363, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 39739 + }, + { + "epoch": 2.089624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 39740 + }, + { + "epoch": 2.0896842105263156, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 39741 + }, + { + "epoch": 2.0897443609022557, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 39742 + }, + { + "epoch": 2.0898045112781953, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 39743 + }, + { + "epoch": 2.0898646616541354, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 39744 + }, + { + "epoch": 2.089924812030075, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 39745 + }, + { + "epoch": 2.089984962406015, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 39746 + }, + { + "epoch": 2.090045112781955, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 39747 + }, + { + "epoch": 2.090105263157895, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 39748 + }, + { + "epoch": 2.0901654135338346, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 39749 + }, + { + "epoch": 2.090225563909774, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 39750 + }, + { + "epoch": 2.0902857142857143, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 39751 + }, + { + "epoch": 2.090345864661654, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 39752 + }, + { + "epoch": 2.090406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 39753 + }, + { + "epoch": 2.0904661654135337, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 39754 + }, + { + "epoch": 2.090526315789474, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 39755 + }, + { + "epoch": 2.0905864661654134, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 39756 + }, + { + "epoch": 2.0906466165413535, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 39757 + }, + { + "epoch": 2.090706766917293, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 39758 + }, + { + "epoch": 2.0907669172932333, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 39759 + }, + { + "epoch": 2.090827067669173, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 39760 + }, + { + "epoch": 2.0908872180451126, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 39761 + }, + { + "epoch": 2.0909473684210527, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 39762 + }, + { + "epoch": 2.0910075187969923, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 39763 + }, + { + "epoch": 2.0910676691729324, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 39764 + }, + { + "epoch": 2.091127819548872, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 39765 + }, + { + "epoch": 2.091187969924812, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 39766 + }, + { + "epoch": 2.091248120300752, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 39767 + }, + { + "epoch": 2.091308270676692, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 39768 + }, + { + "epoch": 2.0913684210526315, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 39769 + }, + { + "epoch": 2.0914285714285716, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 39770 + }, + { + "epoch": 2.0914887218045113, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 39771 + }, + { + "epoch": 2.091548872180451, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9222, + "step": 39772 + }, + { + "epoch": 2.091609022556391, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 39773 + }, + { + "epoch": 2.0916691729323307, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 39774 + }, + { + "epoch": 2.0917293233082708, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 39775 + }, + { + "epoch": 2.0917894736842104, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 39776 + }, + { + "epoch": 2.0918496240601505, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 39777 + }, + { + "epoch": 2.09190977443609, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 39778 + }, + { + "epoch": 2.0919699248120303, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9259, + "step": 39779 + }, + { + "epoch": 2.09203007518797, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9131, + "step": 39780 + }, + { + "epoch": 2.0920902255639096, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 39781 + }, + { + "epoch": 2.0921503759398496, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 39782 + }, + { + "epoch": 2.0922105263157893, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 39783 + }, + { + "epoch": 2.0922706766917294, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 39784 + }, + { + "epoch": 2.092330827067669, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 39785 + }, + { + "epoch": 2.092390977443609, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 39786 + }, + { + "epoch": 2.092451127819549, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 39787 + }, + { + "epoch": 2.092511278195489, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 39788 + }, + { + "epoch": 2.0925714285714285, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 39789 + }, + { + "epoch": 2.0926315789473686, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 39790 + }, + { + "epoch": 2.0926917293233083, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 39791 + }, + { + "epoch": 2.092751879699248, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 39792 + }, + { + "epoch": 2.092812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 39793 + }, + { + "epoch": 2.0928721804511277, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 39794 + }, + { + "epoch": 2.0929323308270678, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 39795 + }, + { + "epoch": 2.0929924812030074, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 39796 + }, + { + "epoch": 2.0930526315789475, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 39797 + }, + { + "epoch": 2.093112781954887, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 39798 + }, + { + "epoch": 2.0931729323308272, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 39799 + }, + { + "epoch": 2.093233082706767, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 39800 + }, + { + "epoch": 2.093293233082707, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 39801 + }, + { + "epoch": 2.0933533834586466, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 39802 + }, + { + "epoch": 2.0934135338345863, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1124, + "step": 39803 + }, + { + "epoch": 2.0934736842105264, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 39804 + }, + { + "epoch": 2.093533834586466, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 39805 + }, + { + "epoch": 2.093593984962406, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 39806 + }, + { + "epoch": 2.0936541353383458, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 39807 + }, + { + "epoch": 2.093714285714286, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 39808 + }, + { + "epoch": 2.0937744360902255, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 39809 + }, + { + "epoch": 2.0938345864661656, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 39810 + }, + { + "epoch": 2.0938947368421053, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 39811 + }, + { + "epoch": 2.093954887218045, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 39812 + }, + { + "epoch": 2.094015037593985, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 39813 + }, + { + "epoch": 2.0940751879699246, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 39814 + }, + { + "epoch": 2.0941353383458647, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 39815 + }, + { + "epoch": 2.0941954887218044, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 39816 + }, + { + "epoch": 2.0942556390977445, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 39817 + }, + { + "epoch": 2.094315789473684, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 39818 + }, + { + "epoch": 2.0943759398496242, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 39819 + }, + { + "epoch": 2.094436090225564, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9459, + "step": 39820 + }, + { + "epoch": 2.094496240601504, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 0.9281, + "step": 39821 + }, + { + "epoch": 2.0945563909774436, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 39822 + }, + { + "epoch": 2.0946165413533833, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 39823 + }, + { + "epoch": 2.0946766917293234, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 39824 + }, + { + "epoch": 2.094736842105263, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 39825 + }, + { + "epoch": 2.094796992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 39826 + }, + { + "epoch": 2.0948571428571428, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 39827 + }, + { + "epoch": 2.094917293233083, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 39828 + }, + { + "epoch": 2.0949774436090225, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 39829 + }, + { + "epoch": 2.0950375939849626, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 39830 + }, + { + "epoch": 2.0950977443609022, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 39831 + }, + { + "epoch": 2.0951578947368423, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 39832 + }, + { + "epoch": 2.095218045112782, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 39833 + }, + { + "epoch": 2.0952781954887216, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 39834 + }, + { + "epoch": 2.0953383458646617, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 39835 + }, + { + "epoch": 2.0953984962406014, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 39836 + }, + { + "epoch": 2.0954586466165415, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 39837 + }, + { + "epoch": 2.095518796992481, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 39838 + }, + { + "epoch": 2.095578947368421, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 39839 + }, + { + "epoch": 2.095639097744361, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 39840 + }, + { + "epoch": 2.095699248120301, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 39841 + }, + { + "epoch": 2.0957593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 39842 + }, + { + "epoch": 2.0958195488721802, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 39843 + }, + { + "epoch": 2.0958796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 39844 + }, + { + "epoch": 2.09593984962406, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 39845 + }, + { + "epoch": 2.096, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 39846 + }, + { + "epoch": 2.0960601503759397, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 39847 + }, + { + "epoch": 2.09612030075188, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 39848 + }, + { + "epoch": 2.0961804511278195, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 39849 + }, + { + "epoch": 2.0962406015037596, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 39850 + }, + { + "epoch": 2.096300751879699, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 39851 + }, + { + "epoch": 2.096360902255639, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 39852 + }, + { + "epoch": 2.096421052631579, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 39853 + }, + { + "epoch": 2.0964812030075186, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 39854 + }, + { + "epoch": 2.0965413533834587, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 39855 + }, + { + "epoch": 2.0966015037593984, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 39856 + }, + { + "epoch": 2.0966616541353384, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 39857 + }, + { + "epoch": 2.096721804511278, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 39858 + }, + { + "epoch": 2.096781954887218, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 39859 + }, + { + "epoch": 2.096842105263158, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9241, + "step": 39860 + }, + { + "epoch": 2.096902255639098, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 39861 + }, + { + "epoch": 2.0969624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 39862 + }, + { + "epoch": 2.0970225563909772, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 39863 + }, + { + "epoch": 2.0970827067669173, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 39864 + }, + { + "epoch": 2.097142857142857, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 39865 + }, + { + "epoch": 2.097203007518797, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 39866 + }, + { + "epoch": 2.0972631578947367, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2258, + "step": 39867 + }, + { + "epoch": 2.097323308270677, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 39868 + }, + { + "epoch": 2.0973834586466165, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 39869 + }, + { + "epoch": 2.0974436090225566, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 39870 + }, + { + "epoch": 2.097503759398496, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 39871 + }, + { + "epoch": 2.0975639097744363, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 39872 + }, + { + "epoch": 2.097624060150376, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 39873 + }, + { + "epoch": 2.0976842105263156, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 39874 + }, + { + "epoch": 2.0977443609022557, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 39875 + }, + { + "epoch": 2.0978045112781953, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 39876 + }, + { + "epoch": 2.0978646616541354, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 39877 + }, + { + "epoch": 2.097924812030075, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 0.9064, + "step": 39878 + }, + { + "epoch": 2.097984962406015, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 39879 + }, + { + "epoch": 2.098045112781955, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 39880 + }, + { + "epoch": 2.098105263157895, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 39881 + }, + { + "epoch": 2.0981654135338346, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 39882 + }, + { + "epoch": 2.098225563909774, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 39883 + }, + { + "epoch": 2.0982857142857143, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 39884 + }, + { + "epoch": 2.098345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 39885 + }, + { + "epoch": 2.098406015037594, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 39886 + }, + { + "epoch": 2.0984661654135337, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 39887 + }, + { + "epoch": 2.098526315789474, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 39888 + }, + { + "epoch": 2.0985864661654134, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 39889 + }, + { + "epoch": 2.0986466165413535, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 39890 + }, + { + "epoch": 2.098706766917293, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 39891 + }, + { + "epoch": 2.0987669172932333, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 39892 + }, + { + "epoch": 2.098827067669173, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 39893 + }, + { + "epoch": 2.0988872180451126, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 39894 + }, + { + "epoch": 2.0989473684210527, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 39895 + }, + { + "epoch": 2.0990075187969923, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1735, + "step": 39896 + }, + { + "epoch": 2.0990676691729324, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 39897 + }, + { + "epoch": 2.099127819548872, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 39898 + }, + { + "epoch": 2.099187969924812, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 39899 + }, + { + "epoch": 2.099248120300752, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 39900 + }, + { + "epoch": 2.099308270676692, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 39901 + }, + { + "epoch": 2.0993684210526316, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 39902 + }, + { + "epoch": 2.0994285714285716, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 39903 + }, + { + "epoch": 2.0994887218045113, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 39904 + }, + { + "epoch": 2.099548872180451, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 39905 + }, + { + "epoch": 2.099609022556391, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 39906 + }, + { + "epoch": 2.0996691729323307, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 39907 + }, + { + "epoch": 2.099729323308271, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 39908 + }, + { + "epoch": 2.0997894736842104, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 39909 + }, + { + "epoch": 2.0998496240601505, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 39910 + }, + { + "epoch": 2.09990977443609, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 39911 + }, + { + "epoch": 2.0999699248120303, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 39912 + }, + { + "epoch": 2.10003007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 39913 + }, + { + "epoch": 2.1000902255639096, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 39914 + }, + { + "epoch": 2.1001503759398497, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 39915 + }, + { + "epoch": 2.1002105263157893, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 39916 + }, + { + "epoch": 2.1002706766917294, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 39917 + }, + { + "epoch": 2.100330827067669, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 39918 + }, + { + "epoch": 2.100390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 39919 + }, + { + "epoch": 2.100451127819549, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 39920 + }, + { + "epoch": 2.100511278195489, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 39921 + }, + { + "epoch": 2.1005714285714285, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 39922 + }, + { + "epoch": 2.1006315789473686, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 39923 + }, + { + "epoch": 2.1006917293233083, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 39924 + }, + { + "epoch": 2.100751879699248, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 39925 + }, + { + "epoch": 2.100812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9332, + "step": 39926 + }, + { + "epoch": 2.1008721804511277, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 39927 + }, + { + "epoch": 2.1009323308270678, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 39928 + }, + { + "epoch": 2.1009924812030074, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 39929 + }, + { + "epoch": 2.1010526315789475, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 39930 + }, + { + "epoch": 2.101112781954887, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 39931 + }, + { + "epoch": 2.1011729323308272, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 39932 + }, + { + "epoch": 2.101233082706767, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 39933 + }, + { + "epoch": 2.101293233082707, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 39934 + }, + { + "epoch": 2.1013533834586466, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 39935 + }, + { + "epoch": 2.1014135338345863, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 39936 + }, + { + "epoch": 2.1014736842105264, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 39937 + }, + { + "epoch": 2.101533834586466, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 39938 + }, + { + "epoch": 2.101593984962406, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 39939 + }, + { + "epoch": 2.1016541353383458, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 39940 + }, + { + "epoch": 2.101714285714286, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9195, + "step": 39941 + }, + { + "epoch": 2.1017744360902255, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 39942 + }, + { + "epoch": 2.1018345864661656, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 39943 + }, + { + "epoch": 2.1018947368421053, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 39944 + }, + { + "epoch": 2.101954887218045, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9276, + "step": 39945 + }, + { + "epoch": 2.102015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 39946 + }, + { + "epoch": 2.1020751879699247, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 39947 + }, + { + "epoch": 2.1021353383458647, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 39948 + }, + { + "epoch": 2.1021954887218044, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 39949 + }, + { + "epoch": 2.1022556390977445, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 39950 + }, + { + "epoch": 2.102315789473684, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 39951 + }, + { + "epoch": 2.1023759398496242, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 39952 + }, + { + "epoch": 2.102436090225564, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 39953 + }, + { + "epoch": 2.102496240601504, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 39954 + }, + { + "epoch": 2.1025563909774436, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1621, + "step": 39955 + }, + { + "epoch": 2.1026165413533833, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 39956 + }, + { + "epoch": 2.1026766917293234, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 39957 + }, + { + "epoch": 2.102736842105263, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 39958 + }, + { + "epoch": 2.102796992481203, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 39959 + }, + { + "epoch": 2.1028571428571428, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 39960 + }, + { + "epoch": 2.102917293233083, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 39961 + }, + { + "epoch": 2.1029774436090225, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 39962 + }, + { + "epoch": 2.1030375939849626, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 39963 + }, + { + "epoch": 2.1030977443609022, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1725, + "step": 39964 + }, + { + "epoch": 2.103157894736842, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 39965 + }, + { + "epoch": 2.103218045112782, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 39966 + }, + { + "epoch": 2.1032781954887216, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.2658, + "step": 39967 + }, + { + "epoch": 2.1033383458646617, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 39968 + }, + { + "epoch": 2.1033984962406014, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 39969 + }, + { + "epoch": 2.1034586466165415, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 39970 + }, + { + "epoch": 2.103518796992481, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 39971 + }, + { + "epoch": 2.103578947368421, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 39972 + }, + { + "epoch": 2.103639097744361, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9211, + "step": 39973 + }, + { + "epoch": 2.103699248120301, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 39974 + }, + { + "epoch": 2.1037593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 39975 + }, + { + "epoch": 2.1038195488721803, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 39976 + }, + { + "epoch": 2.1038796992481203, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 39977 + }, + { + "epoch": 2.10393984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 39978 + }, + { + "epoch": 2.104, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 39979 + }, + { + "epoch": 2.1040601503759397, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 39980 + }, + { + "epoch": 2.10412030075188, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 39981 + }, + { + "epoch": 2.1041804511278195, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 39982 + }, + { + "epoch": 2.1042406015037596, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 39983 + }, + { + "epoch": 2.1043007518796992, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 39984 + }, + { + "epoch": 2.104360902255639, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 39985 + }, + { + "epoch": 2.104421052631579, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 39986 + }, + { + "epoch": 2.1044812030075186, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 39987 + }, + { + "epoch": 2.1045413533834587, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 39988 + }, + { + "epoch": 2.1046015037593984, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 39989 + }, + { + "epoch": 2.1046616541353385, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 39990 + }, + { + "epoch": 2.104721804511278, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 39991 + }, + { + "epoch": 2.104781954887218, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1235, + "step": 39992 + }, + { + "epoch": 2.104842105263158, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 39993 + }, + { + "epoch": 2.104902255639098, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 39994 + }, + { + "epoch": 2.1049624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 39995 + }, + { + "epoch": 2.1050225563909772, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 39996 + }, + { + "epoch": 2.1050827067669173, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1213, + "step": 39997 + }, + { + "epoch": 2.105142857142857, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 39998 + }, + { + "epoch": 2.105203007518797, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 39999 + }, + { + "epoch": 2.1052631578947367, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 40000 + }, + { + "epoch": 2.105323308270677, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 40001 + }, + { + "epoch": 2.1053834586466165, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9518, + "step": 40002 + }, + { + "epoch": 2.1054436090225566, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 40003 + }, + { + "epoch": 2.105503759398496, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 40004 + }, + { + "epoch": 2.1055639097744363, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 40005 + }, + { + "epoch": 2.105624060150376, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 40006 + }, + { + "epoch": 2.1056842105263156, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 40007 + }, + { + "epoch": 2.1057443609022557, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 40008 + }, + { + "epoch": 2.1058045112781953, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 40009 + }, + { + "epoch": 2.1058646616541354, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 40010 + }, + { + "epoch": 2.105924812030075, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 40011 + }, + { + "epoch": 2.105984962406015, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 40012 + }, + { + "epoch": 2.106045112781955, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 40013 + }, + { + "epoch": 2.106105263157895, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 40014 + }, + { + "epoch": 2.1061654135338346, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 40015 + }, + { + "epoch": 2.1062255639097742, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 40016 + }, + { + "epoch": 2.1062857142857143, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 40017 + }, + { + "epoch": 2.106345864661654, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 40018 + }, + { + "epoch": 2.106406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9284, + "step": 40019 + }, + { + "epoch": 2.1064661654135337, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 40020 + }, + { + "epoch": 2.106526315789474, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 40021 + }, + { + "epoch": 2.1065864661654135, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 40022 + }, + { + "epoch": 2.1066466165413535, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 40023 + }, + { + "epoch": 2.106706766917293, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 40024 + }, + { + "epoch": 2.1067669172932333, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 40025 + }, + { + "epoch": 2.106827067669173, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 40026 + }, + { + "epoch": 2.1068872180451126, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9325, + "step": 40027 + }, + { + "epoch": 2.1069473684210527, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 40028 + }, + { + "epoch": 2.1070075187969923, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 40029 + }, + { + "epoch": 2.1070676691729324, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 40030 + }, + { + "epoch": 2.107127819548872, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 40031 + }, + { + "epoch": 2.107187969924812, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.2131, + "step": 40032 + }, + { + "epoch": 2.107248120300752, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 40033 + }, + { + "epoch": 2.107308270676692, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 40034 + }, + { + "epoch": 2.1073684210526316, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 40035 + }, + { + "epoch": 2.1074285714285717, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 40036 + }, + { + "epoch": 2.1074887218045113, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9407, + "step": 40037 + }, + { + "epoch": 2.107548872180451, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 40038 + }, + { + "epoch": 2.107609022556391, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 40039 + }, + { + "epoch": 2.1076691729323307, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 40040 + }, + { + "epoch": 2.107729323308271, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 40041 + }, + { + "epoch": 2.1077894736842104, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 40042 + }, + { + "epoch": 2.1078496240601505, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1262, + "step": 40043 + }, + { + "epoch": 2.10790977443609, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 40044 + }, + { + "epoch": 2.1079699248120303, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 40045 + }, + { + "epoch": 2.10803007518797, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 40046 + }, + { + "epoch": 2.1080902255639096, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 40047 + }, + { + "epoch": 2.1081503759398497, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1098, + "step": 40048 + }, + { + "epoch": 2.1082105263157893, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 40049 + }, + { + "epoch": 2.1082706766917294, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 40050 + }, + { + "epoch": 2.108330827067669, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 40051 + }, + { + "epoch": 2.108390977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 40052 + }, + { + "epoch": 2.108451127819549, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1328, + "step": 40053 + }, + { + "epoch": 2.108511278195489, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9505, + "step": 40054 + }, + { + "epoch": 2.1085714285714285, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 40055 + }, + { + "epoch": 2.1086315789473686, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 40056 + }, + { + "epoch": 2.1086917293233083, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 40057 + }, + { + "epoch": 2.108751879699248, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 40058 + }, + { + "epoch": 2.108812030075188, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 40059 + }, + { + "epoch": 2.1088721804511277, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 40060 + }, + { + "epoch": 2.1089323308270678, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 40061 + }, + { + "epoch": 2.1089924812030074, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 40062 + }, + { + "epoch": 2.1090526315789475, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 40063 + }, + { + "epoch": 2.109112781954887, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 40064 + }, + { + "epoch": 2.1091729323308273, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 40065 + }, + { + "epoch": 2.109233082706767, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 40066 + }, + { + "epoch": 2.1092932330827066, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 40067 + }, + { + "epoch": 2.1093533834586466, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 40068 + }, + { + "epoch": 2.1094135338345863, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 40069 + }, + { + "epoch": 2.1094736842105264, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 40070 + }, + { + "epoch": 2.109533834586466, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 40071 + }, + { + "epoch": 2.109593984962406, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 40072 + }, + { + "epoch": 2.109654135338346, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.2341, + "step": 40073 + }, + { + "epoch": 2.109714285714286, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 40074 + }, + { + "epoch": 2.1097744360902255, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 40075 + }, + { + "epoch": 2.1098345864661656, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 40076 + }, + { + "epoch": 2.1098947368421053, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 40077 + }, + { + "epoch": 2.109954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 40078 + }, + { + "epoch": 2.110015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 40079 + }, + { + "epoch": 2.1100751879699247, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 40080 + }, + { + "epoch": 2.1101353383458648, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 40081 + }, + { + "epoch": 2.1101954887218044, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 40082 + }, + { + "epoch": 2.1102556390977445, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 40083 + }, + { + "epoch": 2.110315789473684, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 40084 + }, + { + "epoch": 2.1103759398496242, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 40085 + }, + { + "epoch": 2.110436090225564, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 40086 + }, + { + "epoch": 2.1104962406015035, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 40087 + }, + { + "epoch": 2.1105563909774436, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 40088 + }, + { + "epoch": 2.1106165413533833, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 40089 + }, + { + "epoch": 2.1106766917293234, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 40090 + }, + { + "epoch": 2.110736842105263, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 40091 + }, + { + "epoch": 2.110796992481203, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 40092 + }, + { + "epoch": 2.1108571428571428, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 40093 + }, + { + "epoch": 2.110917293233083, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 40094 + }, + { + "epoch": 2.1109774436090225, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 40095 + }, + { + "epoch": 2.1110375939849626, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 40096 + }, + { + "epoch": 2.1110977443609023, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 40097 + }, + { + "epoch": 2.111157894736842, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 40098 + }, + { + "epoch": 2.111218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 40099 + }, + { + "epoch": 2.1112781954887216, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 40100 + }, + { + "epoch": 2.1113383458646617, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 40101 + }, + { + "epoch": 2.1113984962406014, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 40102 + }, + { + "epoch": 2.1114586466165415, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.8709, + "step": 40103 + }, + { + "epoch": 2.111518796992481, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 40104 + }, + { + "epoch": 2.111578947368421, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 40105 + }, + { + "epoch": 2.111639097744361, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 40106 + }, + { + "epoch": 2.111699248120301, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 40107 + }, + { + "epoch": 2.1117593984962406, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 40108 + }, + { + "epoch": 2.1118195488721803, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1537, + "step": 40109 + }, + { + "epoch": 2.1118796992481204, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 40110 + }, + { + "epoch": 2.11193984962406, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 40111 + }, + { + "epoch": 2.112, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 40112 + }, + { + "epoch": 2.1120601503759397, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 40113 + }, + { + "epoch": 2.11212030075188, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 40114 + }, + { + "epoch": 2.1121804511278195, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 40115 + }, + { + "epoch": 2.1122406015037596, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 40116 + }, + { + "epoch": 2.1123007518796992, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 40117 + }, + { + "epoch": 2.112360902255639, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.2211, + "step": 40118 + }, + { + "epoch": 2.112421052631579, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 40119 + }, + { + "epoch": 2.1124812030075186, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 40120 + }, + { + "epoch": 2.1125413533834587, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 40121 + }, + { + "epoch": 2.1126015037593984, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 40122 + }, + { + "epoch": 2.1126616541353385, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 40123 + }, + { + "epoch": 2.112721804511278, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 40124 + }, + { + "epoch": 2.112781954887218, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 40125 + }, + { + "epoch": 2.112842105263158, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 40126 + }, + { + "epoch": 2.112902255639098, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 40127 + }, + { + "epoch": 2.1129624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 40128 + }, + { + "epoch": 2.1130225563909772, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 40129 + }, + { + "epoch": 2.1130827067669173, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 40130 + }, + { + "epoch": 2.113142857142857, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 40131 + }, + { + "epoch": 2.113203007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 40132 + }, + { + "epoch": 2.1132631578947367, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 40133 + }, + { + "epoch": 2.113323308270677, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1279, + "step": 40134 + }, + { + "epoch": 2.1133834586466165, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 40135 + }, + { + "epoch": 2.1134436090225566, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 40136 + }, + { + "epoch": 2.113503759398496, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 40137 + }, + { + "epoch": 2.1135639097744363, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 40138 + }, + { + "epoch": 2.113624060150376, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.893, + "step": 40139 + }, + { + "epoch": 2.1136842105263156, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 40140 + }, + { + "epoch": 2.1137443609022557, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 40141 + }, + { + "epoch": 2.1138045112781954, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 40142 + }, + { + "epoch": 2.1138646616541354, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 40143 + }, + { + "epoch": 2.113924812030075, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 40144 + }, + { + "epoch": 2.113984962406015, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 40145 + }, + { + "epoch": 2.114045112781955, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 40146 + }, + { + "epoch": 2.114105263157895, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 40147 + }, + { + "epoch": 2.1141654135338346, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 40148 + }, + { + "epoch": 2.1142255639097742, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 40149 + }, + { + "epoch": 2.1142857142857143, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 40150 + }, + { + "epoch": 2.114345864661654, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 40151 + }, + { + "epoch": 2.114406015037594, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 40152 + }, + { + "epoch": 2.1144661654135337, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 40153 + }, + { + "epoch": 2.114526315789474, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.254, + "step": 40154 + }, + { + "epoch": 2.1145864661654135, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 40155 + }, + { + "epoch": 2.1146466165413536, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 40156 + }, + { + "epoch": 2.114706766917293, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 40157 + }, + { + "epoch": 2.1147669172932333, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 40158 + }, + { + "epoch": 2.114827067669173, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 40159 + }, + { + "epoch": 2.1148872180451126, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 40160 + }, + { + "epoch": 2.1149473684210527, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 40161 + }, + { + "epoch": 2.1150075187969923, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 40162 + }, + { + "epoch": 2.1150676691729324, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 40163 + }, + { + "epoch": 2.115127819548872, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 40164 + }, + { + "epoch": 2.115187969924812, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 40165 + }, + { + "epoch": 2.115248120300752, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 40166 + }, + { + "epoch": 2.115308270676692, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 40167 + }, + { + "epoch": 2.1153684210526316, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 40168 + }, + { + "epoch": 2.115428571428571, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1691, + "step": 40169 + }, + { + "epoch": 2.1154887218045113, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 40170 + }, + { + "epoch": 2.115548872180451, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 40171 + }, + { + "epoch": 2.115609022556391, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 40172 + }, + { + "epoch": 2.1156691729323307, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 40173 + }, + { + "epoch": 2.115729323308271, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 40174 + }, + { + "epoch": 2.1157894736842104, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 40175 + }, + { + "epoch": 2.1158496240601505, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 40176 + }, + { + "epoch": 2.11590977443609, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 40177 + }, + { + "epoch": 2.1159699248120303, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 40178 + }, + { + "epoch": 2.11603007518797, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 40179 + }, + { + "epoch": 2.1160902255639096, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 40180 + }, + { + "epoch": 2.1161503759398497, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 40181 + }, + { + "epoch": 2.1162105263157893, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 40182 + }, + { + "epoch": 2.1162706766917294, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 40183 + }, + { + "epoch": 2.116330827067669, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 40184 + }, + { + "epoch": 2.116390977443609, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 40185 + }, + { + "epoch": 2.116451127819549, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 40186 + }, + { + "epoch": 2.116511278195489, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 40187 + }, + { + "epoch": 2.1165714285714285, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.17, + "step": 40188 + }, + { + "epoch": 2.116631578947368, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 40189 + }, + { + "epoch": 2.1166917293233083, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 40190 + }, + { + "epoch": 2.116751879699248, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 40191 + }, + { + "epoch": 2.116812030075188, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 40192 + }, + { + "epoch": 2.1168721804511277, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 40193 + }, + { + "epoch": 2.1169323308270678, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 40194 + }, + { + "epoch": 2.1169924812030074, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 40195 + }, + { + "epoch": 2.1170526315789475, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 40196 + }, + { + "epoch": 2.117112781954887, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 40197 + }, + { + "epoch": 2.1171729323308273, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 40198 + }, + { + "epoch": 2.117233082706767, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 40199 + }, + { + "epoch": 2.1172932330827066, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 40200 + }, + { + "epoch": 2.1173533834586467, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 40201 + }, + { + "epoch": 2.1174135338345863, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 40202 + }, + { + "epoch": 2.1174736842105264, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 40203 + }, + { + "epoch": 2.117533834586466, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 40204 + }, + { + "epoch": 2.117593984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 40205 + }, + { + "epoch": 2.117654135338346, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 40206 + }, + { + "epoch": 2.117714285714286, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 40207 + }, + { + "epoch": 2.1177744360902255, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 40208 + }, + { + "epoch": 2.1178345864661656, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 40209 + }, + { + "epoch": 2.1178947368421053, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 40210 + }, + { + "epoch": 2.117954887218045, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 40211 + }, + { + "epoch": 2.118015037593985, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 40212 + }, + { + "epoch": 2.1180751879699247, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 40213 + }, + { + "epoch": 2.1181353383458648, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 40214 + }, + { + "epoch": 2.1181954887218044, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 40215 + }, + { + "epoch": 2.1182556390977445, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 40216 + }, + { + "epoch": 2.118315789473684, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 40217 + }, + { + "epoch": 2.1183759398496242, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 40218 + }, + { + "epoch": 2.118436090225564, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 40219 + }, + { + "epoch": 2.1184962406015035, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 40220 + }, + { + "epoch": 2.1185563909774436, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 40221 + }, + { + "epoch": 2.1186165413533833, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 40222 + }, + { + "epoch": 2.1186766917293234, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 40223 + }, + { + "epoch": 2.118736842105263, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 40224 + }, + { + "epoch": 2.118796992481203, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 40225 + }, + { + "epoch": 2.1188571428571428, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 40226 + }, + { + "epoch": 2.118917293233083, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 40227 + }, + { + "epoch": 2.1189774436090225, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 40228 + }, + { + "epoch": 2.1190375939849626, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 40229 + }, + { + "epoch": 2.1190977443609023, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 40230 + }, + { + "epoch": 2.119157894736842, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 40231 + }, + { + "epoch": 2.119218045112782, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 40232 + }, + { + "epoch": 2.1192781954887217, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 40233 + }, + { + "epoch": 2.1193383458646617, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 40234 + }, + { + "epoch": 2.1193984962406014, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 40235 + }, + { + "epoch": 2.1194586466165415, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 40236 + }, + { + "epoch": 2.119518796992481, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.8902, + "step": 40237 + }, + { + "epoch": 2.1195789473684212, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 40238 + }, + { + "epoch": 2.119639097744361, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 40239 + }, + { + "epoch": 2.119699248120301, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 40240 + }, + { + "epoch": 2.1197593984962406, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 40241 + }, + { + "epoch": 2.1198195488721803, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 40242 + }, + { + "epoch": 2.1198796992481204, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1414, + "step": 40243 + }, + { + "epoch": 2.11993984962406, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 40244 + }, + { + "epoch": 2.12, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 40245 + }, + { + "epoch": 2.1200601503759398, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 40246 + }, + { + "epoch": 2.12012030075188, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 40247 + }, + { + "epoch": 2.1201804511278195, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 40248 + }, + { + "epoch": 2.1202406015037596, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 40249 + }, + { + "epoch": 2.1203007518796992, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1691, + "step": 40250 + }, + { + "epoch": 2.120360902255639, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 40251 + }, + { + "epoch": 2.120421052631579, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 40252 + }, + { + "epoch": 2.1204812030075186, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.8562, + "step": 40253 + }, + { + "epoch": 2.1205413533834587, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.936, + "step": 40254 + }, + { + "epoch": 2.1206015037593984, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 40255 + }, + { + "epoch": 2.1206616541353385, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 40256 + }, + { + "epoch": 2.120721804511278, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 40257 + }, + { + "epoch": 2.120781954887218, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 40258 + }, + { + "epoch": 2.120842105263158, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 40259 + }, + { + "epoch": 2.120902255639098, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 40260 + }, + { + "epoch": 2.1209624060150376, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 40261 + }, + { + "epoch": 2.1210225563909773, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 40262 + }, + { + "epoch": 2.1210827067669173, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 40263 + }, + { + "epoch": 2.121142857142857, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 40264 + }, + { + "epoch": 2.121203007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.8928, + "step": 40265 + }, + { + "epoch": 2.1212631578947367, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1541, + "step": 40266 + }, + { + "epoch": 2.121323308270677, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 40267 + }, + { + "epoch": 2.1213834586466165, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 40268 + }, + { + "epoch": 2.1214436090225566, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 40269 + }, + { + "epoch": 2.1215037593984962, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 40270 + }, + { + "epoch": 2.1215639097744363, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 40271 + }, + { + "epoch": 2.121624060150376, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 40272 + }, + { + "epoch": 2.1216842105263156, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 40273 + }, + { + "epoch": 2.1217443609022557, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 40274 + }, + { + "epoch": 2.1218045112781954, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 40275 + }, + { + "epoch": 2.1218646616541355, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9207, + "step": 40276 + }, + { + "epoch": 2.121924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 40277 + }, + { + "epoch": 2.121984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 40278 + }, + { + "epoch": 2.122045112781955, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 40279 + }, + { + "epoch": 2.122105263157895, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9221, + "step": 40280 + }, + { + "epoch": 2.1221654135338346, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 40281 + }, + { + "epoch": 2.1222255639097742, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 40282 + }, + { + "epoch": 2.1222857142857143, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9383, + "step": 40283 + }, + { + "epoch": 2.122345864661654, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.179, + "step": 40284 + }, + { + "epoch": 2.122406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 40285 + }, + { + "epoch": 2.1224661654135337, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 40286 + }, + { + "epoch": 2.122526315789474, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 40287 + }, + { + "epoch": 2.1225864661654135, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 40288 + }, + { + "epoch": 2.1226466165413536, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9667, + "step": 40289 + }, + { + "epoch": 2.122706766917293, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 40290 + }, + { + "epoch": 2.122766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 40291 + }, + { + "epoch": 2.122827067669173, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 40292 + }, + { + "epoch": 2.1228872180451126, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 40293 + }, + { + "epoch": 2.1229473684210527, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 40294 + }, + { + "epoch": 2.1230075187969923, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 40295 + }, + { + "epoch": 2.1230676691729324, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 40296 + }, + { + "epoch": 2.123127819548872, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 40297 + }, + { + "epoch": 2.123187969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 40298 + }, + { + "epoch": 2.123248120300752, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 40299 + }, + { + "epoch": 2.123308270676692, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 40300 + }, + { + "epoch": 2.1233684210526316, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 40301 + }, + { + "epoch": 2.123428571428571, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 40302 + }, + { + "epoch": 2.1234887218045113, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 40303 + }, + { + "epoch": 2.123548872180451, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 40304 + }, + { + "epoch": 2.123609022556391, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 40305 + }, + { + "epoch": 2.1236691729323307, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 40306 + }, + { + "epoch": 2.123729323308271, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 40307 + }, + { + "epoch": 2.1237894736842104, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 40308 + }, + { + "epoch": 2.1238496240601505, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 40309 + }, + { + "epoch": 2.12390977443609, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 40310 + }, + { + "epoch": 2.1239699248120303, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 40311 + }, + { + "epoch": 2.12403007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 40312 + }, + { + "epoch": 2.1240902255639096, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 40313 + }, + { + "epoch": 2.1241503759398497, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 40314 + }, + { + "epoch": 2.1242105263157893, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 40315 + }, + { + "epoch": 2.1242706766917294, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 40316 + }, + { + "epoch": 2.124330827067669, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 40317 + }, + { + "epoch": 2.124390977443609, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 40318 + }, + { + "epoch": 2.124451127819549, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 40319 + }, + { + "epoch": 2.124511278195489, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 40320 + }, + { + "epoch": 2.1245714285714286, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 40321 + }, + { + "epoch": 2.124631578947368, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 40322 + }, + { + "epoch": 2.1246917293233083, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 40323 + }, + { + "epoch": 2.124751879699248, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 40324 + }, + { + "epoch": 2.124812030075188, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 40325 + }, + { + "epoch": 2.1248721804511277, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 40326 + }, + { + "epoch": 2.124932330827068, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 40327 + }, + { + "epoch": 2.1249924812030074, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 40328 + }, + { + "epoch": 2.1250526315789475, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 40329 + }, + { + "epoch": 2.125112781954887, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 40330 + }, + { + "epoch": 2.1251729323308273, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 40331 + }, + { + "epoch": 2.125233082706767, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 40332 + }, + { + "epoch": 2.1252932330827066, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 40333 + }, + { + "epoch": 2.1253533834586467, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 40334 + }, + { + "epoch": 2.1254135338345863, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 40335 + }, + { + "epoch": 2.1254736842105264, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 40336 + }, + { + "epoch": 2.125533834586466, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1905, + "step": 40337 + }, + { + "epoch": 2.125593984962406, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 40338 + }, + { + "epoch": 2.125654135338346, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 40339 + }, + { + "epoch": 2.125714285714286, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 40340 + }, + { + "epoch": 2.1257744360902255, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 40341 + }, + { + "epoch": 2.1258345864661656, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 40342 + }, + { + "epoch": 2.1258947368421053, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.187, + "step": 40343 + }, + { + "epoch": 2.125954887218045, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 40344 + }, + { + "epoch": 2.126015037593985, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 40345 + }, + { + "epoch": 2.1260751879699247, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 40346 + }, + { + "epoch": 2.1261353383458648, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 40347 + }, + { + "epoch": 2.1261954887218044, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 40348 + }, + { + "epoch": 2.1262556390977445, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 40349 + }, + { + "epoch": 2.126315789473684, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 40350 + }, + { + "epoch": 2.1263759398496243, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 40351 + }, + { + "epoch": 2.126436090225564, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 40352 + }, + { + "epoch": 2.1264962406015036, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 40353 + }, + { + "epoch": 2.1265563909774436, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9433, + "step": 40354 + }, + { + "epoch": 2.1266165413533833, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 40355 + }, + { + "epoch": 2.1266766917293234, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 40356 + }, + { + "epoch": 2.126736842105263, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 40357 + }, + { + "epoch": 2.126796992481203, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 40358 + }, + { + "epoch": 2.126857142857143, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 40359 + }, + { + "epoch": 2.126917293233083, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 40360 + }, + { + "epoch": 2.1269774436090225, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 40361 + }, + { + "epoch": 2.127037593984962, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 40362 + }, + { + "epoch": 2.1270977443609023, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 40363 + }, + { + "epoch": 2.127157894736842, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 40364 + }, + { + "epoch": 2.127218045112782, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 40365 + }, + { + "epoch": 2.1272781954887217, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 40366 + }, + { + "epoch": 2.1273383458646618, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.132, + "step": 40367 + }, + { + "epoch": 2.1273984962406014, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 40368 + }, + { + "epoch": 2.1274586466165415, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 40369 + }, + { + "epoch": 2.127518796992481, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 40370 + }, + { + "epoch": 2.1275789473684212, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 40371 + }, + { + "epoch": 2.127639097744361, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 40372 + }, + { + "epoch": 2.127699248120301, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 40373 + }, + { + "epoch": 2.1277593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 40374 + }, + { + "epoch": 2.1278195488721803, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 40375 + }, + { + "epoch": 2.1278796992481204, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 40376 + }, + { + "epoch": 2.12793984962406, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 40377 + }, + { + "epoch": 2.128, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 40378 + }, + { + "epoch": 2.1280601503759398, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 40379 + }, + { + "epoch": 2.12812030075188, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 40380 + }, + { + "epoch": 2.1281804511278195, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 40381 + }, + { + "epoch": 2.1282406015037596, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9389, + "step": 40382 + }, + { + "epoch": 2.1283007518796992, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 40383 + }, + { + "epoch": 2.128360902255639, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 40384 + }, + { + "epoch": 2.128421052631579, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 40385 + }, + { + "epoch": 2.1284812030075186, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 40386 + }, + { + "epoch": 2.1285413533834587, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 40387 + }, + { + "epoch": 2.1286015037593984, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 40388 + }, + { + "epoch": 2.1286616541353385, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 40389 + }, + { + "epoch": 2.128721804511278, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 40390 + }, + { + "epoch": 2.128781954887218, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 40391 + }, + { + "epoch": 2.128842105263158, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 40392 + }, + { + "epoch": 2.1289022556390975, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 40393 + }, + { + "epoch": 2.1289624060150376, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 40394 + }, + { + "epoch": 2.1290225563909773, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 40395 + }, + { + "epoch": 2.1290827067669174, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 40396 + }, + { + "epoch": 2.129142857142857, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 40397 + }, + { + "epoch": 2.129203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 40398 + }, + { + "epoch": 2.1292631578947367, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 40399 + }, + { + "epoch": 2.129323308270677, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 40400 + }, + { + "epoch": 2.1293834586466165, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 40401 + }, + { + "epoch": 2.1294436090225566, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 40402 + }, + { + "epoch": 2.1295037593984962, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 40403 + }, + { + "epoch": 2.1295639097744363, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 40404 + }, + { + "epoch": 2.129624060150376, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 40405 + }, + { + "epoch": 2.1296842105263156, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 40406 + }, + { + "epoch": 2.1297443609022557, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 40407 + }, + { + "epoch": 2.1298045112781954, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 40408 + }, + { + "epoch": 2.1298646616541355, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 40409 + }, + { + "epoch": 2.129924812030075, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 40410 + }, + { + "epoch": 2.129984962406015, + "grad_norm": 1.4921875, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 40411 + }, + { + "epoch": 2.130045112781955, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 40412 + }, + { + "epoch": 2.130105263157895, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 40413 + }, + { + "epoch": 2.1301654135338346, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 40414 + }, + { + "epoch": 2.1302255639097742, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 40415 + }, + { + "epoch": 2.1302857142857143, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 40416 + }, + { + "epoch": 2.130345864661654, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 40417 + }, + { + "epoch": 2.130406015037594, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 40418 + }, + { + "epoch": 2.1304661654135337, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 40419 + }, + { + "epoch": 2.130526315789474, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 40420 + }, + { + "epoch": 2.1305864661654135, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 40421 + }, + { + "epoch": 2.1306466165413536, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 40422 + }, + { + "epoch": 2.130706766917293, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 40423 + }, + { + "epoch": 2.130766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 40424 + }, + { + "epoch": 2.130827067669173, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 40425 + }, + { + "epoch": 2.1308872180451126, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 40426 + }, + { + "epoch": 2.1309473684210527, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 40427 + }, + { + "epoch": 2.1310075187969924, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 40428 + }, + { + "epoch": 2.1310676691729324, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 40429 + }, + { + "epoch": 2.131127819548872, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 40430 + }, + { + "epoch": 2.131187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 40431 + }, + { + "epoch": 2.131248120300752, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 40432 + }, + { + "epoch": 2.131308270676692, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 40433 + }, + { + "epoch": 2.1313684210526316, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 40434 + }, + { + "epoch": 2.1314285714285712, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 40435 + }, + { + "epoch": 2.1314887218045113, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 40436 + }, + { + "epoch": 2.131548872180451, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 40437 + }, + { + "epoch": 2.131609022556391, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 40438 + }, + { + "epoch": 2.1316691729323307, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 40439 + }, + { + "epoch": 2.131729323308271, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 40440 + }, + { + "epoch": 2.1317894736842105, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 40441 + }, + { + "epoch": 2.1318496240601505, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 40442 + }, + { + "epoch": 2.13190977443609, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 40443 + }, + { + "epoch": 2.1319699248120303, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 40444 + }, + { + "epoch": 2.13203007518797, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 40445 + }, + { + "epoch": 2.1320902255639096, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 40446 + }, + { + "epoch": 2.1321503759398497, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 40447 + }, + { + "epoch": 2.1322105263157893, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 40448 + }, + { + "epoch": 2.1322706766917294, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 40449 + }, + { + "epoch": 2.132330827067669, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 40450 + }, + { + "epoch": 2.132390977443609, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1756, + "step": 40451 + }, + { + "epoch": 2.132451127819549, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 40452 + }, + { + "epoch": 2.132511278195489, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 40453 + }, + { + "epoch": 2.1325714285714286, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 40454 + }, + { + "epoch": 2.132631578947368, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 40455 + }, + { + "epoch": 2.1326917293233083, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 40456 + }, + { + "epoch": 2.132751879699248, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 40457 + }, + { + "epoch": 2.132812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 40458 + }, + { + "epoch": 2.1328721804511277, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 40459 + }, + { + "epoch": 2.132932330827068, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 40460 + }, + { + "epoch": 2.1329924812030074, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 40461 + }, + { + "epoch": 2.1330526315789475, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 40462 + }, + { + "epoch": 2.133112781954887, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 40463 + }, + { + "epoch": 2.133172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 40464 + }, + { + "epoch": 2.133233082706767, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 40465 + }, + { + "epoch": 2.1332932330827066, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9205, + "step": 40466 + }, + { + "epoch": 2.1333533834586467, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 40467 + }, + { + "epoch": 2.1334135338345863, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 40468 + }, + { + "epoch": 2.1334736842105264, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 40469 + }, + { + "epoch": 2.133533834586466, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 40470 + }, + { + "epoch": 2.133593984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 40471 + }, + { + "epoch": 2.133654135338346, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 40472 + }, + { + "epoch": 2.133714285714286, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 40473 + }, + { + "epoch": 2.1337744360902255, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 40474 + }, + { + "epoch": 2.1338345864661656, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 40475 + }, + { + "epoch": 2.1338947368421053, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 40476 + }, + { + "epoch": 2.133954887218045, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 40477 + }, + { + "epoch": 2.134015037593985, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 40478 + }, + { + "epoch": 2.1340751879699247, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 40479 + }, + { + "epoch": 2.1341353383458648, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 40480 + }, + { + "epoch": 2.1341954887218044, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 40481 + }, + { + "epoch": 2.1342556390977445, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2285, + "step": 40482 + }, + { + "epoch": 2.134315789473684, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 40483 + }, + { + "epoch": 2.1343759398496243, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 40484 + }, + { + "epoch": 2.134436090225564, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 40485 + }, + { + "epoch": 2.1344962406015036, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 40486 + }, + { + "epoch": 2.1345563909774437, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 40487 + }, + { + "epoch": 2.1346165413533833, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 40488 + }, + { + "epoch": 2.1346766917293234, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 40489 + }, + { + "epoch": 2.134736842105263, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 40490 + }, + { + "epoch": 2.134796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 40491 + }, + { + "epoch": 2.134857142857143, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 40492 + }, + { + "epoch": 2.134917293233083, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 40493 + }, + { + "epoch": 2.1349774436090225, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 40494 + }, + { + "epoch": 2.135037593984962, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 40495 + }, + { + "epoch": 2.1350977443609023, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 40496 + }, + { + "epoch": 2.135157894736842, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 40497 + }, + { + "epoch": 2.135218045112782, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 40498 + }, + { + "epoch": 2.1352781954887217, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 40499 + }, + { + "epoch": 2.1353383458646618, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 40500 + }, + { + "epoch": 2.1353984962406014, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 40501 + }, + { + "epoch": 2.1354586466165415, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 40502 + }, + { + "epoch": 2.135518796992481, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 40503 + }, + { + "epoch": 2.1355789473684212, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 40504 + }, + { + "epoch": 2.135639097744361, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 40505 + }, + { + "epoch": 2.135699248120301, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 40506 + }, + { + "epoch": 2.1357593984962406, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 40507 + }, + { + "epoch": 2.1358195488721803, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 40508 + }, + { + "epoch": 2.1358796992481204, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.8651, + "step": 40509 + }, + { + "epoch": 2.13593984962406, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 40510 + }, + { + "epoch": 2.136, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 40511 + }, + { + "epoch": 2.1360601503759398, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 40512 + }, + { + "epoch": 2.13612030075188, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 40513 + }, + { + "epoch": 2.1361804511278195, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 40514 + }, + { + "epoch": 2.1362406015037596, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 40515 + }, + { + "epoch": 2.1363007518796993, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 40516 + }, + { + "epoch": 2.136360902255639, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 40517 + }, + { + "epoch": 2.136421052631579, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 40518 + }, + { + "epoch": 2.1364812030075186, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 40519 + }, + { + "epoch": 2.1365413533834587, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 40520 + }, + { + "epoch": 2.1366015037593984, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 40521 + }, + { + "epoch": 2.1366616541353385, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 40522 + }, + { + "epoch": 2.136721804511278, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 40523 + }, + { + "epoch": 2.1367819548872182, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 40524 + }, + { + "epoch": 2.136842105263158, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 40525 + }, + { + "epoch": 2.1369022556390975, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 40526 + }, + { + "epoch": 2.1369624060150376, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9382, + "step": 40527 + }, + { + "epoch": 2.1370225563909773, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 40528 + }, + { + "epoch": 2.1370827067669174, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 40529 + }, + { + "epoch": 2.137142857142857, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 40530 + }, + { + "epoch": 2.137203007518797, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 40531 + }, + { + "epoch": 2.1372631578947368, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 40532 + }, + { + "epoch": 2.137323308270677, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9052, + "step": 40533 + }, + { + "epoch": 2.1373834586466165, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 40534 + }, + { + "epoch": 2.1374436090225566, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 40535 + }, + { + "epoch": 2.1375037593984962, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 40536 + }, + { + "epoch": 2.137563909774436, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 40537 + }, + { + "epoch": 2.137624060150376, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9531, + "step": 40538 + }, + { + "epoch": 2.1376842105263156, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 40539 + }, + { + "epoch": 2.1377443609022557, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 40540 + }, + { + "epoch": 2.1378045112781954, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 40541 + }, + { + "epoch": 2.1378646616541355, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 40542 + }, + { + "epoch": 2.137924812030075, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 40543 + }, + { + "epoch": 2.137984962406015, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 40544 + }, + { + "epoch": 2.138045112781955, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 40545 + }, + { + "epoch": 2.138105263157895, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 40546 + }, + { + "epoch": 2.1381654135338346, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 40547 + }, + { + "epoch": 2.1382255639097743, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 40548 + }, + { + "epoch": 2.1382857142857143, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 40549 + }, + { + "epoch": 2.138345864661654, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 40550 + }, + { + "epoch": 2.138406015037594, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 40551 + }, + { + "epoch": 2.1384661654135337, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 40552 + }, + { + "epoch": 2.138526315789474, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 40553 + }, + { + "epoch": 2.1385864661654135, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 40554 + }, + { + "epoch": 2.1386466165413536, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 40555 + }, + { + "epoch": 2.1387067669172932, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 40556 + }, + { + "epoch": 2.138766917293233, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 40557 + }, + { + "epoch": 2.138827067669173, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 40558 + }, + { + "epoch": 2.1388872180451126, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 40559 + }, + { + "epoch": 2.1389473684210527, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 40560 + }, + { + "epoch": 2.1390075187969924, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 40561 + }, + { + "epoch": 2.1390676691729325, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 40562 + }, + { + "epoch": 2.139127819548872, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 40563 + }, + { + "epoch": 2.139187969924812, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.2202, + "step": 40564 + }, + { + "epoch": 2.139248120300752, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 40565 + }, + { + "epoch": 2.139308270676692, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 40566 + }, + { + "epoch": 2.1393684210526316, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 40567 + }, + { + "epoch": 2.1394285714285712, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 40568 + }, + { + "epoch": 2.1394887218045113, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 40569 + }, + { + "epoch": 2.139548872180451, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 40570 + }, + { + "epoch": 2.139609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9227, + "step": 40571 + }, + { + "epoch": 2.1396691729323307, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 40572 + }, + { + "epoch": 2.139729323308271, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 40573 + }, + { + "epoch": 2.1397894736842105, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 40574 + }, + { + "epoch": 2.1398496240601506, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 40575 + }, + { + "epoch": 2.13990977443609, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 40576 + }, + { + "epoch": 2.1399699248120303, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.1572, + "step": 40577 + }, + { + "epoch": 2.14003007518797, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 40578 + }, + { + "epoch": 2.1400902255639096, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 40579 + }, + { + "epoch": 2.1401503759398497, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 40580 + }, + { + "epoch": 2.1402105263157893, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 40581 + }, + { + "epoch": 2.1402706766917294, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 40582 + }, + { + "epoch": 2.140330827067669, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 40583 + }, + { + "epoch": 2.140390977443609, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 40584 + }, + { + "epoch": 2.140451127819549, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 40585 + }, + { + "epoch": 2.140511278195489, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 40586 + }, + { + "epoch": 2.1405714285714286, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9396, + "step": 40587 + }, + { + "epoch": 2.140631578947368, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 40588 + }, + { + "epoch": 2.1406917293233083, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 40589 + }, + { + "epoch": 2.140751879699248, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 40590 + }, + { + "epoch": 2.140812030075188, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 40591 + }, + { + "epoch": 2.1408721804511277, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 40592 + }, + { + "epoch": 2.140932330827068, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 40593 + }, + { + "epoch": 2.1409924812030074, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 40594 + }, + { + "epoch": 2.1410526315789475, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 40595 + }, + { + "epoch": 2.141112781954887, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 40596 + }, + { + "epoch": 2.141172932330827, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 40597 + }, + { + "epoch": 2.141233082706767, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 40598 + }, + { + "epoch": 2.1412932330827066, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 40599 + }, + { + "epoch": 2.1413533834586467, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 40600 + }, + { + "epoch": 2.1414135338345863, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 40601 + }, + { + "epoch": 2.1414736842105264, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 40602 + }, + { + "epoch": 2.141533834586466, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.8499, + "step": 40603 + }, + { + "epoch": 2.141593984962406, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.946, + "step": 40604 + }, + { + "epoch": 2.141654135338346, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 40605 + }, + { + "epoch": 2.141714285714286, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 40606 + }, + { + "epoch": 2.1417744360902256, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 40607 + }, + { + "epoch": 2.1418345864661656, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9222, + "step": 40608 + }, + { + "epoch": 2.1418947368421053, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 40609 + }, + { + "epoch": 2.141954887218045, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 40610 + }, + { + "epoch": 2.142015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 40611 + }, + { + "epoch": 2.1420751879699247, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 40612 + }, + { + "epoch": 2.142135338345865, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 40613 + }, + { + "epoch": 2.1421954887218044, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 40614 + }, + { + "epoch": 2.1422556390977445, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 40615 + }, + { + "epoch": 2.142315789473684, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 40616 + }, + { + "epoch": 2.1423759398496243, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 40617 + }, + { + "epoch": 2.142436090225564, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 40618 + }, + { + "epoch": 2.1424962406015036, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 40619 + }, + { + "epoch": 2.1425563909774437, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 40620 + }, + { + "epoch": 2.1426165413533833, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 40621 + }, + { + "epoch": 2.1426766917293234, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 40622 + }, + { + "epoch": 2.142736842105263, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 40623 + }, + { + "epoch": 2.142796992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 40624 + }, + { + "epoch": 2.142857142857143, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 40625 + }, + { + "epoch": 2.142917293233083, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9343, + "step": 40626 + }, + { + "epoch": 2.1429774436090225, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 40627 + }, + { + "epoch": 2.143037593984962, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 40628 + }, + { + "epoch": 2.1430977443609023, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 40629 + }, + { + "epoch": 2.143157894736842, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 40630 + }, + { + "epoch": 2.143218045112782, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 40631 + }, + { + "epoch": 2.1432781954887217, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 40632 + }, + { + "epoch": 2.1433383458646618, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 40633 + }, + { + "epoch": 2.1433984962406014, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 40634 + }, + { + "epoch": 2.1434586466165415, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 40635 + }, + { + "epoch": 2.143518796992481, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 40636 + }, + { + "epoch": 2.1435789473684213, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 40637 + }, + { + "epoch": 2.143639097744361, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 40638 + }, + { + "epoch": 2.1436992481203005, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 40639 + }, + { + "epoch": 2.1437593984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 40640 + }, + { + "epoch": 2.1438195488721803, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2394, + "step": 40641 + }, + { + "epoch": 2.1438796992481204, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 40642 + }, + { + "epoch": 2.14393984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 40643 + }, + { + "epoch": 2.144, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 40644 + }, + { + "epoch": 2.1440601503759398, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 40645 + }, + { + "epoch": 2.14412030075188, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 40646 + }, + { + "epoch": 2.1441804511278195, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 40647 + }, + { + "epoch": 2.1442406015037596, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 40648 + }, + { + "epoch": 2.1443007518796993, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 40649 + }, + { + "epoch": 2.144360902255639, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 40650 + }, + { + "epoch": 2.144421052631579, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9225, + "step": 40651 + }, + { + "epoch": 2.1444812030075187, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 40652 + }, + { + "epoch": 2.1445413533834587, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 40653 + }, + { + "epoch": 2.1446015037593984, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 40654 + }, + { + "epoch": 2.1446616541353385, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 40655 + }, + { + "epoch": 2.144721804511278, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 40656 + }, + { + "epoch": 2.1447819548872182, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 40657 + }, + { + "epoch": 2.144842105263158, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 40658 + }, + { + "epoch": 2.1449022556390975, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 40659 + }, + { + "epoch": 2.1449624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 40660 + }, + { + "epoch": 2.1450225563909773, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 40661 + }, + { + "epoch": 2.1450827067669174, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 40662 + }, + { + "epoch": 2.145142857142857, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 40663 + }, + { + "epoch": 2.145203007518797, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 40664 + }, + { + "epoch": 2.1452631578947368, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 40665 + }, + { + "epoch": 2.145323308270677, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 40666 + }, + { + "epoch": 2.1453834586466165, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 40667 + }, + { + "epoch": 2.1454436090225566, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 40668 + }, + { + "epoch": 2.1455037593984962, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 40669 + }, + { + "epoch": 2.145563909774436, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 40670 + }, + { + "epoch": 2.145624060150376, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 40671 + }, + { + "epoch": 2.1456842105263156, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 40672 + }, + { + "epoch": 2.1457443609022557, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 40673 + }, + { + "epoch": 2.1458045112781954, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 40674 + }, + { + "epoch": 2.1458646616541355, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 40675 + }, + { + "epoch": 2.145924812030075, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 40676 + }, + { + "epoch": 2.145984962406015, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 40677 + }, + { + "epoch": 2.146045112781955, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 40678 + }, + { + "epoch": 2.146105263157895, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 40679 + }, + { + "epoch": 2.1461654135338346, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 40680 + }, + { + "epoch": 2.1462255639097743, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 40681 + }, + { + "epoch": 2.1462857142857144, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 40682 + }, + { + "epoch": 2.146345864661654, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 40683 + }, + { + "epoch": 2.146406015037594, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 40684 + }, + { + "epoch": 2.1464661654135337, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 40685 + }, + { + "epoch": 2.146526315789474, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 40686 + }, + { + "epoch": 2.1465864661654135, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 40687 + }, + { + "epoch": 2.1466466165413536, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 40688 + }, + { + "epoch": 2.1467067669172932, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 40689 + }, + { + "epoch": 2.146766917293233, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 40690 + }, + { + "epoch": 2.146827067669173, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 40691 + }, + { + "epoch": 2.1468872180451126, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 40692 + }, + { + "epoch": 2.1469473684210527, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 40693 + }, + { + "epoch": 2.1470075187969924, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 40694 + }, + { + "epoch": 2.1470676691729325, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 40695 + }, + { + "epoch": 2.147127819548872, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 40696 + }, + { + "epoch": 2.147187969924812, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 40697 + }, + { + "epoch": 2.147248120300752, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 40698 + }, + { + "epoch": 2.1473082706766915, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 40699 + }, + { + "epoch": 2.1473684210526316, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 40700 + }, + { + "epoch": 2.1474285714285712, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 40701 + }, + { + "epoch": 2.1474887218045113, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 40702 + }, + { + "epoch": 2.147548872180451, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 40703 + }, + { + "epoch": 2.147609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 40704 + }, + { + "epoch": 2.1476691729323307, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 40705 + }, + { + "epoch": 2.147729323308271, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 40706 + }, + { + "epoch": 2.1477894736842105, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 40707 + }, + { + "epoch": 2.1478496240601506, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 40708 + }, + { + "epoch": 2.14790977443609, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 40709 + }, + { + "epoch": 2.1479699248120303, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9392, + "step": 40710 + }, + { + "epoch": 2.14803007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 40711 + }, + { + "epoch": 2.1480902255639096, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 40712 + }, + { + "epoch": 2.1481503759398497, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 40713 + }, + { + "epoch": 2.1482105263157893, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 40714 + }, + { + "epoch": 2.1482706766917294, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 40715 + }, + { + "epoch": 2.148330827067669, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 40716 + }, + { + "epoch": 2.148390977443609, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 40717 + }, + { + "epoch": 2.148451127819549, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 40718 + }, + { + "epoch": 2.148511278195489, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 40719 + }, + { + "epoch": 2.1485714285714286, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 40720 + }, + { + "epoch": 2.1486315789473682, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 40721 + }, + { + "epoch": 2.1486917293233083, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 40722 + }, + { + "epoch": 2.148751879699248, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 40723 + }, + { + "epoch": 2.148812030075188, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 40724 + }, + { + "epoch": 2.1488721804511277, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 40725 + }, + { + "epoch": 2.148932330827068, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 40726 + }, + { + "epoch": 2.1489924812030075, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1371, + "step": 40727 + }, + { + "epoch": 2.1490526315789475, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9333, + "step": 40728 + }, + { + "epoch": 2.149112781954887, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 40729 + }, + { + "epoch": 2.149172932330827, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 40730 + }, + { + "epoch": 2.149233082706767, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 40731 + }, + { + "epoch": 2.1492932330827066, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 40732 + }, + { + "epoch": 2.1493533834586467, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 40733 + }, + { + "epoch": 2.1494135338345863, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 40734 + }, + { + "epoch": 2.1494736842105264, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 40735 + }, + { + "epoch": 2.149533834586466, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 40736 + }, + { + "epoch": 2.149593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 40737 + }, + { + "epoch": 2.149654135338346, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 40738 + }, + { + "epoch": 2.149714285714286, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 40739 + }, + { + "epoch": 2.1497744360902256, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 40740 + }, + { + "epoch": 2.149834586466165, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 40741 + }, + { + "epoch": 2.1498947368421053, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 40742 + }, + { + "epoch": 2.149954887218045, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 40743 + }, + { + "epoch": 2.150015037593985, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 40744 + }, + { + "epoch": 2.1500751879699247, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 40745 + }, + { + "epoch": 2.150135338345865, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 40746 + }, + { + "epoch": 2.1501954887218044, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 40747 + }, + { + "epoch": 2.1502556390977445, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 40748 + }, + { + "epoch": 2.150315789473684, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 40749 + }, + { + "epoch": 2.1503759398496243, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 40750 + }, + { + "epoch": 2.150436090225564, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 40751 + }, + { + "epoch": 2.1504962406015036, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 40752 + }, + { + "epoch": 2.1505563909774437, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 40753 + }, + { + "epoch": 2.1506165413533833, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 40754 + }, + { + "epoch": 2.1506766917293234, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.929, + "step": 40755 + }, + { + "epoch": 2.150736842105263, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 40756 + }, + { + "epoch": 2.150796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 40757 + }, + { + "epoch": 2.150857142857143, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 40758 + }, + { + "epoch": 2.150917293233083, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 40759 + }, + { + "epoch": 2.1509774436090225, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 40760 + }, + { + "epoch": 2.151037593984962, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 40761 + }, + { + "epoch": 2.1510977443609023, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 40762 + }, + { + "epoch": 2.151157894736842, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 40763 + }, + { + "epoch": 2.151218045112782, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 40764 + }, + { + "epoch": 2.1512781954887217, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 40765 + }, + { + "epoch": 2.1513383458646618, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 40766 + }, + { + "epoch": 2.1513984962406014, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 40767 + }, + { + "epoch": 2.1514586466165415, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 40768 + }, + { + "epoch": 2.151518796992481, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 40769 + }, + { + "epoch": 2.1515789473684213, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 40770 + }, + { + "epoch": 2.151639097744361, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 40771 + }, + { + "epoch": 2.1516992481203006, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 40772 + }, + { + "epoch": 2.1517593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 40773 + }, + { + "epoch": 2.1518195488721803, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 40774 + }, + { + "epoch": 2.1518796992481204, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 40775 + }, + { + "epoch": 2.15193984962406, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 40776 + }, + { + "epoch": 2.152, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 40777 + }, + { + "epoch": 2.15206015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 40778 + }, + { + "epoch": 2.15212030075188, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 40779 + }, + { + "epoch": 2.1521804511278195, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 40780 + }, + { + "epoch": 2.1522406015037596, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 40781 + }, + { + "epoch": 2.1523007518796993, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 40782 + }, + { + "epoch": 2.152360902255639, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 40783 + }, + { + "epoch": 2.152421052631579, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 40784 + }, + { + "epoch": 2.1524812030075187, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 40785 + }, + { + "epoch": 2.1525413533834588, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 40786 + }, + { + "epoch": 2.1526015037593984, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 40787 + }, + { + "epoch": 2.1526616541353385, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 40788 + }, + { + "epoch": 2.152721804511278, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 40789 + }, + { + "epoch": 2.1527819548872182, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 40790 + }, + { + "epoch": 2.152842105263158, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 40791 + }, + { + "epoch": 2.1529022556390975, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 40792 + }, + { + "epoch": 2.1529624060150376, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 40793 + }, + { + "epoch": 2.1530225563909773, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1069, + "step": 40794 + }, + { + "epoch": 2.1530827067669174, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 40795 + }, + { + "epoch": 2.153142857142857, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 40796 + }, + { + "epoch": 2.153203007518797, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 40797 + }, + { + "epoch": 2.1532631578947368, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 40798 + }, + { + "epoch": 2.153323308270677, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 40799 + }, + { + "epoch": 2.1533834586466165, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 40800 + }, + { + "epoch": 2.153443609022556, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 40801 + }, + { + "epoch": 2.1535037593984963, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 40802 + }, + { + "epoch": 2.153563909774436, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 40803 + }, + { + "epoch": 2.153624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 40804 + }, + { + "epoch": 2.1536842105263156, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 40805 + }, + { + "epoch": 2.1537443609022557, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 40806 + }, + { + "epoch": 2.1538045112781954, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 40807 + }, + { + "epoch": 2.1538646616541355, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1871, + "step": 40808 + }, + { + "epoch": 2.153924812030075, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 40809 + }, + { + "epoch": 2.1539849624060152, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 40810 + }, + { + "epoch": 2.154045112781955, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 40811 + }, + { + "epoch": 2.154105263157895, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 40812 + }, + { + "epoch": 2.1541654135338346, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9434, + "step": 40813 + }, + { + "epoch": 2.1542255639097743, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 40814 + }, + { + "epoch": 2.1542857142857144, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 40815 + }, + { + "epoch": 2.154345864661654, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 40816 + }, + { + "epoch": 2.154406015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 40817 + }, + { + "epoch": 2.1544661654135338, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 40818 + }, + { + "epoch": 2.154526315789474, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 40819 + }, + { + "epoch": 2.1545864661654135, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 40820 + }, + { + "epoch": 2.1546466165413536, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 40821 + }, + { + "epoch": 2.1547067669172932, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2483, + "step": 40822 + }, + { + "epoch": 2.154766917293233, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 40823 + }, + { + "epoch": 2.154827067669173, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 40824 + }, + { + "epoch": 2.1548872180451126, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 40825 + }, + { + "epoch": 2.1549473684210527, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 40826 + }, + { + "epoch": 2.1550075187969924, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 40827 + }, + { + "epoch": 2.1550676691729325, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 40828 + }, + { + "epoch": 2.155127819548872, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 40829 + }, + { + "epoch": 2.155187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 40830 + }, + { + "epoch": 2.155248120300752, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 40831 + }, + { + "epoch": 2.1553082706766915, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 40832 + }, + { + "epoch": 2.1553684210526316, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 40833 + }, + { + "epoch": 2.1554285714285712, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 40834 + }, + { + "epoch": 2.1554887218045113, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 40835 + }, + { + "epoch": 2.155548872180451, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 40836 + }, + { + "epoch": 2.155609022556391, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 40837 + }, + { + "epoch": 2.1556691729323307, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 40838 + }, + { + "epoch": 2.155729323308271, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 40839 + }, + { + "epoch": 2.1557894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 40840 + }, + { + "epoch": 2.1558496240601506, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 40841 + }, + { + "epoch": 2.15590977443609, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 40842 + }, + { + "epoch": 2.15596992481203, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9459, + "step": 40843 + }, + { + "epoch": 2.15603007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 40844 + }, + { + "epoch": 2.1560902255639096, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 40845 + }, + { + "epoch": 2.1561503759398497, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 40846 + }, + { + "epoch": 2.1562105263157894, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 40847 + }, + { + "epoch": 2.1562706766917294, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 40848 + }, + { + "epoch": 2.156330827067669, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 40849 + }, + { + "epoch": 2.156390977443609, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 40850 + }, + { + "epoch": 2.156451127819549, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 40851 + }, + { + "epoch": 2.156511278195489, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9432, + "step": 40852 + }, + { + "epoch": 2.1565714285714286, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 40853 + }, + { + "epoch": 2.1566315789473682, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 40854 + }, + { + "epoch": 2.1566917293233083, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 40855 + }, + { + "epoch": 2.156751879699248, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 40856 + }, + { + "epoch": 2.156812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 40857 + }, + { + "epoch": 2.1568721804511277, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 40858 + }, + { + "epoch": 2.156932330827068, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 40859 + }, + { + "epoch": 2.1569924812030075, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 40860 + }, + { + "epoch": 2.1570526315789476, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 40861 + }, + { + "epoch": 2.157112781954887, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 40862 + }, + { + "epoch": 2.157172932330827, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 40863 + }, + { + "epoch": 2.157233082706767, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 40864 + }, + { + "epoch": 2.1572932330827066, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 40865 + }, + { + "epoch": 2.1573533834586467, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 40866 + }, + { + "epoch": 2.1574135338345863, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 40867 + }, + { + "epoch": 2.1574736842105264, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 40868 + }, + { + "epoch": 2.157533834586466, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9293, + "step": 40869 + }, + { + "epoch": 2.157593984962406, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 40870 + }, + { + "epoch": 2.157654135338346, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 40871 + }, + { + "epoch": 2.157714285714286, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 40872 + }, + { + "epoch": 2.1577744360902256, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 40873 + }, + { + "epoch": 2.157834586466165, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 40874 + }, + { + "epoch": 2.1578947368421053, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 40875 + }, + { + "epoch": 2.157954887218045, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 40876 + }, + { + "epoch": 2.158015037593985, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 40877 + }, + { + "epoch": 2.1580751879699247, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 40878 + }, + { + "epoch": 2.158135338345865, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 40879 + }, + { + "epoch": 2.1581954887218044, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.2357, + "step": 40880 + }, + { + "epoch": 2.1582556390977445, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 40881 + }, + { + "epoch": 2.158315789473684, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 40882 + }, + { + "epoch": 2.1583759398496243, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 40883 + }, + { + "epoch": 2.158436090225564, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 40884 + }, + { + "epoch": 2.1584962406015036, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 40885 + }, + { + "epoch": 2.1585563909774437, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 40886 + }, + { + "epoch": 2.1586165413533833, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 40887 + }, + { + "epoch": 2.1586766917293234, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 40888 + }, + { + "epoch": 2.158736842105263, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 40889 + }, + { + "epoch": 2.158796992481203, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 40890 + }, + { + "epoch": 2.158857142857143, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 40891 + }, + { + "epoch": 2.158917293233083, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 40892 + }, + { + "epoch": 2.1589774436090226, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 40893 + }, + { + "epoch": 2.159037593984962, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 40894 + }, + { + "epoch": 2.1590977443609023, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 40895 + }, + { + "epoch": 2.159157894736842, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 40896 + }, + { + "epoch": 2.159218045112782, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 40897 + }, + { + "epoch": 2.1592781954887217, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 40898 + }, + { + "epoch": 2.159338345864662, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 40899 + }, + { + "epoch": 2.1593984962406014, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 40900 + }, + { + "epoch": 2.1594586466165415, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 40901 + }, + { + "epoch": 2.159518796992481, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 40902 + }, + { + "epoch": 2.159578947368421, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 40903 + }, + { + "epoch": 2.159639097744361, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 40904 + }, + { + "epoch": 2.1596992481203006, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 40905 + }, + { + "epoch": 2.1597593984962407, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 40906 + }, + { + "epoch": 2.1598195488721803, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 40907 + }, + { + "epoch": 2.1598796992481204, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 40908 + }, + { + "epoch": 2.15993984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 40909 + }, + { + "epoch": 2.16, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 40910 + }, + { + "epoch": 2.16006015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 40911 + }, + { + "epoch": 2.16012030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9488, + "step": 40912 + }, + { + "epoch": 2.1601804511278195, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 40913 + }, + { + "epoch": 2.1602406015037596, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 40914 + }, + { + "epoch": 2.1603007518796993, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 40915 + }, + { + "epoch": 2.160360902255639, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 40916 + }, + { + "epoch": 2.160421052631579, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 40917 + }, + { + "epoch": 2.1604812030075187, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 40918 + }, + { + "epoch": 2.1605413533834588, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 40919 + }, + { + "epoch": 2.1606015037593984, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 40920 + }, + { + "epoch": 2.1606616541353385, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 40921 + }, + { + "epoch": 2.160721804511278, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 40922 + }, + { + "epoch": 2.1607819548872182, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 40923 + }, + { + "epoch": 2.160842105263158, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 40924 + }, + { + "epoch": 2.1609022556390975, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 40925 + }, + { + "epoch": 2.1609624060150376, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 40926 + }, + { + "epoch": 2.1610225563909773, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 40927 + }, + { + "epoch": 2.1610827067669174, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 40928 + }, + { + "epoch": 2.161142857142857, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 40929 + }, + { + "epoch": 2.161203007518797, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 40930 + }, + { + "epoch": 2.1612631578947368, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 40931 + }, + { + "epoch": 2.161323308270677, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 40932 + }, + { + "epoch": 2.1613834586466165, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 40933 + }, + { + "epoch": 2.161443609022556, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9164, + "step": 40934 + }, + { + "epoch": 2.1615037593984963, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 40935 + }, + { + "epoch": 2.161563909774436, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 40936 + }, + { + "epoch": 2.161624060150376, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 40937 + }, + { + "epoch": 2.1616842105263157, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 40938 + }, + { + "epoch": 2.1617443609022557, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 40939 + }, + { + "epoch": 2.1618045112781954, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 40940 + }, + { + "epoch": 2.1618646616541355, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 40941 + }, + { + "epoch": 2.161924812030075, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 40942 + }, + { + "epoch": 2.1619849624060152, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 40943 + }, + { + "epoch": 2.162045112781955, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 40944 + }, + { + "epoch": 2.1621052631578945, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 40945 + }, + { + "epoch": 2.1621654135338346, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 40946 + }, + { + "epoch": 2.1622255639097743, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 40947 + }, + { + "epoch": 2.1622857142857144, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 40948 + }, + { + "epoch": 2.162345864661654, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 40949 + }, + { + "epoch": 2.162406015037594, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 40950 + }, + { + "epoch": 2.1624661654135338, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 40951 + }, + { + "epoch": 2.162526315789474, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 40952 + }, + { + "epoch": 2.1625864661654135, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 40953 + }, + { + "epoch": 2.1626466165413536, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 40954 + }, + { + "epoch": 2.1627067669172932, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 40955 + }, + { + "epoch": 2.162766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 40956 + }, + { + "epoch": 2.162827067669173, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 40957 + }, + { + "epoch": 2.1628872180451126, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 40958 + }, + { + "epoch": 2.1629473684210527, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 40959 + }, + { + "epoch": 2.1630075187969924, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 40960 + }, + { + "epoch": 2.1630676691729325, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 40961 + }, + { + "epoch": 2.163127819548872, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 40962 + }, + { + "epoch": 2.163187969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 40963 + }, + { + "epoch": 2.163248120300752, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 40964 + }, + { + "epoch": 2.1633082706766915, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 40965 + }, + { + "epoch": 2.1633684210526316, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 40966 + }, + { + "epoch": 2.1634285714285713, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 40967 + }, + { + "epoch": 2.1634887218045113, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 40968 + }, + { + "epoch": 2.163548872180451, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 40969 + }, + { + "epoch": 2.163609022556391, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 40970 + }, + { + "epoch": 2.1636691729323307, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 40971 + }, + { + "epoch": 2.163729323308271, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 40972 + }, + { + "epoch": 2.1637894736842105, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 40973 + }, + { + "epoch": 2.1638496240601506, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 40974 + }, + { + "epoch": 2.1639097744360902, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1961, + "step": 40975 + }, + { + "epoch": 2.16396992481203, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 40976 + }, + { + "epoch": 2.16403007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 40977 + }, + { + "epoch": 2.1640902255639096, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 40978 + }, + { + "epoch": 2.1641503759398497, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 40979 + }, + { + "epoch": 2.1642105263157894, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 40980 + }, + { + "epoch": 2.1642706766917295, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 40981 + }, + { + "epoch": 2.164330827067669, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 40982 + }, + { + "epoch": 2.164390977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 40983 + }, + { + "epoch": 2.164451127819549, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 40984 + }, + { + "epoch": 2.164511278195489, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.195, + "step": 40985 + }, + { + "epoch": 2.1645714285714286, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 40986 + }, + { + "epoch": 2.1646315789473682, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 40987 + }, + { + "epoch": 2.1646917293233083, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 40988 + }, + { + "epoch": 2.164751879699248, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 40989 + }, + { + "epoch": 2.164812030075188, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 40990 + }, + { + "epoch": 2.1648721804511277, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 40991 + }, + { + "epoch": 2.164932330827068, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 40992 + }, + { + "epoch": 2.1649924812030075, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 40993 + }, + { + "epoch": 2.1650526315789476, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 40994 + }, + { + "epoch": 2.165112781954887, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 40995 + }, + { + "epoch": 2.165172932330827, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 40996 + }, + { + "epoch": 2.165233082706767, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 40997 + }, + { + "epoch": 2.1652932330827066, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 40998 + }, + { + "epoch": 2.1653533834586467, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 40999 + }, + { + "epoch": 2.1654135338345863, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 41000 + }, + { + "epoch": 2.1654736842105264, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 41001 + }, + { + "epoch": 2.165533834586466, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 41002 + }, + { + "epoch": 2.165593984962406, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 41003 + }, + { + "epoch": 2.165654135338346, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 41004 + }, + { + "epoch": 2.1657142857142855, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 41005 + }, + { + "epoch": 2.1657744360902256, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 41006 + }, + { + "epoch": 2.1658345864661652, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 41007 + }, + { + "epoch": 2.1658947368421053, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 41008 + }, + { + "epoch": 2.165954887218045, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 41009 + }, + { + "epoch": 2.166015037593985, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 41010 + }, + { + "epoch": 2.1660751879699247, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 41011 + }, + { + "epoch": 2.166135338345865, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 41012 + }, + { + "epoch": 2.1661954887218045, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 41013 + }, + { + "epoch": 2.1662556390977445, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 41014 + }, + { + "epoch": 2.166315789473684, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 41015 + }, + { + "epoch": 2.1663759398496243, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 41016 + }, + { + "epoch": 2.166436090225564, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 41017 + }, + { + "epoch": 2.1664962406015036, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 41018 + }, + { + "epoch": 2.1665563909774437, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 41019 + }, + { + "epoch": 2.1666165413533833, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 41020 + }, + { + "epoch": 2.1666766917293234, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 41021 + }, + { + "epoch": 2.166736842105263, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 41022 + }, + { + "epoch": 2.166796992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 41023 + }, + { + "epoch": 2.166857142857143, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 41024 + }, + { + "epoch": 2.166917293233083, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 41025 + }, + { + "epoch": 2.1669774436090226, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 41026 + }, + { + "epoch": 2.167037593984962, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 41027 + }, + { + "epoch": 2.1670977443609023, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9347, + "step": 41028 + }, + { + "epoch": 2.167157894736842, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 41029 + }, + { + "epoch": 2.167218045112782, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 41030 + }, + { + "epoch": 2.1672781954887217, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 41031 + }, + { + "epoch": 2.167338345864662, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 41032 + }, + { + "epoch": 2.1673984962406014, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 41033 + }, + { + "epoch": 2.1674586466165415, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 41034 + }, + { + "epoch": 2.167518796992481, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 41035 + }, + { + "epoch": 2.167578947368421, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 41036 + }, + { + "epoch": 2.167639097744361, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 41037 + }, + { + "epoch": 2.1676992481203006, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 41038 + }, + { + "epoch": 2.1677593984962407, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 41039 + }, + { + "epoch": 2.1678195488721803, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 41040 + }, + { + "epoch": 2.1678796992481204, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 41041 + }, + { + "epoch": 2.16793984962406, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 41042 + }, + { + "epoch": 2.168, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 41043 + }, + { + "epoch": 2.16806015037594, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 41044 + }, + { + "epoch": 2.16812030075188, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 41045 + }, + { + "epoch": 2.1681804511278195, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 41046 + }, + { + "epoch": 2.168240601503759, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 41047 + }, + { + "epoch": 2.1683007518796993, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 41048 + }, + { + "epoch": 2.168360902255639, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 41049 + }, + { + "epoch": 2.168421052631579, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 41050 + }, + { + "epoch": 2.1684812030075187, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 41051 + }, + { + "epoch": 2.1685413533834588, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 41052 + }, + { + "epoch": 2.1686015037593984, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 41053 + }, + { + "epoch": 2.1686616541353385, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 41054 + }, + { + "epoch": 2.168721804511278, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 41055 + }, + { + "epoch": 2.1687819548872183, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 41056 + }, + { + "epoch": 2.168842105263158, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 41057 + }, + { + "epoch": 2.1689022556390976, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 41058 + }, + { + "epoch": 2.1689624060150376, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1633, + "step": 41059 + }, + { + "epoch": 2.1690225563909773, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 41060 + }, + { + "epoch": 2.1690827067669174, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 41061 + }, + { + "epoch": 2.169142857142857, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 41062 + }, + { + "epoch": 2.169203007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 41063 + }, + { + "epoch": 2.169263157894737, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 41064 + }, + { + "epoch": 2.169323308270677, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 41065 + }, + { + "epoch": 2.1693834586466165, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 41066 + }, + { + "epoch": 2.169443609022556, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 41067 + }, + { + "epoch": 2.1695037593984963, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 41068 + }, + { + "epoch": 2.169563909774436, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 41069 + }, + { + "epoch": 2.169624060150376, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1799, + "step": 41070 + }, + { + "epoch": 2.1696842105263157, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 41071 + }, + { + "epoch": 2.1697443609022558, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 41072 + }, + { + "epoch": 2.1698045112781954, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 41073 + }, + { + "epoch": 2.1698646616541355, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 41074 + }, + { + "epoch": 2.169924812030075, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 41075 + }, + { + "epoch": 2.1699849624060152, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 41076 + }, + { + "epoch": 2.170045112781955, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 41077 + }, + { + "epoch": 2.1701052631578945, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.94, + "step": 41078 + }, + { + "epoch": 2.1701654135338346, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 41079 + }, + { + "epoch": 2.1702255639097743, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 41080 + }, + { + "epoch": 2.1702857142857144, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 41081 + }, + { + "epoch": 2.170345864661654, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 41082 + }, + { + "epoch": 2.170406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 41083 + }, + { + "epoch": 2.1704661654135338, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 41084 + }, + { + "epoch": 2.170526315789474, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 41085 + }, + { + "epoch": 2.1705864661654135, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 41086 + }, + { + "epoch": 2.1706466165413536, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 41087 + }, + { + "epoch": 2.1707067669172933, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 41088 + }, + { + "epoch": 2.170766917293233, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 41089 + }, + { + "epoch": 2.170827067669173, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 41090 + }, + { + "epoch": 2.1708872180451126, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 41091 + }, + { + "epoch": 2.1709473684210527, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 41092 + }, + { + "epoch": 2.1710075187969924, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 41093 + }, + { + "epoch": 2.1710676691729325, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 41094 + }, + { + "epoch": 2.171127819548872, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 41095 + }, + { + "epoch": 2.1711879699248122, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 41096 + }, + { + "epoch": 2.171248120300752, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 41097 + }, + { + "epoch": 2.1713082706766915, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 41098 + }, + { + "epoch": 2.1713684210526316, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 41099 + }, + { + "epoch": 2.1714285714285713, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 41100 + }, + { + "epoch": 2.1714887218045114, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 41101 + }, + { + "epoch": 2.171548872180451, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 41102 + }, + { + "epoch": 2.171609022556391, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 41103 + }, + { + "epoch": 2.1716691729323307, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 41104 + }, + { + "epoch": 2.171729323308271, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9412, + "step": 41105 + }, + { + "epoch": 2.1717894736842105, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 41106 + }, + { + "epoch": 2.17184962406015, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 41107 + }, + { + "epoch": 2.1719097744360902, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 41108 + }, + { + "epoch": 2.17196992481203, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 41109 + }, + { + "epoch": 2.17203007518797, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 41110 + }, + { + "epoch": 2.1720902255639096, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 41111 + }, + { + "epoch": 2.1721503759398497, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 41112 + }, + { + "epoch": 2.1722105263157894, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 41113 + }, + { + "epoch": 2.1722706766917295, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 41114 + }, + { + "epoch": 2.172330827067669, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9434, + "step": 41115 + }, + { + "epoch": 2.172390977443609, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 41116 + }, + { + "epoch": 2.172451127819549, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 41117 + }, + { + "epoch": 2.172511278195489, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 41118 + }, + { + "epoch": 2.1725714285714286, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 41119 + }, + { + "epoch": 2.1726315789473682, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 41120 + }, + { + "epoch": 2.1726917293233083, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 41121 + }, + { + "epoch": 2.172751879699248, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 41122 + }, + { + "epoch": 2.172812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 41123 + }, + { + "epoch": 2.1728721804511277, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1969, + "step": 41124 + }, + { + "epoch": 2.172932330827068, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 41125 + }, + { + "epoch": 2.1729924812030075, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 41126 + }, + { + "epoch": 2.1730526315789476, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 41127 + }, + { + "epoch": 2.173112781954887, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 41128 + }, + { + "epoch": 2.173172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 41129 + }, + { + "epoch": 2.173233082706767, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 41130 + }, + { + "epoch": 2.1732932330827066, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 41131 + }, + { + "epoch": 2.1733533834586467, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 41132 + }, + { + "epoch": 2.1734135338345864, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9407, + "step": 41133 + }, + { + "epoch": 2.1734736842105264, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 41134 + }, + { + "epoch": 2.173533834586466, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 41135 + }, + { + "epoch": 2.173593984962406, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 41136 + }, + { + "epoch": 2.173654135338346, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 41137 + }, + { + "epoch": 2.1737142857142855, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 41138 + }, + { + "epoch": 2.1737744360902256, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 41139 + }, + { + "epoch": 2.1738345864661652, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 41140 + }, + { + "epoch": 2.1738947368421053, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 41141 + }, + { + "epoch": 2.173954887218045, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 41142 + }, + { + "epoch": 2.174015037593985, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 41143 + }, + { + "epoch": 2.1740751879699247, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 41144 + }, + { + "epoch": 2.174135338345865, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 41145 + }, + { + "epoch": 2.1741954887218045, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 41146 + }, + { + "epoch": 2.1742556390977446, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 41147 + }, + { + "epoch": 2.174315789473684, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 41148 + }, + { + "epoch": 2.1743759398496243, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 41149 + }, + { + "epoch": 2.174436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 41150 + }, + { + "epoch": 2.1744962406015036, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 41151 + }, + { + "epoch": 2.1745563909774437, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 41152 + }, + { + "epoch": 2.1746165413533833, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 41153 + }, + { + "epoch": 2.1746766917293234, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 41154 + }, + { + "epoch": 2.174736842105263, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 41155 + }, + { + "epoch": 2.174796992481203, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 41156 + }, + { + "epoch": 2.174857142857143, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 41157 + }, + { + "epoch": 2.174917293233083, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 41158 + }, + { + "epoch": 2.1749774436090226, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 41159 + }, + { + "epoch": 2.175037593984962, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 41160 + }, + { + "epoch": 2.1750977443609023, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 41161 + }, + { + "epoch": 2.175157894736842, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 41162 + }, + { + "epoch": 2.175218045112782, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 41163 + }, + { + "epoch": 2.1752781954887217, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 41164 + }, + { + "epoch": 2.175338345864662, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 41165 + }, + { + "epoch": 2.1753984962406014, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 41166 + }, + { + "epoch": 2.1754586466165415, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 41167 + }, + { + "epoch": 2.175518796992481, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 41168 + }, + { + "epoch": 2.175578947368421, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 41169 + }, + { + "epoch": 2.175639097744361, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 41170 + }, + { + "epoch": 2.1756992481203006, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 41171 + }, + { + "epoch": 2.1757593984962407, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 41172 + }, + { + "epoch": 2.1758195488721803, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 41173 + }, + { + "epoch": 2.1758796992481204, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 41174 + }, + { + "epoch": 2.17593984962406, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 41175 + }, + { + "epoch": 2.176, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 41176 + }, + { + "epoch": 2.17606015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 41177 + }, + { + "epoch": 2.17612030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 41178 + }, + { + "epoch": 2.1761804511278195, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 41179 + }, + { + "epoch": 2.176240601503759, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 41180 + }, + { + "epoch": 2.1763007518796993, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 41181 + }, + { + "epoch": 2.176360902255639, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 41182 + }, + { + "epoch": 2.176421052631579, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 41183 + }, + { + "epoch": 2.1764812030075187, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 41184 + }, + { + "epoch": 2.1765413533834588, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 41185 + }, + { + "epoch": 2.1766015037593984, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 41186 + }, + { + "epoch": 2.1766616541353385, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 41187 + }, + { + "epoch": 2.176721804511278, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 41188 + }, + { + "epoch": 2.1767819548872183, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 41189 + }, + { + "epoch": 2.176842105263158, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 41190 + }, + { + "epoch": 2.1769022556390976, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 41191 + }, + { + "epoch": 2.1769624060150377, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 41192 + }, + { + "epoch": 2.1770225563909773, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 41193 + }, + { + "epoch": 2.1770827067669174, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 41194 + }, + { + "epoch": 2.177142857142857, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 41195 + }, + { + "epoch": 2.177203007518797, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 41196 + }, + { + "epoch": 2.177263157894737, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 41197 + }, + { + "epoch": 2.177323308270677, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 41198 + }, + { + "epoch": 2.1773834586466165, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 41199 + }, + { + "epoch": 2.177443609022556, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 41200 + }, + { + "epoch": 2.1775037593984963, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 41201 + }, + { + "epoch": 2.177563909774436, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 41202 + }, + { + "epoch": 2.177624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 41203 + }, + { + "epoch": 2.1776842105263157, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 41204 + }, + { + "epoch": 2.1777443609022558, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 41205 + }, + { + "epoch": 2.1778045112781954, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 41206 + }, + { + "epoch": 2.1778646616541355, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 41207 + }, + { + "epoch": 2.177924812030075, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 41208 + }, + { + "epoch": 2.177984962406015, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 41209 + }, + { + "epoch": 2.178045112781955, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 41210 + }, + { + "epoch": 2.1781052631578945, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 41211 + }, + { + "epoch": 2.1781654135338346, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 41212 + }, + { + "epoch": 2.1782255639097743, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 41213 + }, + { + "epoch": 2.1782857142857144, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 41214 + }, + { + "epoch": 2.178345864661654, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 41215 + }, + { + "epoch": 2.178406015037594, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 41216 + }, + { + "epoch": 2.1784661654135338, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 41217 + }, + { + "epoch": 2.178526315789474, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 41218 + }, + { + "epoch": 2.1785864661654135, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 41219 + }, + { + "epoch": 2.1786466165413536, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9469, + "step": 41220 + }, + { + "epoch": 2.1787067669172933, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 41221 + }, + { + "epoch": 2.178766917293233, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 41222 + }, + { + "epoch": 2.178827067669173, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 41223 + }, + { + "epoch": 2.1788872180451127, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 41224 + }, + { + "epoch": 2.1789473684210527, + "grad_norm": 1.6796875, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 41225 + }, + { + "epoch": 2.1790075187969924, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 41226 + }, + { + "epoch": 2.1790676691729325, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 41227 + }, + { + "epoch": 2.179127819548872, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 41228 + }, + { + "epoch": 2.1791879699248122, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 41229 + }, + { + "epoch": 2.179248120300752, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 41230 + }, + { + "epoch": 2.1793082706766915, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 41231 + }, + { + "epoch": 2.1793684210526316, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 41232 + }, + { + "epoch": 2.1794285714285713, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 41233 + }, + { + "epoch": 2.1794887218045114, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 41234 + }, + { + "epoch": 2.179548872180451, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 41235 + }, + { + "epoch": 2.179609022556391, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 41236 + }, + { + "epoch": 2.1796691729323308, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 41237 + }, + { + "epoch": 2.179729323308271, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 41238 + }, + { + "epoch": 2.1797894736842105, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 41239 + }, + { + "epoch": 2.17984962406015, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 41240 + }, + { + "epoch": 2.1799097744360902, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 41241 + }, + { + "epoch": 2.17996992481203, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 41242 + }, + { + "epoch": 2.18003007518797, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 41243 + }, + { + "epoch": 2.1800902255639096, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 41244 + }, + { + "epoch": 2.1801503759398497, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 41245 + }, + { + "epoch": 2.1802105263157894, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 41246 + }, + { + "epoch": 2.1802706766917295, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1842, + "step": 41247 + }, + { + "epoch": 2.180330827067669, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 41248 + }, + { + "epoch": 2.180390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 41249 + }, + { + "epoch": 2.180451127819549, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 41250 + }, + { + "epoch": 2.180511278195489, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 41251 + }, + { + "epoch": 2.1805714285714286, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 41252 + }, + { + "epoch": 2.1806315789473683, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 41253 + }, + { + "epoch": 2.1806917293233083, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 41254 + }, + { + "epoch": 2.180751879699248, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.8931, + "step": 41255 + }, + { + "epoch": 2.180812030075188, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 41256 + }, + { + "epoch": 2.1808721804511277, + "grad_norm": 1.3984375, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 41257 + }, + { + "epoch": 2.180932330827068, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 41258 + }, + { + "epoch": 2.1809924812030075, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 41259 + }, + { + "epoch": 2.1810526315789476, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 41260 + }, + { + "epoch": 2.1811127819548872, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 41261 + }, + { + "epoch": 2.181172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 41262 + }, + { + "epoch": 2.181233082706767, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 41263 + }, + { + "epoch": 2.1812932330827066, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 41264 + }, + { + "epoch": 2.1813533834586467, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 41265 + }, + { + "epoch": 2.1814135338345864, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9211, + "step": 41266 + }, + { + "epoch": 2.1814736842105265, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9203, + "step": 41267 + }, + { + "epoch": 2.181533834586466, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 41268 + }, + { + "epoch": 2.181593984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1773, + "step": 41269 + }, + { + "epoch": 2.181654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.18, + "step": 41270 + }, + { + "epoch": 2.1817142857142855, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 41271 + }, + { + "epoch": 2.1817744360902256, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 41272 + }, + { + "epoch": 2.1818345864661652, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 41273 + }, + { + "epoch": 2.1818947368421053, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 41274 + }, + { + "epoch": 2.181954887218045, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 41275 + }, + { + "epoch": 2.182015037593985, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2536, + "step": 41276 + }, + { + "epoch": 2.1820751879699247, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 41277 + }, + { + "epoch": 2.182135338345865, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 41278 + }, + { + "epoch": 2.1821954887218045, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 41279 + }, + { + "epoch": 2.1822556390977446, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 41280 + }, + { + "epoch": 2.182315789473684, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 41281 + }, + { + "epoch": 2.182375939849624, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9098, + "step": 41282 + }, + { + "epoch": 2.182436090225564, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 41283 + }, + { + "epoch": 2.1824962406015036, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 41284 + }, + { + "epoch": 2.1825563909774437, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 41285 + }, + { + "epoch": 2.1826165413533833, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 41286 + }, + { + "epoch": 2.1826766917293234, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 41287 + }, + { + "epoch": 2.182736842105263, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 41288 + }, + { + "epoch": 2.182796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 41289 + }, + { + "epoch": 2.182857142857143, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 41290 + }, + { + "epoch": 2.182917293233083, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 41291 + }, + { + "epoch": 2.1829774436090226, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 41292 + }, + { + "epoch": 2.183037593984962, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 41293 + }, + { + "epoch": 2.1830977443609023, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 41294 + }, + { + "epoch": 2.183157894736842, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 41295 + }, + { + "epoch": 2.183218045112782, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 41296 + }, + { + "epoch": 2.1832781954887217, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 41297 + }, + { + "epoch": 2.183338345864662, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 41298 + }, + { + "epoch": 2.1833984962406014, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 41299 + }, + { + "epoch": 2.1834586466165415, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 41300 + }, + { + "epoch": 2.183518796992481, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 41301 + }, + { + "epoch": 2.183578947368421, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 41302 + }, + { + "epoch": 2.183639097744361, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 41303 + }, + { + "epoch": 2.1836992481203006, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 41304 + }, + { + "epoch": 2.1837593984962407, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 41305 + }, + { + "epoch": 2.1838195488721803, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 41306 + }, + { + "epoch": 2.1838796992481204, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 41307 + }, + { + "epoch": 2.18393984962406, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 41308 + }, + { + "epoch": 2.184, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 41309 + }, + { + "epoch": 2.18406015037594, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 41310 + }, + { + "epoch": 2.18412030075188, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 41311 + }, + { + "epoch": 2.1841804511278196, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 41312 + }, + { + "epoch": 2.184240601503759, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 41313 + }, + { + "epoch": 2.1843007518796993, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9387, + "step": 41314 + }, + { + "epoch": 2.184360902255639, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 41315 + }, + { + "epoch": 2.184421052631579, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 41316 + }, + { + "epoch": 2.1844812030075187, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 41317 + }, + { + "epoch": 2.184541353383459, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9413, + "step": 41318 + }, + { + "epoch": 2.1846015037593984, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 41319 + }, + { + "epoch": 2.1846616541353385, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 41320 + }, + { + "epoch": 2.184721804511278, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 41321 + }, + { + "epoch": 2.1847819548872183, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 41322 + }, + { + "epoch": 2.184842105263158, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 41323 + }, + { + "epoch": 2.1849022556390976, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 41324 + }, + { + "epoch": 2.1849624060150377, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 41325 + }, + { + "epoch": 2.1850225563909773, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 41326 + }, + { + "epoch": 2.1850827067669174, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 41327 + }, + { + "epoch": 2.185142857142857, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 41328 + }, + { + "epoch": 2.185203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 41329 + }, + { + "epoch": 2.185263157894737, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 41330 + }, + { + "epoch": 2.185323308270677, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 41331 + }, + { + "epoch": 2.1853834586466165, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 41332 + }, + { + "epoch": 2.185443609022556, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 41333 + }, + { + "epoch": 2.1855037593984963, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 41334 + }, + { + "epoch": 2.185563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 41335 + }, + { + "epoch": 2.185624060150376, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 41336 + }, + { + "epoch": 2.1856842105263157, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 41337 + }, + { + "epoch": 2.1857443609022558, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 41338 + }, + { + "epoch": 2.1858045112781954, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 41339 + }, + { + "epoch": 2.1858646616541355, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 41340 + }, + { + "epoch": 2.185924812030075, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 41341 + }, + { + "epoch": 2.185984962406015, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 41342 + }, + { + "epoch": 2.186045112781955, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 41343 + }, + { + "epoch": 2.1861052631578946, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 41344 + }, + { + "epoch": 2.1861654135338346, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.8684, + "step": 41345 + }, + { + "epoch": 2.1862255639097743, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 41346 + }, + { + "epoch": 2.1862857142857144, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 41347 + }, + { + "epoch": 2.186345864661654, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 41348 + }, + { + "epoch": 2.186406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 41349 + }, + { + "epoch": 2.186466165413534, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 41350 + }, + { + "epoch": 2.186526315789474, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 41351 + }, + { + "epoch": 2.1865864661654135, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 41352 + }, + { + "epoch": 2.1866466165413536, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 41353 + }, + { + "epoch": 2.1867067669172933, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 41354 + }, + { + "epoch": 2.186766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 41355 + }, + { + "epoch": 2.186827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 41356 + }, + { + "epoch": 2.1868872180451127, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 41357 + }, + { + "epoch": 2.1869473684210528, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 41358 + }, + { + "epoch": 2.1870075187969924, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 41359 + }, + { + "epoch": 2.1870676691729325, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 41360 + }, + { + "epoch": 2.187127819548872, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 41361 + }, + { + "epoch": 2.1871879699248122, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 41362 + }, + { + "epoch": 2.187248120300752, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 41363 + }, + { + "epoch": 2.1873082706766915, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 41364 + }, + { + "epoch": 2.1873684210526316, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 41365 + }, + { + "epoch": 2.1874285714285713, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 41366 + }, + { + "epoch": 2.1874887218045114, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 41367 + }, + { + "epoch": 2.187548872180451, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 41368 + }, + { + "epoch": 2.187609022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 41369 + }, + { + "epoch": 2.1876691729323308, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 41370 + }, + { + "epoch": 2.187729323308271, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 41371 + }, + { + "epoch": 2.1877894736842105, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 41372 + }, + { + "epoch": 2.18784962406015, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 41373 + }, + { + "epoch": 2.1879097744360902, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 41374 + }, + { + "epoch": 2.18796992481203, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 41375 + }, + { + "epoch": 2.18803007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 41376 + }, + { + "epoch": 2.1880902255639096, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 41377 + }, + { + "epoch": 2.1881503759398497, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 41378 + }, + { + "epoch": 2.1882105263157894, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 41379 + }, + { + "epoch": 2.1882706766917295, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 41380 + }, + { + "epoch": 2.188330827067669, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 41381 + }, + { + "epoch": 2.188390977443609, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 41382 + }, + { + "epoch": 2.188451127819549, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 41383 + }, + { + "epoch": 2.1885112781954885, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 41384 + }, + { + "epoch": 2.1885714285714286, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1804, + "step": 41385 + }, + { + "epoch": 2.1886315789473683, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 41386 + }, + { + "epoch": 2.1886917293233084, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 41387 + }, + { + "epoch": 2.188751879699248, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 41388 + }, + { + "epoch": 2.188812030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 41389 + }, + { + "epoch": 2.1888721804511277, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 41390 + }, + { + "epoch": 2.188932330827068, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 41391 + }, + { + "epoch": 2.1889924812030075, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 41392 + }, + { + "epoch": 2.1890526315789476, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 41393 + }, + { + "epoch": 2.1891127819548872, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 41394 + }, + { + "epoch": 2.189172932330827, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 41395 + }, + { + "epoch": 2.189233082706767, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 41396 + }, + { + "epoch": 2.1892932330827066, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 41397 + }, + { + "epoch": 2.1893533834586467, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 41398 + }, + { + "epoch": 2.1894135338345864, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 41399 + }, + { + "epoch": 2.1894736842105265, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 41400 + }, + { + "epoch": 2.189533834586466, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 41401 + }, + { + "epoch": 2.189593984962406, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 41402 + }, + { + "epoch": 2.189654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 41403 + }, + { + "epoch": 2.1897142857142855, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 41404 + }, + { + "epoch": 2.1897744360902256, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1414, + "step": 41405 + }, + { + "epoch": 2.1898345864661652, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 41406 + }, + { + "epoch": 2.1898947368421053, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9667, + "step": 41407 + }, + { + "epoch": 2.189954887218045, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 41408 + }, + { + "epoch": 2.190015037593985, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 41409 + }, + { + "epoch": 2.1900751879699247, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9187, + "step": 41410 + }, + { + "epoch": 2.190135338345865, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 41411 + }, + { + "epoch": 2.1901954887218045, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9456, + "step": 41412 + }, + { + "epoch": 2.1902556390977446, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 41413 + }, + { + "epoch": 2.190315789473684, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 41414 + }, + { + "epoch": 2.190375939849624, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 41415 + }, + { + "epoch": 2.190436090225564, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 41416 + }, + { + "epoch": 2.1904962406015036, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 41417 + }, + { + "epoch": 2.1905563909774437, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 41418 + }, + { + "epoch": 2.1906165413533834, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 41419 + }, + { + "epoch": 2.1906766917293234, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 41420 + }, + { + "epoch": 2.190736842105263, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.3119, + "step": 41421 + }, + { + "epoch": 2.190796992481203, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 41422 + }, + { + "epoch": 2.190857142857143, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 41423 + }, + { + "epoch": 2.190917293233083, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 41424 + }, + { + "epoch": 2.1909774436090226, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 41425 + }, + { + "epoch": 2.1910375939849622, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 41426 + }, + { + "epoch": 2.1910977443609023, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 41427 + }, + { + "epoch": 2.191157894736842, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 41428 + }, + { + "epoch": 2.191218045112782, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 41429 + }, + { + "epoch": 2.1912781954887217, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 41430 + }, + { + "epoch": 2.191338345864662, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 41431 + }, + { + "epoch": 2.1913984962406015, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 41432 + }, + { + "epoch": 2.1914586466165416, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 41433 + }, + { + "epoch": 2.191518796992481, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.3181, + "step": 41434 + }, + { + "epoch": 2.191578947368421, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 41435 + }, + { + "epoch": 2.191639097744361, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 41436 + }, + { + "epoch": 2.1916992481203006, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 41437 + }, + { + "epoch": 2.1917593984962407, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 41438 + }, + { + "epoch": 2.1918195488721803, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 41439 + }, + { + "epoch": 2.1918796992481204, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 41440 + }, + { + "epoch": 2.19193984962406, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 41441 + }, + { + "epoch": 2.192, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 41442 + }, + { + "epoch": 2.19206015037594, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 41443 + }, + { + "epoch": 2.1921203007518795, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 41444 + }, + { + "epoch": 2.1921804511278196, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 41445 + }, + { + "epoch": 2.192240601503759, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 41446 + }, + { + "epoch": 2.1923007518796993, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 41447 + }, + { + "epoch": 2.192360902255639, + "grad_norm": 1.875, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 41448 + }, + { + "epoch": 2.192421052631579, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 41449 + }, + { + "epoch": 2.1924812030075187, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 41450 + }, + { + "epoch": 2.192541353383459, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 41451 + }, + { + "epoch": 2.1926015037593984, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 41452 + }, + { + "epoch": 2.1926616541353385, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 41453 + }, + { + "epoch": 2.192721804511278, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9209, + "step": 41454 + }, + { + "epoch": 2.1927819548872183, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 41455 + }, + { + "epoch": 2.192842105263158, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 41456 + }, + { + "epoch": 2.1929022556390976, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 41457 + }, + { + "epoch": 2.1929624060150377, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 41458 + }, + { + "epoch": 2.1930225563909773, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 41459 + }, + { + "epoch": 2.1930827067669174, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 41460 + }, + { + "epoch": 2.193142857142857, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1978, + "step": 41461 + }, + { + "epoch": 2.193203007518797, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 41462 + }, + { + "epoch": 2.193263157894737, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 41463 + }, + { + "epoch": 2.193323308270677, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 41464 + }, + { + "epoch": 2.1933834586466165, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 41465 + }, + { + "epoch": 2.193443609022556, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 41466 + }, + { + "epoch": 2.1935037593984963, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 41467 + }, + { + "epoch": 2.193563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 41468 + }, + { + "epoch": 2.193624060150376, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1716, + "step": 41469 + }, + { + "epoch": 2.1936842105263157, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 41470 + }, + { + "epoch": 2.1937443609022558, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 41471 + }, + { + "epoch": 2.1938045112781954, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 41472 + }, + { + "epoch": 2.1938646616541355, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 41473 + }, + { + "epoch": 2.193924812030075, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9442, + "step": 41474 + }, + { + "epoch": 2.193984962406015, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 41475 + }, + { + "epoch": 2.194045112781955, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 41476 + }, + { + "epoch": 2.1941052631578946, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 41477 + }, + { + "epoch": 2.1941654135338347, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 41478 + }, + { + "epoch": 2.1942255639097743, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 41479 + }, + { + "epoch": 2.1942857142857144, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 41480 + }, + { + "epoch": 2.194345864661654, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 41481 + }, + { + "epoch": 2.194406015037594, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 41482 + }, + { + "epoch": 2.194466165413534, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 41483 + }, + { + "epoch": 2.194526315789474, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 41484 + }, + { + "epoch": 2.1945864661654135, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 41485 + }, + { + "epoch": 2.194646616541353, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 41486 + }, + { + "epoch": 2.1947067669172933, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 41487 + }, + { + "epoch": 2.194766917293233, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 41488 + }, + { + "epoch": 2.194827067669173, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 41489 + }, + { + "epoch": 2.1948872180451127, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 41490 + }, + { + "epoch": 2.1949473684210528, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 41491 + }, + { + "epoch": 2.1950075187969924, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 41492 + }, + { + "epoch": 2.1950676691729325, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 41493 + }, + { + "epoch": 2.195127819548872, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9297, + "step": 41494 + }, + { + "epoch": 2.1951879699248122, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 41495 + }, + { + "epoch": 2.195248120300752, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 41496 + }, + { + "epoch": 2.1953082706766915, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 41497 + }, + { + "epoch": 2.1953684210526316, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 41498 + }, + { + "epoch": 2.1954285714285713, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 41499 + }, + { + "epoch": 2.1954887218045114, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 41500 + }, + { + "epoch": 2.195548872180451, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 41501 + }, + { + "epoch": 2.195609022556391, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 41502 + }, + { + "epoch": 2.1956691729323308, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9208, + "step": 41503 + }, + { + "epoch": 2.195729323308271, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 41504 + }, + { + "epoch": 2.1957894736842105, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 41505 + }, + { + "epoch": 2.19584962406015, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 41506 + }, + { + "epoch": 2.1959097744360903, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 41507 + }, + { + "epoch": 2.19596992481203, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 41508 + }, + { + "epoch": 2.19603007518797, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 41509 + }, + { + "epoch": 2.1960902255639096, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 41510 + }, + { + "epoch": 2.1961503759398497, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 41511 + }, + { + "epoch": 2.1962105263157894, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 41512 + }, + { + "epoch": 2.1962706766917295, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 41513 + }, + { + "epoch": 2.196330827067669, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 41514 + }, + { + "epoch": 2.1963909774436092, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 41515 + }, + { + "epoch": 2.196451127819549, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 41516 + }, + { + "epoch": 2.1965112781954885, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 41517 + }, + { + "epoch": 2.1965714285714286, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 41518 + }, + { + "epoch": 2.1966315789473683, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 41519 + }, + { + "epoch": 2.1966917293233084, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 41520 + }, + { + "epoch": 2.196751879699248, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 41521 + }, + { + "epoch": 2.196812030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 41522 + }, + { + "epoch": 2.1968721804511278, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 41523 + }, + { + "epoch": 2.196932330827068, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 41524 + }, + { + "epoch": 2.1969924812030075, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 41525 + }, + { + "epoch": 2.1970526315789476, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 41526 + }, + { + "epoch": 2.1971127819548872, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 41527 + }, + { + "epoch": 2.197172932330827, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 41528 + }, + { + "epoch": 2.197233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 41529 + }, + { + "epoch": 2.1972932330827066, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 41530 + }, + { + "epoch": 2.1973533834586467, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 41531 + }, + { + "epoch": 2.1974135338345864, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 41532 + }, + { + "epoch": 2.1974736842105265, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 41533 + }, + { + "epoch": 2.197533834586466, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 41534 + }, + { + "epoch": 2.197593984962406, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 41535 + }, + { + "epoch": 2.197654135338346, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9491, + "step": 41536 + }, + { + "epoch": 2.1977142857142855, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 41537 + }, + { + "epoch": 2.1977744360902256, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 41538 + }, + { + "epoch": 2.1978345864661653, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 41539 + }, + { + "epoch": 2.1978947368421053, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 41540 + }, + { + "epoch": 2.197954887218045, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 41541 + }, + { + "epoch": 2.198015037593985, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 41542 + }, + { + "epoch": 2.1980751879699247, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 41543 + }, + { + "epoch": 2.198135338345865, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 41544 + }, + { + "epoch": 2.1981954887218045, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 41545 + }, + { + "epoch": 2.198255639097744, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 41546 + }, + { + "epoch": 2.1983157894736842, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 41547 + }, + { + "epoch": 2.198375939849624, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 41548 + }, + { + "epoch": 2.198436090225564, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 41549 + }, + { + "epoch": 2.1984962406015036, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 41550 + }, + { + "epoch": 2.1985563909774437, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 41551 + }, + { + "epoch": 2.1986165413533834, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 41552 + }, + { + "epoch": 2.1986766917293235, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 41553 + }, + { + "epoch": 2.198736842105263, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 41554 + }, + { + "epoch": 2.198796992481203, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 41555 + }, + { + "epoch": 2.198857142857143, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 41556 + }, + { + "epoch": 2.198917293233083, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 41557 + }, + { + "epoch": 2.1989774436090226, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 41558 + }, + { + "epoch": 2.1990375939849622, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 41559 + }, + { + "epoch": 2.1990977443609023, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 41560 + }, + { + "epoch": 2.199157894736842, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 41561 + }, + { + "epoch": 2.199218045112782, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 41562 + }, + { + "epoch": 2.1992781954887217, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 41563 + }, + { + "epoch": 2.199338345864662, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 41564 + }, + { + "epoch": 2.1993984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 41565 + }, + { + "epoch": 2.1994586466165416, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 41566 + }, + { + "epoch": 2.199518796992481, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 41567 + }, + { + "epoch": 2.199578947368421, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 41568 + }, + { + "epoch": 2.199639097744361, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 41569 + }, + { + "epoch": 2.1996992481203006, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 41570 + }, + { + "epoch": 2.1997593984962407, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 41571 + }, + { + "epoch": 2.1998195488721803, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 41572 + }, + { + "epoch": 2.1998796992481204, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 41573 + }, + { + "epoch": 2.19993984962406, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 41574 + }, + { + "epoch": 2.2, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 41575 + }, + { + "epoch": 2.20006015037594, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1898, + "step": 41576 + }, + { + "epoch": 2.2001203007518795, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 41577 + }, + { + "epoch": 2.2001804511278196, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 41578 + }, + { + "epoch": 2.200240601503759, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 41579 + }, + { + "epoch": 2.2003007518796993, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9069, + "step": 41580 + }, + { + "epoch": 2.200360902255639, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 41581 + }, + { + "epoch": 2.200421052631579, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 41582 + }, + { + "epoch": 2.2004812030075187, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 41583 + }, + { + "epoch": 2.200541353383459, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 41584 + }, + { + "epoch": 2.2006015037593984, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 41585 + }, + { + "epoch": 2.2006616541353385, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 41586 + }, + { + "epoch": 2.200721804511278, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 41587 + }, + { + "epoch": 2.200781954887218, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 41588 + }, + { + "epoch": 2.200842105263158, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 41589 + }, + { + "epoch": 2.2009022556390976, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 41590 + }, + { + "epoch": 2.2009624060150377, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 41591 + }, + { + "epoch": 2.2010225563909773, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 41592 + }, + { + "epoch": 2.2010827067669174, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 41593 + }, + { + "epoch": 2.201142857142857, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 41594 + }, + { + "epoch": 2.201203007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 41595 + }, + { + "epoch": 2.201263157894737, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 41596 + }, + { + "epoch": 2.201323308270677, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9387, + "step": 41597 + }, + { + "epoch": 2.2013834586466166, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 41598 + }, + { + "epoch": 2.201443609022556, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 41599 + }, + { + "epoch": 2.2015037593984963, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 41600 + }, + { + "epoch": 2.201563909774436, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 41601 + }, + { + "epoch": 2.201624060150376, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 41602 + }, + { + "epoch": 2.2016842105263157, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 41603 + }, + { + "epoch": 2.201744360902256, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 41604 + }, + { + "epoch": 2.2018045112781954, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 41605 + }, + { + "epoch": 2.2018646616541355, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 41606 + }, + { + "epoch": 2.201924812030075, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 41607 + }, + { + "epoch": 2.201984962406015, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 41608 + }, + { + "epoch": 2.202045112781955, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 41609 + }, + { + "epoch": 2.2021052631578946, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 41610 + }, + { + "epoch": 2.2021654135338347, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 41611 + }, + { + "epoch": 2.2022255639097743, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 41612 + }, + { + "epoch": 2.2022857142857144, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 41613 + }, + { + "epoch": 2.202345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 41614 + }, + { + "epoch": 2.202406015037594, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 41615 + }, + { + "epoch": 2.202466165413534, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 41616 + }, + { + "epoch": 2.202526315789474, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 41617 + }, + { + "epoch": 2.2025864661654135, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 41618 + }, + { + "epoch": 2.202646616541353, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1459, + "step": 41619 + }, + { + "epoch": 2.2027067669172933, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 41620 + }, + { + "epoch": 2.202766917293233, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 41621 + }, + { + "epoch": 2.202827067669173, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1618, + "step": 41622 + }, + { + "epoch": 2.2028872180451127, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 41623 + }, + { + "epoch": 2.2029473684210528, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 41624 + }, + { + "epoch": 2.2030075187969924, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.8549, + "step": 41625 + }, + { + "epoch": 2.2030676691729325, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2039, + "step": 41626 + }, + { + "epoch": 2.203127819548872, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 41627 + }, + { + "epoch": 2.2031879699248123, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 41628 + }, + { + "epoch": 2.203248120300752, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 41629 + }, + { + "epoch": 2.2033082706766915, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 41630 + }, + { + "epoch": 2.2033684210526316, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 41631 + }, + { + "epoch": 2.2034285714285713, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 41632 + }, + { + "epoch": 2.2034887218045114, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 41633 + }, + { + "epoch": 2.203548872180451, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 41634 + }, + { + "epoch": 2.203609022556391, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 41635 + }, + { + "epoch": 2.2036691729323308, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 41636 + }, + { + "epoch": 2.203729323308271, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 41637 + }, + { + "epoch": 2.2037894736842105, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 41638 + }, + { + "epoch": 2.20384962406015, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 41639 + }, + { + "epoch": 2.2039097744360903, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 41640 + }, + { + "epoch": 2.20396992481203, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 41641 + }, + { + "epoch": 2.20403007518797, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1736, + "step": 41642 + }, + { + "epoch": 2.2040902255639097, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 41643 + }, + { + "epoch": 2.2041503759398497, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 41644 + }, + { + "epoch": 2.2042105263157894, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 41645 + }, + { + "epoch": 2.2042706766917295, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 41646 + }, + { + "epoch": 2.204330827067669, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 41647 + }, + { + "epoch": 2.204390977443609, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 41648 + }, + { + "epoch": 2.204451127819549, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 41649 + }, + { + "epoch": 2.2045112781954885, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 41650 + }, + { + "epoch": 2.2045714285714286, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 41651 + }, + { + "epoch": 2.2046315789473683, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 41652 + }, + { + "epoch": 2.2046917293233084, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 41653 + }, + { + "epoch": 2.204751879699248, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 41654 + }, + { + "epoch": 2.204812030075188, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 41655 + }, + { + "epoch": 2.2048721804511278, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.931, + "step": 41656 + }, + { + "epoch": 2.204932330827068, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 41657 + }, + { + "epoch": 2.2049924812030075, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 41658 + }, + { + "epoch": 2.2050526315789476, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.2054, + "step": 41659 + }, + { + "epoch": 2.2051127819548872, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 41660 + }, + { + "epoch": 2.205172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 41661 + }, + { + "epoch": 2.205233082706767, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 41662 + }, + { + "epoch": 2.2052932330827066, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 41663 + }, + { + "epoch": 2.2053533834586467, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 41664 + }, + { + "epoch": 2.2054135338345864, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 41665 + }, + { + "epoch": 2.2054736842105265, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 41666 + }, + { + "epoch": 2.205533834586466, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 41667 + }, + { + "epoch": 2.205593984962406, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 41668 + }, + { + "epoch": 2.205654135338346, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 41669 + }, + { + "epoch": 2.2057142857142855, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9491, + "step": 41670 + }, + { + "epoch": 2.2057744360902256, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 41671 + }, + { + "epoch": 2.2058345864661653, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 41672 + }, + { + "epoch": 2.2058947368421054, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 41673 + }, + { + "epoch": 2.205954887218045, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 41674 + }, + { + "epoch": 2.206015037593985, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 41675 + }, + { + "epoch": 2.2060751879699247, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 41676 + }, + { + "epoch": 2.206135338345865, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 41677 + }, + { + "epoch": 2.2061954887218045, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 41678 + }, + { + "epoch": 2.206255639097744, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 41679 + }, + { + "epoch": 2.2063157894736842, + "grad_norm": 1.9140625, + "learning_rate": 3e-05, + "loss": 0.9372, + "step": 41680 + }, + { + "epoch": 2.206375939849624, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 41681 + }, + { + "epoch": 2.206436090225564, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 41682 + }, + { + "epoch": 2.2064962406015036, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 41683 + }, + { + "epoch": 2.2065563909774437, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 41684 + }, + { + "epoch": 2.2066165413533834, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 41685 + }, + { + "epoch": 2.2066766917293235, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 41686 + }, + { + "epoch": 2.206736842105263, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 41687 + }, + { + "epoch": 2.206796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 41688 + }, + { + "epoch": 2.206857142857143, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 41689 + }, + { + "epoch": 2.2069172932330825, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 41690 + }, + { + "epoch": 2.2069774436090226, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 41691 + }, + { + "epoch": 2.2070375939849622, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 41692 + }, + { + "epoch": 2.2070977443609023, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 41693 + }, + { + "epoch": 2.207157894736842, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 41694 + }, + { + "epoch": 2.207218045112782, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 41695 + }, + { + "epoch": 2.2072781954887217, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 41696 + }, + { + "epoch": 2.207338345864662, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2193, + "step": 41697 + }, + { + "epoch": 2.2073984962406015, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 41698 + }, + { + "epoch": 2.2074586466165416, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 41699 + }, + { + "epoch": 2.207518796992481, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 41700 + }, + { + "epoch": 2.207578947368421, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 41701 + }, + { + "epoch": 2.207639097744361, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 41702 + }, + { + "epoch": 2.2076992481203006, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 41703 + }, + { + "epoch": 2.2077593984962407, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 41704 + }, + { + "epoch": 2.2078195488721803, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 41705 + }, + { + "epoch": 2.2078796992481204, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 41706 + }, + { + "epoch": 2.20793984962406, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 41707 + }, + { + "epoch": 2.208, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 41708 + }, + { + "epoch": 2.20806015037594, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 41709 + }, + { + "epoch": 2.2081203007518795, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 41710 + }, + { + "epoch": 2.2081804511278196, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 41711 + }, + { + "epoch": 2.2082406015037592, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 41712 + }, + { + "epoch": 2.2083007518796993, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 41713 + }, + { + "epoch": 2.208360902255639, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 41714 + }, + { + "epoch": 2.208421052631579, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 41715 + }, + { + "epoch": 2.2084812030075187, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 41716 + }, + { + "epoch": 2.208541353383459, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 41717 + }, + { + "epoch": 2.2086015037593985, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 41718 + }, + { + "epoch": 2.2086616541353385, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 41719 + }, + { + "epoch": 2.208721804511278, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 41720 + }, + { + "epoch": 2.208781954887218, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 41721 + }, + { + "epoch": 2.208842105263158, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9339, + "step": 41722 + }, + { + "epoch": 2.2089022556390976, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 41723 + }, + { + "epoch": 2.2089624060150377, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 41724 + }, + { + "epoch": 2.2090225563909773, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 41725 + }, + { + "epoch": 2.2090827067669174, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 41726 + }, + { + "epoch": 2.209142857142857, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 41727 + }, + { + "epoch": 2.209203007518797, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 41728 + }, + { + "epoch": 2.209263157894737, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.5624, + "step": 41729 + }, + { + "epoch": 2.209323308270677, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 41730 + }, + { + "epoch": 2.2093834586466166, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.907, + "step": 41731 + }, + { + "epoch": 2.209443609022556, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 41732 + }, + { + "epoch": 2.2095037593984963, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 41733 + }, + { + "epoch": 2.209563909774436, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 41734 + }, + { + "epoch": 2.209624060150376, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 41735 + }, + { + "epoch": 2.2096842105263157, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 41736 + }, + { + "epoch": 2.209744360902256, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 41737 + }, + { + "epoch": 2.2098045112781954, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 41738 + }, + { + "epoch": 2.2098646616541355, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 41739 + }, + { + "epoch": 2.209924812030075, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.932, + "step": 41740 + }, + { + "epoch": 2.209984962406015, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 41741 + }, + { + "epoch": 2.210045112781955, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 41742 + }, + { + "epoch": 2.2101052631578946, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 41743 + }, + { + "epoch": 2.2101654135338347, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 41744 + }, + { + "epoch": 2.2102255639097743, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 41745 + }, + { + "epoch": 2.2102857142857144, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 41746 + }, + { + "epoch": 2.210345864661654, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.902, + "step": 41747 + }, + { + "epoch": 2.210406015037594, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 41748 + }, + { + "epoch": 2.210466165413534, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 41749 + }, + { + "epoch": 2.2105263157894735, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 41750 + }, + { + "epoch": 2.2105864661654135, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.2167, + "step": 41751 + }, + { + "epoch": 2.210646616541353, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 41752 + }, + { + "epoch": 2.2107067669172933, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 41753 + }, + { + "epoch": 2.210766917293233, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 41754 + }, + { + "epoch": 2.210827067669173, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 41755 + }, + { + "epoch": 2.2108872180451127, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 41756 + }, + { + "epoch": 2.2109473684210528, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 41757 + }, + { + "epoch": 2.2110075187969924, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 41758 + }, + { + "epoch": 2.2110676691729325, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 41759 + }, + { + "epoch": 2.211127819548872, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0901, + "step": 41760 + }, + { + "epoch": 2.2111879699248123, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 41761 + }, + { + "epoch": 2.211248120300752, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 41762 + }, + { + "epoch": 2.2113082706766916, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 41763 + }, + { + "epoch": 2.2113684210526316, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 41764 + }, + { + "epoch": 2.2114285714285713, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 41765 + }, + { + "epoch": 2.2114887218045114, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 41766 + }, + { + "epoch": 2.211548872180451, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 41767 + }, + { + "epoch": 2.211609022556391, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 41768 + }, + { + "epoch": 2.211669172932331, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 41769 + }, + { + "epoch": 2.211729323308271, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 41770 + }, + { + "epoch": 2.2117894736842105, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 41771 + }, + { + "epoch": 2.21184962406015, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 41772 + }, + { + "epoch": 2.2119097744360903, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 41773 + }, + { + "epoch": 2.21196992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 41774 + }, + { + "epoch": 2.21203007518797, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 41775 + }, + { + "epoch": 2.2120902255639097, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 41776 + }, + { + "epoch": 2.2121503759398498, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 41777 + }, + { + "epoch": 2.2122105263157894, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 41778 + }, + { + "epoch": 2.2122706766917295, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 41779 + }, + { + "epoch": 2.212330827067669, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 41780 + }, + { + "epoch": 2.212390977443609, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1916, + "step": 41781 + }, + { + "epoch": 2.212451127819549, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 41782 + }, + { + "epoch": 2.2125112781954885, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 41783 + }, + { + "epoch": 2.2125714285714286, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 41784 + }, + { + "epoch": 2.2126315789473683, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.2248, + "step": 41785 + }, + { + "epoch": 2.2126917293233084, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9324, + "step": 41786 + }, + { + "epoch": 2.212751879699248, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 41787 + }, + { + "epoch": 2.212812030075188, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 41788 + }, + { + "epoch": 2.2128721804511278, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 41789 + }, + { + "epoch": 2.212932330827068, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 41790 + }, + { + "epoch": 2.2129924812030075, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 41791 + }, + { + "epoch": 2.213052631578947, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 41792 + }, + { + "epoch": 2.2131127819548873, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 41793 + }, + { + "epoch": 2.213172932330827, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 41794 + }, + { + "epoch": 2.213233082706767, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 41795 + }, + { + "epoch": 2.2132932330827066, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 41796 + }, + { + "epoch": 2.2133533834586467, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 41797 + }, + { + "epoch": 2.2134135338345864, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 41798 + }, + { + "epoch": 2.2134736842105265, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 41799 + }, + { + "epoch": 2.213533834586466, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 41800 + }, + { + "epoch": 2.2135939849624062, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 41801 + }, + { + "epoch": 2.213654135338346, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9362, + "step": 41802 + }, + { + "epoch": 2.2137142857142855, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1535, + "step": 41803 + }, + { + "epoch": 2.2137744360902256, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 41804 + }, + { + "epoch": 2.2138345864661653, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 41805 + }, + { + "epoch": 2.2138947368421054, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 41806 + }, + { + "epoch": 2.213954887218045, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 41807 + }, + { + "epoch": 2.214015037593985, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 41808 + }, + { + "epoch": 2.2140751879699248, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 41809 + }, + { + "epoch": 2.214135338345865, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 41810 + }, + { + "epoch": 2.2141954887218045, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 41811 + }, + { + "epoch": 2.214255639097744, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 41812 + }, + { + "epoch": 2.2143157894736842, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 41813 + }, + { + "epoch": 2.214375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 41814 + }, + { + "epoch": 2.214436090225564, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.204, + "step": 41815 + }, + { + "epoch": 2.2144962406015036, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 41816 + }, + { + "epoch": 2.2145563909774437, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 41817 + }, + { + "epoch": 2.2146165413533834, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 41818 + }, + { + "epoch": 2.2146766917293235, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 41819 + }, + { + "epoch": 2.214736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 41820 + }, + { + "epoch": 2.214796992481203, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9382, + "step": 41821 + }, + { + "epoch": 2.214857142857143, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 41822 + }, + { + "epoch": 2.2149172932330825, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 41823 + }, + { + "epoch": 2.2149774436090226, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 41824 + }, + { + "epoch": 2.2150375939849622, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 41825 + }, + { + "epoch": 2.2150977443609023, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 41826 + }, + { + "epoch": 2.215157894736842, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 41827 + }, + { + "epoch": 2.215218045112782, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 41828 + }, + { + "epoch": 2.2152781954887217, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 41829 + }, + { + "epoch": 2.215338345864662, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 41830 + }, + { + "epoch": 2.2153984962406015, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 41831 + }, + { + "epoch": 2.2154586466165416, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 41832 + }, + { + "epoch": 2.215518796992481, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 41833 + }, + { + "epoch": 2.215578947368421, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 41834 + }, + { + "epoch": 2.215639097744361, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 41835 + }, + { + "epoch": 2.2156992481203006, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 41836 + }, + { + "epoch": 2.2157593984962407, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 41837 + }, + { + "epoch": 2.2158195488721804, + "grad_norm": 1.859375, + "learning_rate": 3e-05, + "loss": 1.1454, + "step": 41838 + }, + { + "epoch": 2.2158796992481204, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 41839 + }, + { + "epoch": 2.21593984962406, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 41840 + }, + { + "epoch": 2.216, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 41841 + }, + { + "epoch": 2.21606015037594, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 41842 + }, + { + "epoch": 2.2161203007518795, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 41843 + }, + { + "epoch": 2.2161804511278196, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 41844 + }, + { + "epoch": 2.2162406015037592, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 41845 + }, + { + "epoch": 2.2163007518796993, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 41846 + }, + { + "epoch": 2.216360902255639, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 41847 + }, + { + "epoch": 2.216421052631579, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 41848 + }, + { + "epoch": 2.2164812030075187, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 41849 + }, + { + "epoch": 2.216541353383459, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 41850 + }, + { + "epoch": 2.2166015037593985, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 41851 + }, + { + "epoch": 2.216661654135338, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 41852 + }, + { + "epoch": 2.216721804511278, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 41853 + }, + { + "epoch": 2.216781954887218, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 41854 + }, + { + "epoch": 2.216842105263158, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 41855 + }, + { + "epoch": 2.2169022556390976, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 41856 + }, + { + "epoch": 2.2169624060150377, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 41857 + }, + { + "epoch": 2.2170225563909773, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 41858 + }, + { + "epoch": 2.2170827067669174, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 41859 + }, + { + "epoch": 2.217142857142857, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 41860 + }, + { + "epoch": 2.217203007518797, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 41861 + }, + { + "epoch": 2.217263157894737, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 41862 + }, + { + "epoch": 2.217323308270677, + "grad_norm": 1.6171875, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 41863 + }, + { + "epoch": 2.2173834586466166, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 41864 + }, + { + "epoch": 2.217443609022556, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 41865 + }, + { + "epoch": 2.2175037593984963, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 41866 + }, + { + "epoch": 2.217563909774436, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 41867 + }, + { + "epoch": 2.217624060150376, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 41868 + }, + { + "epoch": 2.2176842105263157, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 41869 + }, + { + "epoch": 2.217744360902256, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 41870 + }, + { + "epoch": 2.2178045112781954, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 41871 + }, + { + "epoch": 2.2178646616541355, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 41872 + }, + { + "epoch": 2.217924812030075, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 41873 + }, + { + "epoch": 2.217984962406015, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 41874 + }, + { + "epoch": 2.218045112781955, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 41875 + }, + { + "epoch": 2.2181052631578946, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 41876 + }, + { + "epoch": 2.2181654135338347, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 41877 + }, + { + "epoch": 2.2182255639097743, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 41878 + }, + { + "epoch": 2.2182857142857144, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.931, + "step": 41879 + }, + { + "epoch": 2.218345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 41880 + }, + { + "epoch": 2.218406015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 41881 + }, + { + "epoch": 2.218466165413534, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 41882 + }, + { + "epoch": 2.2185263157894735, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 41883 + }, + { + "epoch": 2.2185864661654136, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 41884 + }, + { + "epoch": 2.218646616541353, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 41885 + }, + { + "epoch": 2.2187067669172933, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 41886 + }, + { + "epoch": 2.218766917293233, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 41887 + }, + { + "epoch": 2.218827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 41888 + }, + { + "epoch": 2.2188872180451127, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 41889 + }, + { + "epoch": 2.218947368421053, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 41890 + }, + { + "epoch": 2.2190075187969924, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 41891 + }, + { + "epoch": 2.2190676691729325, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 41892 + }, + { + "epoch": 2.219127819548872, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 41893 + }, + { + "epoch": 2.2191879699248123, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 41894 + }, + { + "epoch": 2.219248120300752, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 41895 + }, + { + "epoch": 2.2193082706766916, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 41896 + }, + { + "epoch": 2.2193684210526317, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 41897 + }, + { + "epoch": 2.2194285714285713, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 41898 + }, + { + "epoch": 2.2194887218045114, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 41899 + }, + { + "epoch": 2.219548872180451, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 41900 + }, + { + "epoch": 2.219609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 41901 + }, + { + "epoch": 2.219669172932331, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 41902 + }, + { + "epoch": 2.219729323308271, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 41903 + }, + { + "epoch": 2.2197894736842105, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1676, + "step": 41904 + }, + { + "epoch": 2.21984962406015, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 41905 + }, + { + "epoch": 2.2199097744360903, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 41906 + }, + { + "epoch": 2.21996992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9305, + "step": 41907 + }, + { + "epoch": 2.22003007518797, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 41908 + }, + { + "epoch": 2.2200902255639097, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 41909 + }, + { + "epoch": 2.2201503759398498, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 41910 + }, + { + "epoch": 2.2202105263157894, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 41911 + }, + { + "epoch": 2.2202706766917295, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 41912 + }, + { + "epoch": 2.220330827067669, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 41913 + }, + { + "epoch": 2.220390977443609, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 41914 + }, + { + "epoch": 2.220451127819549, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 41915 + }, + { + "epoch": 2.2205112781954885, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1446, + "step": 41916 + }, + { + "epoch": 2.2205714285714286, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 41917 + }, + { + "epoch": 2.2206315789473683, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 41918 + }, + { + "epoch": 2.2206917293233084, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9322, + "step": 41919 + }, + { + "epoch": 2.220751879699248, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 41920 + }, + { + "epoch": 2.220812030075188, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 41921 + }, + { + "epoch": 2.2208721804511278, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 41922 + }, + { + "epoch": 2.220932330827068, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 41923 + }, + { + "epoch": 2.2209924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 41924 + }, + { + "epoch": 2.221052631578947, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 41925 + }, + { + "epoch": 2.2211127819548873, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 41926 + }, + { + "epoch": 2.221172932330827, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 41927 + }, + { + "epoch": 2.221233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 41928 + }, + { + "epoch": 2.2212932330827067, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 41929 + }, + { + "epoch": 2.2213533834586467, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 41930 + }, + { + "epoch": 2.2214135338345864, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9247, + "step": 41931 + }, + { + "epoch": 2.2214736842105265, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 41932 + }, + { + "epoch": 2.221533834586466, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9411, + "step": 41933 + }, + { + "epoch": 2.2215939849624062, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 41934 + }, + { + "epoch": 2.221654135338346, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 41935 + }, + { + "epoch": 2.2217142857142855, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 41936 + }, + { + "epoch": 2.2217744360902256, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 41937 + }, + { + "epoch": 2.2218345864661653, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 41938 + }, + { + "epoch": 2.2218947368421054, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 41939 + }, + { + "epoch": 2.221954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 41940 + }, + { + "epoch": 2.222015037593985, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 41941 + }, + { + "epoch": 2.2220751879699248, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 41942 + }, + { + "epoch": 2.222135338345865, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 41943 + }, + { + "epoch": 2.2221954887218045, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 41944 + }, + { + "epoch": 2.222255639097744, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 41945 + }, + { + "epoch": 2.2223157894736842, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 41946 + }, + { + "epoch": 2.222375939849624, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 41947 + }, + { + "epoch": 2.222436090225564, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 41948 + }, + { + "epoch": 2.2224962406015036, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 41949 + }, + { + "epoch": 2.2225563909774437, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 41950 + }, + { + "epoch": 2.2226165413533834, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 41951 + }, + { + "epoch": 2.2226766917293235, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 41952 + }, + { + "epoch": 2.222736842105263, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 41953 + }, + { + "epoch": 2.222796992481203, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 41954 + }, + { + "epoch": 2.222857142857143, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 41955 + }, + { + "epoch": 2.2229172932330825, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 41956 + }, + { + "epoch": 2.2229774436090226, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 41957 + }, + { + "epoch": 2.2230375939849623, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 41958 + }, + { + "epoch": 2.2230977443609024, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 41959 + }, + { + "epoch": 2.223157894736842, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 41960 + }, + { + "epoch": 2.223218045112782, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 41961 + }, + { + "epoch": 2.2232781954887217, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 41962 + }, + { + "epoch": 2.223338345864662, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 41963 + }, + { + "epoch": 2.2233984962406015, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 41964 + }, + { + "epoch": 2.2234586466165416, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 41965 + }, + { + "epoch": 2.2235187969924812, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 41966 + }, + { + "epoch": 2.223578947368421, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 41967 + }, + { + "epoch": 2.223639097744361, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 41968 + }, + { + "epoch": 2.2236992481203006, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 41969 + }, + { + "epoch": 2.2237593984962407, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 41970 + }, + { + "epoch": 2.2238195488721804, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 41971 + }, + { + "epoch": 2.2238796992481205, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 41972 + }, + { + "epoch": 2.22393984962406, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 41973 + }, + { + "epoch": 2.224, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 41974 + }, + { + "epoch": 2.22406015037594, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 41975 + }, + { + "epoch": 2.2241203007518795, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 41976 + }, + { + "epoch": 2.2241804511278196, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 41977 + }, + { + "epoch": 2.2242406015037592, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 41978 + }, + { + "epoch": 2.2243007518796993, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 41979 + }, + { + "epoch": 2.224360902255639, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 41980 + }, + { + "epoch": 2.224421052631579, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 41981 + }, + { + "epoch": 2.2244812030075187, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 41982 + }, + { + "epoch": 2.224541353383459, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 41983 + }, + { + "epoch": 2.2246015037593985, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 41984 + }, + { + "epoch": 2.224661654135338, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 41985 + }, + { + "epoch": 2.224721804511278, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 41986 + }, + { + "epoch": 2.224781954887218, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 41987 + }, + { + "epoch": 2.224842105263158, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 41988 + }, + { + "epoch": 2.2249022556390976, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 41989 + }, + { + "epoch": 2.2249624060150377, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 41990 + }, + { + "epoch": 2.2250225563909773, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 41991 + }, + { + "epoch": 2.2250827067669174, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 41992 + }, + { + "epoch": 2.225142857142857, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 41993 + }, + { + "epoch": 2.225203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 41994 + }, + { + "epoch": 2.225263157894737, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 41995 + }, + { + "epoch": 2.225323308270677, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 41996 + }, + { + "epoch": 2.2253834586466166, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 41997 + }, + { + "epoch": 2.2254436090225562, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 41998 + }, + { + "epoch": 2.2255037593984963, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9187, + "step": 41999 + }, + { + "epoch": 2.225563909774436, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 42000 + }, + { + "epoch": 2.225624060150376, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 42001 + }, + { + "epoch": 2.2256842105263157, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 42002 + }, + { + "epoch": 2.225744360902256, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 42003 + }, + { + "epoch": 2.2258045112781955, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 42004 + }, + { + "epoch": 2.2258646616541355, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 42005 + }, + { + "epoch": 2.225924812030075, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 42006 + }, + { + "epoch": 2.225984962406015, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 42007 + }, + { + "epoch": 2.226045112781955, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 42008 + }, + { + "epoch": 2.2261052631578946, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 42009 + }, + { + "epoch": 2.2261654135338347, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1857, + "step": 42010 + }, + { + "epoch": 2.2262255639097743, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 42011 + }, + { + "epoch": 2.2262857142857144, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 42012 + }, + { + "epoch": 2.226345864661654, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 42013 + }, + { + "epoch": 2.226406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 42014 + }, + { + "epoch": 2.226466165413534, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 42015 + }, + { + "epoch": 2.2265263157894735, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 42016 + }, + { + "epoch": 2.2265864661654136, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 42017 + }, + { + "epoch": 2.226646616541353, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 42018 + }, + { + "epoch": 2.2267067669172933, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 42019 + }, + { + "epoch": 2.226766917293233, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 42020 + }, + { + "epoch": 2.226827067669173, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 42021 + }, + { + "epoch": 2.2268872180451127, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 42022 + }, + { + "epoch": 2.226947368421053, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9445, + "step": 42023 + }, + { + "epoch": 2.2270075187969924, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 42024 + }, + { + "epoch": 2.2270676691729325, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 42025 + }, + { + "epoch": 2.227127819548872, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 42026 + }, + { + "epoch": 2.227187969924812, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 42027 + }, + { + "epoch": 2.227248120300752, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 42028 + }, + { + "epoch": 2.2273082706766916, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 42029 + }, + { + "epoch": 2.2273684210526317, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 42030 + }, + { + "epoch": 2.2274285714285713, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 42031 + }, + { + "epoch": 2.2274887218045114, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 42032 + }, + { + "epoch": 2.227548872180451, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 42033 + }, + { + "epoch": 2.227609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 42034 + }, + { + "epoch": 2.227669172932331, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 42035 + }, + { + "epoch": 2.227729323308271, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 42036 + }, + { + "epoch": 2.2277894736842105, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 42037 + }, + { + "epoch": 2.22784962406015, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 42038 + }, + { + "epoch": 2.2279097744360903, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 42039 + }, + { + "epoch": 2.22796992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 42040 + }, + { + "epoch": 2.22803007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 42041 + }, + { + "epoch": 2.2280902255639097, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 42042 + }, + { + "epoch": 2.2281503759398498, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 42043 + }, + { + "epoch": 2.2282105263157894, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 42044 + }, + { + "epoch": 2.2282706766917295, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 42045 + }, + { + "epoch": 2.228330827067669, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 42046 + }, + { + "epoch": 2.228390977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 42047 + }, + { + "epoch": 2.228451127819549, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 42048 + }, + { + "epoch": 2.2285112781954886, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 42049 + }, + { + "epoch": 2.2285714285714286, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 42050 + }, + { + "epoch": 2.2286315789473683, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 42051 + }, + { + "epoch": 2.2286917293233084, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 42052 + }, + { + "epoch": 2.228751879699248, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 42053 + }, + { + "epoch": 2.228812030075188, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 42054 + }, + { + "epoch": 2.228872180451128, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 42055 + }, + { + "epoch": 2.228932330827068, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 42056 + }, + { + "epoch": 2.2289924812030075, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 42057 + }, + { + "epoch": 2.229052631578947, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 42058 + }, + { + "epoch": 2.2291127819548873, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9096, + "step": 42059 + }, + { + "epoch": 2.229172932330827, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 42060 + }, + { + "epoch": 2.229233082706767, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 42061 + }, + { + "epoch": 2.2292932330827067, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 42062 + }, + { + "epoch": 2.2293533834586468, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 42063 + }, + { + "epoch": 2.2294135338345864, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 42064 + }, + { + "epoch": 2.2294736842105265, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 42065 + }, + { + "epoch": 2.229533834586466, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 42066 + }, + { + "epoch": 2.2295939849624062, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.3078, + "step": 42067 + }, + { + "epoch": 2.229654135338346, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 42068 + }, + { + "epoch": 2.2297142857142855, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 42069 + }, + { + "epoch": 2.2297744360902256, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 42070 + }, + { + "epoch": 2.2298345864661653, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 42071 + }, + { + "epoch": 2.2298947368421054, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9232, + "step": 42072 + }, + { + "epoch": 2.229954887218045, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 42073 + }, + { + "epoch": 2.230015037593985, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 42074 + }, + { + "epoch": 2.2300751879699248, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 42075 + }, + { + "epoch": 2.230135338345865, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 42076 + }, + { + "epoch": 2.2301954887218045, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 42077 + }, + { + "epoch": 2.230255639097744, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 42078 + }, + { + "epoch": 2.2303157894736843, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9277, + "step": 42079 + }, + { + "epoch": 2.230375939849624, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 42080 + }, + { + "epoch": 2.230436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 42081 + }, + { + "epoch": 2.2304962406015036, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 42082 + }, + { + "epoch": 2.2305563909774437, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 42083 + }, + { + "epoch": 2.2306165413533834, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 42084 + }, + { + "epoch": 2.2306766917293235, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 42085 + }, + { + "epoch": 2.230736842105263, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 42086 + }, + { + "epoch": 2.230796992481203, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 42087 + }, + { + "epoch": 2.230857142857143, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 42088 + }, + { + "epoch": 2.2309172932330825, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 42089 + }, + { + "epoch": 2.2309774436090226, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 42090 + }, + { + "epoch": 2.2310375939849623, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 42091 + }, + { + "epoch": 2.2310977443609024, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 42092 + }, + { + "epoch": 2.231157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 42093 + }, + { + "epoch": 2.231218045112782, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 42094 + }, + { + "epoch": 2.2312781954887217, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 42095 + }, + { + "epoch": 2.231338345864662, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 42096 + }, + { + "epoch": 2.2313984962406015, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 42097 + }, + { + "epoch": 2.2314586466165416, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 42098 + }, + { + "epoch": 2.2315187969924812, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 42099 + }, + { + "epoch": 2.231578947368421, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 42100 + }, + { + "epoch": 2.231639097744361, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 42101 + }, + { + "epoch": 2.2316992481203006, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 42102 + }, + { + "epoch": 2.2317593984962407, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 42103 + }, + { + "epoch": 2.2318195488721804, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 42104 + }, + { + "epoch": 2.2318796992481205, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 42105 + }, + { + "epoch": 2.23193984962406, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 42106 + }, + { + "epoch": 2.232, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1308, + "step": 42107 + }, + { + "epoch": 2.23206015037594, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 42108 + }, + { + "epoch": 2.2321203007518795, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 42109 + }, + { + "epoch": 2.2321804511278196, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 42110 + }, + { + "epoch": 2.2322406015037592, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 42111 + }, + { + "epoch": 2.2323007518796993, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 42112 + }, + { + "epoch": 2.232360902255639, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 42113 + }, + { + "epoch": 2.232421052631579, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 42114 + }, + { + "epoch": 2.2324812030075187, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 42115 + }, + { + "epoch": 2.232541353383459, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 42116 + }, + { + "epoch": 2.2326015037593985, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.149, + "step": 42117 + }, + { + "epoch": 2.232661654135338, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 42118 + }, + { + "epoch": 2.232721804511278, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 42119 + }, + { + "epoch": 2.232781954887218, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 42120 + }, + { + "epoch": 2.232842105263158, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 42121 + }, + { + "epoch": 2.2329022556390976, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 42122 + }, + { + "epoch": 2.2329624060150377, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 42123 + }, + { + "epoch": 2.2330225563909774, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 42124 + }, + { + "epoch": 2.2330827067669174, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 42125 + }, + { + "epoch": 2.233142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 42126 + }, + { + "epoch": 2.233203007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 42127 + }, + { + "epoch": 2.233263157894737, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 42128 + }, + { + "epoch": 2.2333233082706765, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 42129 + }, + { + "epoch": 2.2333834586466166, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 42130 + }, + { + "epoch": 2.2334436090225562, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1844, + "step": 42131 + }, + { + "epoch": 2.2335037593984963, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 42132 + }, + { + "epoch": 2.233563909774436, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 42133 + }, + { + "epoch": 2.233624060150376, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 42134 + }, + { + "epoch": 2.2336842105263157, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 42135 + }, + { + "epoch": 2.233744360902256, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 42136 + }, + { + "epoch": 2.2338045112781955, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 42137 + }, + { + "epoch": 2.2338646616541356, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 42138 + }, + { + "epoch": 2.233924812030075, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 42139 + }, + { + "epoch": 2.233984962406015, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1772, + "step": 42140 + }, + { + "epoch": 2.234045112781955, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 42141 + }, + { + "epoch": 2.2341052631578946, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 42142 + }, + { + "epoch": 2.2341654135338347, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9433, + "step": 42143 + }, + { + "epoch": 2.2342255639097743, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 42144 + }, + { + "epoch": 2.2342857142857144, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 42145 + }, + { + "epoch": 2.234345864661654, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 42146 + }, + { + "epoch": 2.234406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 42147 + }, + { + "epoch": 2.234466165413534, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 42148 + }, + { + "epoch": 2.2345263157894735, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 42149 + }, + { + "epoch": 2.2345864661654136, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 42150 + }, + { + "epoch": 2.234646616541353, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 42151 + }, + { + "epoch": 2.2347067669172933, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 42152 + }, + { + "epoch": 2.234766917293233, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 42153 + }, + { + "epoch": 2.234827067669173, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 42154 + }, + { + "epoch": 2.2348872180451127, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 42155 + }, + { + "epoch": 2.234947368421053, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 42156 + }, + { + "epoch": 2.2350075187969924, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 42157 + }, + { + "epoch": 2.2350676691729325, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 42158 + }, + { + "epoch": 2.235127819548872, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 42159 + }, + { + "epoch": 2.235187969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 42160 + }, + { + "epoch": 2.235248120300752, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 42161 + }, + { + "epoch": 2.2353082706766916, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 42162 + }, + { + "epoch": 2.2353684210526317, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 42163 + }, + { + "epoch": 2.2354285714285713, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 42164 + }, + { + "epoch": 2.2354887218045114, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 42165 + }, + { + "epoch": 2.235548872180451, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 42166 + }, + { + "epoch": 2.235609022556391, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 42167 + }, + { + "epoch": 2.235669172932331, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 42168 + }, + { + "epoch": 2.235729323308271, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 42169 + }, + { + "epoch": 2.2357894736842105, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 42170 + }, + { + "epoch": 2.23584962406015, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 42171 + }, + { + "epoch": 2.2359097744360903, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 42172 + }, + { + "epoch": 2.23596992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 42173 + }, + { + "epoch": 2.23603007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 42174 + }, + { + "epoch": 2.2360902255639097, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 42175 + }, + { + "epoch": 2.2361503759398498, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 42176 + }, + { + "epoch": 2.2362105263157894, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 42177 + }, + { + "epoch": 2.2362706766917295, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 42178 + }, + { + "epoch": 2.236330827067669, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 42179 + }, + { + "epoch": 2.236390977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 42180 + }, + { + "epoch": 2.236451127819549, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 42181 + }, + { + "epoch": 2.2365112781954886, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 42182 + }, + { + "epoch": 2.2365714285714287, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 42183 + }, + { + "epoch": 2.2366315789473683, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 42184 + }, + { + "epoch": 2.2366917293233084, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 42185 + }, + { + "epoch": 2.236751879699248, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 42186 + }, + { + "epoch": 2.236812030075188, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 42187 + }, + { + "epoch": 2.236872180451128, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 42188 + }, + { + "epoch": 2.2369323308270674, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 42189 + }, + { + "epoch": 2.2369924812030075, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 42190 + }, + { + "epoch": 2.237052631578947, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 42191 + }, + { + "epoch": 2.2371127819548873, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 42192 + }, + { + "epoch": 2.237172932330827, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 42193 + }, + { + "epoch": 2.237233082706767, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 42194 + }, + { + "epoch": 2.2372932330827067, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 42195 + }, + { + "epoch": 2.2373533834586468, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 42196 + }, + { + "epoch": 2.2374135338345864, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 42197 + }, + { + "epoch": 2.2374736842105265, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 42198 + }, + { + "epoch": 2.237533834586466, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 42199 + }, + { + "epoch": 2.2375939849624062, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 42200 + }, + { + "epoch": 2.237654135338346, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 42201 + }, + { + "epoch": 2.2377142857142855, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 42202 + }, + { + "epoch": 2.2377744360902256, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 42203 + }, + { + "epoch": 2.2378345864661653, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 42204 + }, + { + "epoch": 2.2378947368421054, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 42205 + }, + { + "epoch": 2.237954887218045, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 42206 + }, + { + "epoch": 2.238015037593985, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9438, + "step": 42207 + }, + { + "epoch": 2.2380751879699248, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 42208 + }, + { + "epoch": 2.238135338345865, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 42209 + }, + { + "epoch": 2.2381954887218045, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 42210 + }, + { + "epoch": 2.238255639097744, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 42211 + }, + { + "epoch": 2.2383157894736843, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 42212 + }, + { + "epoch": 2.238375939849624, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 42213 + }, + { + "epoch": 2.238436090225564, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 42214 + }, + { + "epoch": 2.2384962406015037, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 42215 + }, + { + "epoch": 2.2385563909774437, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 42216 + }, + { + "epoch": 2.2386165413533834, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 42217 + }, + { + "epoch": 2.2386766917293235, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 42218 + }, + { + "epoch": 2.238736842105263, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 42219 + }, + { + "epoch": 2.238796992481203, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 42220 + }, + { + "epoch": 2.238857142857143, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 42221 + }, + { + "epoch": 2.2389172932330825, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 42222 + }, + { + "epoch": 2.2389774436090226, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 42223 + }, + { + "epoch": 2.2390375939849623, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 42224 + }, + { + "epoch": 2.2390977443609024, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 42225 + }, + { + "epoch": 2.239157894736842, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 42226 + }, + { + "epoch": 2.239218045112782, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 42227 + }, + { + "epoch": 2.2392781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 42228 + }, + { + "epoch": 2.239338345864662, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9386, + "step": 42229 + }, + { + "epoch": 2.2393984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 42230 + }, + { + "epoch": 2.239458646616541, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 42231 + }, + { + "epoch": 2.2395187969924812, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.8904, + "step": 42232 + }, + { + "epoch": 2.239578947368421, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 42233 + }, + { + "epoch": 2.239639097744361, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 42234 + }, + { + "epoch": 2.2396992481203006, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 42235 + }, + { + "epoch": 2.2397593984962407, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 42236 + }, + { + "epoch": 2.2398195488721804, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 42237 + }, + { + "epoch": 2.2398796992481205, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 42238 + }, + { + "epoch": 2.23993984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 42239 + }, + { + "epoch": 2.24, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 42240 + }, + { + "epoch": 2.24006015037594, + "grad_norm": 1.890625, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 42241 + }, + { + "epoch": 2.2401203007518795, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 42242 + }, + { + "epoch": 2.2401804511278196, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 42243 + }, + { + "epoch": 2.2402406015037593, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 42244 + }, + { + "epoch": 2.2403007518796993, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 42245 + }, + { + "epoch": 2.240360902255639, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 42246 + }, + { + "epoch": 2.240421052631579, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1867, + "step": 42247 + }, + { + "epoch": 2.2404812030075187, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 42248 + }, + { + "epoch": 2.240541353383459, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 42249 + }, + { + "epoch": 2.2406015037593985, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 42250 + }, + { + "epoch": 2.240661654135338, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 42251 + }, + { + "epoch": 2.2407218045112782, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 42252 + }, + { + "epoch": 2.240781954887218, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 42253 + }, + { + "epoch": 2.240842105263158, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 42254 + }, + { + "epoch": 2.2409022556390976, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9518, + "step": 42255 + }, + { + "epoch": 2.2409624060150377, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 42256 + }, + { + "epoch": 2.2410225563909774, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 42257 + }, + { + "epoch": 2.2410827067669175, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 42258 + }, + { + "epoch": 2.241142857142857, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 42259 + }, + { + "epoch": 2.241203007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 42260 + }, + { + "epoch": 2.241263157894737, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 42261 + }, + { + "epoch": 2.2413233082706765, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 42262 + }, + { + "epoch": 2.2413834586466166, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.2193, + "step": 42263 + }, + { + "epoch": 2.2414436090225562, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 42264 + }, + { + "epoch": 2.2415037593984963, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 42265 + }, + { + "epoch": 2.241563909774436, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 42266 + }, + { + "epoch": 2.241624060150376, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 42267 + }, + { + "epoch": 2.2416842105263157, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 42268 + }, + { + "epoch": 2.241744360902256, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 42269 + }, + { + "epoch": 2.2418045112781955, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 42270 + }, + { + "epoch": 2.2418646616541356, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 42271 + }, + { + "epoch": 2.241924812030075, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 42272 + }, + { + "epoch": 2.241984962406015, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 42273 + }, + { + "epoch": 2.242045112781955, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1804, + "step": 42274 + }, + { + "epoch": 2.2421052631578946, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 42275 + }, + { + "epoch": 2.2421654135338347, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 42276 + }, + { + "epoch": 2.2422255639097743, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 42277 + }, + { + "epoch": 2.2422857142857144, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 42278 + }, + { + "epoch": 2.242345864661654, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 42279 + }, + { + "epoch": 2.242406015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 42280 + }, + { + "epoch": 2.242466165413534, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 42281 + }, + { + "epoch": 2.2425263157894735, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 42282 + }, + { + "epoch": 2.2425864661654136, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 42283 + }, + { + "epoch": 2.242646616541353, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 42284 + }, + { + "epoch": 2.2427067669172933, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 42285 + }, + { + "epoch": 2.242766917293233, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 42286 + }, + { + "epoch": 2.242827067669173, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 42287 + }, + { + "epoch": 2.2428872180451127, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.946, + "step": 42288 + }, + { + "epoch": 2.242947368421053, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 42289 + }, + { + "epoch": 2.2430075187969924, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 42290 + }, + { + "epoch": 2.243067669172932, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 42291 + }, + { + "epoch": 2.243127819548872, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 42292 + }, + { + "epoch": 2.243187969924812, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 42293 + }, + { + "epoch": 2.243248120300752, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 42294 + }, + { + "epoch": 2.2433082706766916, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 42295 + }, + { + "epoch": 2.2433684210526317, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 42296 + }, + { + "epoch": 2.2434285714285713, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 42297 + }, + { + "epoch": 2.2434887218045114, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 42298 + }, + { + "epoch": 2.243548872180451, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 42299 + }, + { + "epoch": 2.243609022556391, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 42300 + }, + { + "epoch": 2.243669172932331, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 42301 + }, + { + "epoch": 2.243729323308271, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 42302 + }, + { + "epoch": 2.2437894736842106, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 42303 + }, + { + "epoch": 2.24384962406015, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 42304 + }, + { + "epoch": 2.2439097744360903, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 42305 + }, + { + "epoch": 2.24396992481203, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 42306 + }, + { + "epoch": 2.24403007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 42307 + }, + { + "epoch": 2.2440902255639097, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 42308 + }, + { + "epoch": 2.24415037593985, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2126, + "step": 42309 + }, + { + "epoch": 2.2442105263157894, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9408, + "step": 42310 + }, + { + "epoch": 2.2442706766917295, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 42311 + }, + { + "epoch": 2.244330827067669, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 42312 + }, + { + "epoch": 2.244390977443609, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 42313 + }, + { + "epoch": 2.244451127819549, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 42314 + }, + { + "epoch": 2.2445112781954886, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 42315 + }, + { + "epoch": 2.2445714285714287, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 42316 + }, + { + "epoch": 2.2446315789473683, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 42317 + }, + { + "epoch": 2.2446917293233084, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 42318 + }, + { + "epoch": 2.244751879699248, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 42319 + }, + { + "epoch": 2.244812030075188, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 42320 + }, + { + "epoch": 2.244872180451128, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.936, + "step": 42321 + }, + { + "epoch": 2.2449323308270674, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 42322 + }, + { + "epoch": 2.2449924812030075, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 42323 + }, + { + "epoch": 2.245052631578947, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 42324 + }, + { + "epoch": 2.2451127819548873, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 42325 + }, + { + "epoch": 2.245172932330827, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.8839, + "step": 42326 + }, + { + "epoch": 2.245233082706767, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 42327 + }, + { + "epoch": 2.2452932330827067, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 42328 + }, + { + "epoch": 2.2453533834586468, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 42329 + }, + { + "epoch": 2.2454135338345864, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 42330 + }, + { + "epoch": 2.2454736842105265, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 42331 + }, + { + "epoch": 2.245533834586466, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 42332 + }, + { + "epoch": 2.245593984962406, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 42333 + }, + { + "epoch": 2.245654135338346, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 42334 + }, + { + "epoch": 2.2457142857142856, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 42335 + }, + { + "epoch": 2.2457744360902256, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 42336 + }, + { + "epoch": 2.2458345864661653, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 42337 + }, + { + "epoch": 2.2458947368421054, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 42338 + }, + { + "epoch": 2.245954887218045, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 42339 + }, + { + "epoch": 2.246015037593985, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1811, + "step": 42340 + }, + { + "epoch": 2.246075187969925, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 42341 + }, + { + "epoch": 2.246135338345865, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 42342 + }, + { + "epoch": 2.2461954887218045, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 42343 + }, + { + "epoch": 2.246255639097744, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 42344 + }, + { + "epoch": 2.2463157894736843, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 42345 + }, + { + "epoch": 2.246375939849624, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 42346 + }, + { + "epoch": 2.246436090225564, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 42347 + }, + { + "epoch": 2.2464962406015037, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 42348 + }, + { + "epoch": 2.2465563909774438, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 42349 + }, + { + "epoch": 2.2466165413533834, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 42350 + }, + { + "epoch": 2.2466766917293235, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 42351 + }, + { + "epoch": 2.246736842105263, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 42352 + }, + { + "epoch": 2.246796992481203, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 42353 + }, + { + "epoch": 2.246857142857143, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 42354 + }, + { + "epoch": 2.2469172932330825, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 42355 + }, + { + "epoch": 2.2469774436090226, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1388, + "step": 42356 + }, + { + "epoch": 2.2470375939849623, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 42357 + }, + { + "epoch": 2.2470977443609024, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 42358 + }, + { + "epoch": 2.247157894736842, + "grad_norm": 1.5625, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 42359 + }, + { + "epoch": 2.247218045112782, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 42360 + }, + { + "epoch": 2.2472781954887218, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 42361 + }, + { + "epoch": 2.247338345864662, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 42362 + }, + { + "epoch": 2.2473984962406015, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.1149, + "step": 42363 + }, + { + "epoch": 2.247458646616541, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 42364 + }, + { + "epoch": 2.2475187969924812, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 42365 + }, + { + "epoch": 2.247578947368421, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 42366 + }, + { + "epoch": 2.247639097744361, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9244, + "step": 42367 + }, + { + "epoch": 2.2476992481203006, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 42368 + }, + { + "epoch": 2.2477593984962407, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 42369 + }, + { + "epoch": 2.2478195488721804, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 42370 + }, + { + "epoch": 2.2478796992481205, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 42371 + }, + { + "epoch": 2.24793984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 42372 + }, + { + "epoch": 2.248, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 42373 + }, + { + "epoch": 2.24806015037594, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 42374 + }, + { + "epoch": 2.2481203007518795, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 42375 + }, + { + "epoch": 2.2481804511278196, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 42376 + }, + { + "epoch": 2.2482406015037593, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 42377 + }, + { + "epoch": 2.2483007518796994, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 42378 + }, + { + "epoch": 2.248360902255639, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 42379 + }, + { + "epoch": 2.248421052631579, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 42380 + }, + { + "epoch": 2.2484812030075187, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 42381 + }, + { + "epoch": 2.248541353383459, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 42382 + }, + { + "epoch": 2.2486015037593985, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 42383 + }, + { + "epoch": 2.248661654135338, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 42384 + }, + { + "epoch": 2.2487218045112782, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 42385 + }, + { + "epoch": 2.248781954887218, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 42386 + }, + { + "epoch": 2.248842105263158, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 42387 + }, + { + "epoch": 2.2489022556390976, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 42388 + }, + { + "epoch": 2.2489624060150377, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 42389 + }, + { + "epoch": 2.2490225563909774, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1767, + "step": 42390 + }, + { + "epoch": 2.2490827067669175, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 42391 + }, + { + "epoch": 2.249142857142857, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1735, + "step": 42392 + }, + { + "epoch": 2.2492030075187968, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 42393 + }, + { + "epoch": 2.249263157894737, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 42394 + }, + { + "epoch": 2.2493233082706765, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 42395 + }, + { + "epoch": 2.2493834586466166, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 42396 + }, + { + "epoch": 2.2494436090225562, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 42397 + }, + { + "epoch": 2.2495037593984963, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 42398 + }, + { + "epoch": 2.249563909774436, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 42399 + }, + { + "epoch": 2.249624060150376, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 42400 + }, + { + "epoch": 2.2496842105263157, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 42401 + }, + { + "epoch": 2.249744360902256, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 42402 + }, + { + "epoch": 2.2498045112781955, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 42403 + }, + { + "epoch": 2.2498646616541356, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 42404 + }, + { + "epoch": 2.249924812030075, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 42405 + }, + { + "epoch": 2.249984962406015, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 42406 + }, + { + "epoch": 2.250045112781955, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 42407 + }, + { + "epoch": 2.2501052631578946, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 42408 + }, + { + "epoch": 2.2501654135338347, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 42409 + }, + { + "epoch": 2.2502255639097744, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 42410 + }, + { + "epoch": 2.2502857142857144, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 42411 + }, + { + "epoch": 2.250345864661654, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 42412 + }, + { + "epoch": 2.250406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 42413 + }, + { + "epoch": 2.250466165413534, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 42414 + }, + { + "epoch": 2.2505263157894735, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 42415 + }, + { + "epoch": 2.2505864661654136, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 42416 + }, + { + "epoch": 2.2506466165413532, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 42417 + }, + { + "epoch": 2.2507067669172933, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 42418 + }, + { + "epoch": 2.250766917293233, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 42419 + }, + { + "epoch": 2.250827067669173, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 42420 + }, + { + "epoch": 2.2508872180451127, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 42421 + }, + { + "epoch": 2.250947368421053, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9533, + "step": 42422 + }, + { + "epoch": 2.2510075187969925, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 42423 + }, + { + "epoch": 2.251067669172932, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 42424 + }, + { + "epoch": 2.251127819548872, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 42425 + }, + { + "epoch": 2.251187969924812, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 42426 + }, + { + "epoch": 2.251248120300752, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 42427 + }, + { + "epoch": 2.2513082706766916, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 42428 + }, + { + "epoch": 2.2513684210526317, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 42429 + }, + { + "epoch": 2.2514285714285713, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 42430 + }, + { + "epoch": 2.2514887218045114, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 42431 + }, + { + "epoch": 2.251548872180451, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 42432 + }, + { + "epoch": 2.251609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 42433 + }, + { + "epoch": 2.251669172932331, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 42434 + }, + { + "epoch": 2.251729323308271, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 42435 + }, + { + "epoch": 2.2517894736842106, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 42436 + }, + { + "epoch": 2.25184962406015, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 42437 + }, + { + "epoch": 2.2519097744360903, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 42438 + }, + { + "epoch": 2.25196992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 42439 + }, + { + "epoch": 2.25203007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 42440 + }, + { + "epoch": 2.2520902255639097, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 42441 + }, + { + "epoch": 2.25215037593985, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 42442 + }, + { + "epoch": 2.2522105263157894, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 42443 + }, + { + "epoch": 2.2522706766917295, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 42444 + }, + { + "epoch": 2.252330827067669, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 42445 + }, + { + "epoch": 2.252390977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 42446 + }, + { + "epoch": 2.252451127819549, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 42447 + }, + { + "epoch": 2.2525112781954886, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 42448 + }, + { + "epoch": 2.2525714285714287, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 42449 + }, + { + "epoch": 2.2526315789473683, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 42450 + }, + { + "epoch": 2.2526917293233084, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 42451 + }, + { + "epoch": 2.252751879699248, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 42452 + }, + { + "epoch": 2.252812030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 42453 + }, + { + "epoch": 2.252872180451128, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 42454 + }, + { + "epoch": 2.2529323308270675, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 42455 + }, + { + "epoch": 2.2529924812030075, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 42456 + }, + { + "epoch": 2.253052631578947, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9351, + "step": 42457 + }, + { + "epoch": 2.2531127819548873, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 42458 + }, + { + "epoch": 2.253172932330827, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 42459 + }, + { + "epoch": 2.253233082706767, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 42460 + }, + { + "epoch": 2.2532932330827067, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 42461 + }, + { + "epoch": 2.2533533834586468, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 42462 + }, + { + "epoch": 2.2534135338345864, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 42463 + }, + { + "epoch": 2.253473684210526, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 42464 + }, + { + "epoch": 2.253533834586466, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9331, + "step": 42465 + }, + { + "epoch": 2.2535939849624063, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 42466 + }, + { + "epoch": 2.253654135338346, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.121, + "step": 42467 + }, + { + "epoch": 2.2537142857142856, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 42468 + }, + { + "epoch": 2.2537744360902257, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 42469 + }, + { + "epoch": 2.2538345864661653, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 42470 + }, + { + "epoch": 2.2538947368421054, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 42471 + }, + { + "epoch": 2.253954887218045, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.8898, + "step": 42472 + }, + { + "epoch": 2.254015037593985, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 42473 + }, + { + "epoch": 2.254075187969925, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 42474 + }, + { + "epoch": 2.254135338345865, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 42475 + }, + { + "epoch": 2.2541954887218045, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 42476 + }, + { + "epoch": 2.254255639097744, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 42477 + }, + { + "epoch": 2.2543157894736843, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 42478 + }, + { + "epoch": 2.254375939849624, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 42479 + }, + { + "epoch": 2.254436090225564, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 42480 + }, + { + "epoch": 2.2544962406015037, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 42481 + }, + { + "epoch": 2.2545563909774438, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 42482 + }, + { + "epoch": 2.2546165413533834, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 42483 + }, + { + "epoch": 2.2546766917293235, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9445, + "step": 42484 + }, + { + "epoch": 2.254736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 42485 + }, + { + "epoch": 2.254796992481203, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 42486 + }, + { + "epoch": 2.254857142857143, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 42487 + }, + { + "epoch": 2.2549172932330825, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0877, + "step": 42488 + }, + { + "epoch": 2.2549774436090226, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 42489 + }, + { + "epoch": 2.2550375939849623, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 42490 + }, + { + "epoch": 2.2550977443609024, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 42491 + }, + { + "epoch": 2.255157894736842, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 42492 + }, + { + "epoch": 2.255218045112782, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9321, + "step": 42493 + }, + { + "epoch": 2.2552781954887218, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 42494 + }, + { + "epoch": 2.2553383458646614, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 42495 + }, + { + "epoch": 2.2553984962406015, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 42496 + }, + { + "epoch": 2.255458646616541, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 42497 + }, + { + "epoch": 2.2555187969924813, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 42498 + }, + { + "epoch": 2.255578947368421, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 42499 + }, + { + "epoch": 2.255639097744361, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.8955, + "step": 42500 + }, + { + "epoch": 2.2556992481203006, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 42501 + }, + { + "epoch": 2.2557593984962407, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 42502 + }, + { + "epoch": 2.2558195488721804, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 42503 + }, + { + "epoch": 2.2558796992481205, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 42504 + }, + { + "epoch": 2.25593984962406, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 42505 + }, + { + "epoch": 2.2560000000000002, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 42506 + }, + { + "epoch": 2.25606015037594, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.158, + "step": 42507 + }, + { + "epoch": 2.2561203007518795, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 42508 + }, + { + "epoch": 2.2561804511278196, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 42509 + }, + { + "epoch": 2.2562406015037593, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 42510 + }, + { + "epoch": 2.2563007518796994, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 42511 + }, + { + "epoch": 2.256360902255639, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 42512 + }, + { + "epoch": 2.256421052631579, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 42513 + }, + { + "epoch": 2.2564812030075188, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 42514 + }, + { + "epoch": 2.256541353383459, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 42515 + }, + { + "epoch": 2.2566015037593985, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 42516 + }, + { + "epoch": 2.256661654135338, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 42517 + }, + { + "epoch": 2.2567218045112782, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 42518 + }, + { + "epoch": 2.256781954887218, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1672, + "step": 42519 + }, + { + "epoch": 2.256842105263158, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 42520 + }, + { + "epoch": 2.2569022556390976, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 42521 + }, + { + "epoch": 2.2569624060150377, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 42522 + }, + { + "epoch": 2.2570225563909774, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 42523 + }, + { + "epoch": 2.2570827067669175, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 42524 + }, + { + "epoch": 2.257142857142857, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 42525 + }, + { + "epoch": 2.2572030075187968, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 42526 + }, + { + "epoch": 2.257263157894737, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 42527 + }, + { + "epoch": 2.2573233082706765, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 42528 + }, + { + "epoch": 2.2573834586466166, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9297, + "step": 42529 + }, + { + "epoch": 2.2574436090225563, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 42530 + }, + { + "epoch": 2.2575037593984963, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 42531 + }, + { + "epoch": 2.257563909774436, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 42532 + }, + { + "epoch": 2.257624060150376, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 42533 + }, + { + "epoch": 2.2576842105263157, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 42534 + }, + { + "epoch": 2.257744360902256, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 42535 + }, + { + "epoch": 2.2578045112781955, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 42536 + }, + { + "epoch": 2.2578646616541356, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 42537 + }, + { + "epoch": 2.2579248120300752, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 0.9518, + "step": 42538 + }, + { + "epoch": 2.257984962406015, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 42539 + }, + { + "epoch": 2.258045112781955, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 42540 + }, + { + "epoch": 2.2581052631578946, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 42541 + }, + { + "epoch": 2.2581654135338347, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 42542 + }, + { + "epoch": 2.2582255639097744, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 42543 + }, + { + "epoch": 2.2582857142857145, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 42544 + }, + { + "epoch": 2.258345864661654, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 42545 + }, + { + "epoch": 2.258406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 42546 + }, + { + "epoch": 2.258466165413534, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 42547 + }, + { + "epoch": 2.2585263157894735, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 42548 + }, + { + "epoch": 2.2585864661654136, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 42549 + }, + { + "epoch": 2.2586466165413532, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 42550 + }, + { + "epoch": 2.2587067669172933, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 42551 + }, + { + "epoch": 2.258766917293233, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 42552 + }, + { + "epoch": 2.258827067669173, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9019, + "step": 42553 + }, + { + "epoch": 2.2588872180451127, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 42554 + }, + { + "epoch": 2.258947368421053, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 42555 + }, + { + "epoch": 2.2590075187969925, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9114, + "step": 42556 + }, + { + "epoch": 2.259067669172932, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 42557 + }, + { + "epoch": 2.259127819548872, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 42558 + }, + { + "epoch": 2.259187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 42559 + }, + { + "epoch": 2.259248120300752, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 42560 + }, + { + "epoch": 2.2593082706766916, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 42561 + }, + { + "epoch": 2.2593684210526317, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 42562 + }, + { + "epoch": 2.2594285714285713, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 42563 + }, + { + "epoch": 2.2594887218045114, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.948, + "step": 42564 + }, + { + "epoch": 2.259548872180451, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 42565 + }, + { + "epoch": 2.2596090225563907, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 42566 + }, + { + "epoch": 2.259669172932331, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 42567 + }, + { + "epoch": 2.259729323308271, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 42568 + }, + { + "epoch": 2.2597894736842106, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 42569 + }, + { + "epoch": 2.25984962406015, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 42570 + }, + { + "epoch": 2.2599097744360903, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 42571 + }, + { + "epoch": 2.25996992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 42572 + }, + { + "epoch": 2.26003007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 42573 + }, + { + "epoch": 2.2600902255639097, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 42574 + }, + { + "epoch": 2.26015037593985, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 42575 + }, + { + "epoch": 2.2602105263157894, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 42576 + }, + { + "epoch": 2.2602706766917295, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 42577 + }, + { + "epoch": 2.260330827067669, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 42578 + }, + { + "epoch": 2.260390977443609, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 42579 + }, + { + "epoch": 2.260451127819549, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 42580 + }, + { + "epoch": 2.2605112781954886, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 42581 + }, + { + "epoch": 2.2605714285714287, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 42582 + }, + { + "epoch": 2.2606315789473683, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 42583 + }, + { + "epoch": 2.2606917293233084, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1521, + "step": 42584 + }, + { + "epoch": 2.260751879699248, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 42585 + }, + { + "epoch": 2.260812030075188, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 42586 + }, + { + "epoch": 2.260872180451128, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 42587 + }, + { + "epoch": 2.2609323308270675, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 42588 + }, + { + "epoch": 2.2609924812030076, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 42589 + }, + { + "epoch": 2.261052631578947, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 42590 + }, + { + "epoch": 2.2611127819548873, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 42591 + }, + { + "epoch": 2.261172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 42592 + }, + { + "epoch": 2.261233082706767, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 42593 + }, + { + "epoch": 2.2612932330827067, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 42594 + }, + { + "epoch": 2.261353383458647, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 42595 + }, + { + "epoch": 2.2614135338345864, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 42596 + }, + { + "epoch": 2.261473684210526, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 42597 + }, + { + "epoch": 2.261533834586466, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 42598 + }, + { + "epoch": 2.261593984962406, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 42599 + }, + { + "epoch": 2.261654135338346, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 42600 + }, + { + "epoch": 2.2617142857142856, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 42601 + }, + { + "epoch": 2.2617744360902257, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 42602 + }, + { + "epoch": 2.2618345864661653, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 42603 + }, + { + "epoch": 2.2618947368421054, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 42604 + }, + { + "epoch": 2.261954887218045, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 42605 + }, + { + "epoch": 2.262015037593985, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 42606 + }, + { + "epoch": 2.262075187969925, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 42607 + }, + { + "epoch": 2.262135338345865, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 42608 + }, + { + "epoch": 2.2621954887218045, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 42609 + }, + { + "epoch": 2.262255639097744, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 42610 + }, + { + "epoch": 2.2623157894736843, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 42611 + }, + { + "epoch": 2.262375939849624, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 42612 + }, + { + "epoch": 2.262436090225564, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9176, + "step": 42613 + }, + { + "epoch": 2.2624962406015037, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 42614 + }, + { + "epoch": 2.2625563909774438, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 42615 + }, + { + "epoch": 2.2626165413533834, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 42616 + }, + { + "epoch": 2.2626766917293235, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 42617 + }, + { + "epoch": 2.262736842105263, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 42618 + }, + { + "epoch": 2.262796992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 42619 + }, + { + "epoch": 2.262857142857143, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9329, + "step": 42620 + }, + { + "epoch": 2.2629172932330825, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 42621 + }, + { + "epoch": 2.2629774436090226, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 42622 + }, + { + "epoch": 2.2630375939849623, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 42623 + }, + { + "epoch": 2.2630977443609024, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 42624 + }, + { + "epoch": 2.263157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9222, + "step": 42625 + }, + { + "epoch": 2.263218045112782, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 42626 + }, + { + "epoch": 2.2632781954887218, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 42627 + }, + { + "epoch": 2.2633383458646614, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 42628 + }, + { + "epoch": 2.2633984962406015, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 42629 + }, + { + "epoch": 2.263458646616541, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 42630 + }, + { + "epoch": 2.2635187969924813, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 42631 + }, + { + "epoch": 2.263578947368421, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 42632 + }, + { + "epoch": 2.263639097744361, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 42633 + }, + { + "epoch": 2.2636992481203007, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 42634 + }, + { + "epoch": 2.2637593984962407, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 42635 + }, + { + "epoch": 2.2638195488721804, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 42636 + }, + { + "epoch": 2.2638796992481205, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9635, + "step": 42637 + }, + { + "epoch": 2.26393984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 42638 + }, + { + "epoch": 2.2640000000000002, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 42639 + }, + { + "epoch": 2.26406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 42640 + }, + { + "epoch": 2.2641203007518795, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 42641 + }, + { + "epoch": 2.2641804511278196, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 42642 + }, + { + "epoch": 2.2642406015037593, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 42643 + }, + { + "epoch": 2.2643007518796994, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 42644 + }, + { + "epoch": 2.264360902255639, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 42645 + }, + { + "epoch": 2.264421052631579, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.2106, + "step": 42646 + }, + { + "epoch": 2.2644812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 42647 + }, + { + "epoch": 2.264541353383459, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.954, + "step": 42648 + }, + { + "epoch": 2.2646015037593985, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 42649 + }, + { + "epoch": 2.264661654135338, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 42650 + }, + { + "epoch": 2.2647218045112782, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 42651 + }, + { + "epoch": 2.264781954887218, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 42652 + }, + { + "epoch": 2.264842105263158, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 42653 + }, + { + "epoch": 2.2649022556390976, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9122, + "step": 42654 + }, + { + "epoch": 2.2649624060150377, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 42655 + }, + { + "epoch": 2.2650225563909774, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 42656 + }, + { + "epoch": 2.2650827067669175, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9341, + "step": 42657 + }, + { + "epoch": 2.265142857142857, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 42658 + }, + { + "epoch": 2.2652030075187968, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 42659 + }, + { + "epoch": 2.265263157894737, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9354, + "step": 42660 + }, + { + "epoch": 2.2653233082706765, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 42661 + }, + { + "epoch": 2.2653834586466166, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 42662 + }, + { + "epoch": 2.2654436090225563, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 42663 + }, + { + "epoch": 2.2655037593984964, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9537, + "step": 42664 + }, + { + "epoch": 2.265563909774436, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 42665 + }, + { + "epoch": 2.265624060150376, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 42666 + }, + { + "epoch": 2.2656842105263157, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 42667 + }, + { + "epoch": 2.2657443609022554, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.918, + "step": 42668 + }, + { + "epoch": 2.2658045112781955, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 42669 + }, + { + "epoch": 2.2658646616541356, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 42670 + }, + { + "epoch": 2.2659248120300752, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 42671 + }, + { + "epoch": 2.265984962406015, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 42672 + }, + { + "epoch": 2.266045112781955, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 42673 + }, + { + "epoch": 2.2661052631578946, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 42674 + }, + { + "epoch": 2.2661654135338347, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 42675 + }, + { + "epoch": 2.2662255639097744, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 42676 + }, + { + "epoch": 2.2662857142857145, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 42677 + }, + { + "epoch": 2.266345864661654, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 42678 + }, + { + "epoch": 2.266406015037594, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 42679 + }, + { + "epoch": 2.266466165413534, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 42680 + }, + { + "epoch": 2.2665263157894735, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 42681 + }, + { + "epoch": 2.2665864661654136, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 42682 + }, + { + "epoch": 2.2666466165413532, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 42683 + }, + { + "epoch": 2.2667067669172933, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 42684 + }, + { + "epoch": 2.266766917293233, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 42685 + }, + { + "epoch": 2.266827067669173, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 42686 + }, + { + "epoch": 2.2668872180451127, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 42687 + }, + { + "epoch": 2.266947368421053, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 42688 + }, + { + "epoch": 2.2670075187969925, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 42689 + }, + { + "epoch": 2.267067669172932, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 42690 + }, + { + "epoch": 2.267127819548872, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 42691 + }, + { + "epoch": 2.267187969924812, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 42692 + }, + { + "epoch": 2.267248120300752, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1783, + "step": 42693 + }, + { + "epoch": 2.2673082706766916, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 42694 + }, + { + "epoch": 2.2673684210526317, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 42695 + }, + { + "epoch": 2.2674285714285713, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 42696 + }, + { + "epoch": 2.2674887218045114, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.196, + "step": 42697 + }, + { + "epoch": 2.267548872180451, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 42698 + }, + { + "epoch": 2.2676090225563907, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 42699 + }, + { + "epoch": 2.267669172932331, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 42700 + }, + { + "epoch": 2.2677293233082705, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 42701 + }, + { + "epoch": 2.2677894736842106, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 42702 + }, + { + "epoch": 2.2678496240601502, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 42703 + }, + { + "epoch": 2.2679097744360903, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 42704 + }, + { + "epoch": 2.26796992481203, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 42705 + }, + { + "epoch": 2.26803007518797, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 42706 + }, + { + "epoch": 2.2680902255639097, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 42707 + }, + { + "epoch": 2.26815037593985, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 42708 + }, + { + "epoch": 2.2682105263157895, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 42709 + }, + { + "epoch": 2.2682706766917295, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 42710 + }, + { + "epoch": 2.268330827067669, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2049, + "step": 42711 + }, + { + "epoch": 2.268390977443609, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 42712 + }, + { + "epoch": 2.268451127819549, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 42713 + }, + { + "epoch": 2.2685112781954886, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 42714 + }, + { + "epoch": 2.2685714285714287, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 42715 + }, + { + "epoch": 2.2686315789473683, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 42716 + }, + { + "epoch": 2.2686917293233084, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 42717 + }, + { + "epoch": 2.268751879699248, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 42718 + }, + { + "epoch": 2.268812030075188, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 42719 + }, + { + "epoch": 2.268872180451128, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 42720 + }, + { + "epoch": 2.2689323308270675, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 42721 + }, + { + "epoch": 2.2689924812030076, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 42722 + }, + { + "epoch": 2.269052631578947, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 42723 + }, + { + "epoch": 2.2691127819548873, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 42724 + }, + { + "epoch": 2.269172932330827, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 42725 + }, + { + "epoch": 2.269233082706767, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 42726 + }, + { + "epoch": 2.2692932330827067, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 42727 + }, + { + "epoch": 2.269353383458647, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 42728 + }, + { + "epoch": 2.2694135338345864, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 42729 + }, + { + "epoch": 2.269473684210526, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 42730 + }, + { + "epoch": 2.269533834586466, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 42731 + }, + { + "epoch": 2.269593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 42732 + }, + { + "epoch": 2.269654135338346, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 42733 + }, + { + "epoch": 2.2697142857142856, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 42734 + }, + { + "epoch": 2.2697744360902257, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 42735 + }, + { + "epoch": 2.2698345864661653, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 42736 + }, + { + "epoch": 2.2698947368421054, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 42737 + }, + { + "epoch": 2.269954887218045, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 42738 + }, + { + "epoch": 2.270015037593985, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 42739 + }, + { + "epoch": 2.270075187969925, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 42740 + }, + { + "epoch": 2.270135338345865, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 42741 + }, + { + "epoch": 2.2701954887218045, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 42742 + }, + { + "epoch": 2.270255639097744, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 42743 + }, + { + "epoch": 2.2703157894736843, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 42744 + }, + { + "epoch": 2.270375939849624, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 42745 + }, + { + "epoch": 2.270436090225564, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 42746 + }, + { + "epoch": 2.2704962406015037, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 42747 + }, + { + "epoch": 2.2705563909774438, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 42748 + }, + { + "epoch": 2.2706165413533834, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9346, + "step": 42749 + }, + { + "epoch": 2.2706766917293235, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 42750 + }, + { + "epoch": 2.270736842105263, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 42751 + }, + { + "epoch": 2.270796992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 42752 + }, + { + "epoch": 2.270857142857143, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 42753 + }, + { + "epoch": 2.2709172932330826, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 42754 + }, + { + "epoch": 2.2709774436090227, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 42755 + }, + { + "epoch": 2.2710375939849623, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 42756 + }, + { + "epoch": 2.2710977443609024, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 42757 + }, + { + "epoch": 2.271157894736842, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 42758 + }, + { + "epoch": 2.271218045112782, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1935, + "step": 42759 + }, + { + "epoch": 2.271278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 42760 + }, + { + "epoch": 2.2713383458646614, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 42761 + }, + { + "epoch": 2.2713984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 42762 + }, + { + "epoch": 2.271458646616541, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 42763 + }, + { + "epoch": 2.2715187969924813, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 42764 + }, + { + "epoch": 2.271578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 42765 + }, + { + "epoch": 2.271639097744361, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 42766 + }, + { + "epoch": 2.2716992481203007, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 42767 + }, + { + "epoch": 2.2717593984962408, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 42768 + }, + { + "epoch": 2.2718195488721804, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 42769 + }, + { + "epoch": 2.27187969924812, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9491, + "step": 42770 + }, + { + "epoch": 2.27193984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 42771 + }, + { + "epoch": 2.2720000000000002, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 42772 + }, + { + "epoch": 2.27206015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 42773 + }, + { + "epoch": 2.2721203007518795, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9157, + "step": 42774 + }, + { + "epoch": 2.2721804511278196, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 42775 + }, + { + "epoch": 2.2722406015037593, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 42776 + }, + { + "epoch": 2.2723007518796994, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 42777 + }, + { + "epoch": 2.272360902255639, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 42778 + }, + { + "epoch": 2.272421052631579, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 42779 + }, + { + "epoch": 2.2724812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 42780 + }, + { + "epoch": 2.272541353383459, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 42781 + }, + { + "epoch": 2.2726015037593985, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 42782 + }, + { + "epoch": 2.272661654135338, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 42783 + }, + { + "epoch": 2.2727218045112783, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 42784 + }, + { + "epoch": 2.272781954887218, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 42785 + }, + { + "epoch": 2.272842105263158, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 42786 + }, + { + "epoch": 2.2729022556390976, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 42787 + }, + { + "epoch": 2.2729624060150377, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 42788 + }, + { + "epoch": 2.2730225563909774, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 42789 + }, + { + "epoch": 2.2730827067669175, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 42790 + }, + { + "epoch": 2.273142857142857, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 42791 + }, + { + "epoch": 2.273203007518797, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 42792 + }, + { + "epoch": 2.273263157894737, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 42793 + }, + { + "epoch": 2.2733233082706765, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9189, + "step": 42794 + }, + { + "epoch": 2.2733834586466166, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 42795 + }, + { + "epoch": 2.2734436090225563, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 42796 + }, + { + "epoch": 2.2735037593984964, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 42797 + }, + { + "epoch": 2.273563909774436, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 42798 + }, + { + "epoch": 2.273624060150376, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 42799 + }, + { + "epoch": 2.2736842105263158, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 42800 + }, + { + "epoch": 2.2737443609022554, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 42801 + }, + { + "epoch": 2.2738045112781955, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 42802 + }, + { + "epoch": 2.273864661654135, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 42803 + }, + { + "epoch": 2.2739248120300752, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9164, + "step": 42804 + }, + { + "epoch": 2.273984962406015, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 42805 + }, + { + "epoch": 2.274045112781955, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 42806 + }, + { + "epoch": 2.2741052631578946, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 42807 + }, + { + "epoch": 2.2741654135338347, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 42808 + }, + { + "epoch": 2.2742255639097744, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 42809 + }, + { + "epoch": 2.2742857142857145, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 42810 + }, + { + "epoch": 2.274345864661654, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 42811 + }, + { + "epoch": 2.274406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 42812 + }, + { + "epoch": 2.274466165413534, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 42813 + }, + { + "epoch": 2.2745263157894735, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 42814 + }, + { + "epoch": 2.2745864661654136, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 42815 + }, + { + "epoch": 2.2746466165413533, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 42816 + }, + { + "epoch": 2.2747067669172933, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 42817 + }, + { + "epoch": 2.274766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 42818 + }, + { + "epoch": 2.274827067669173, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9139, + "step": 42819 + }, + { + "epoch": 2.2748872180451127, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1657, + "step": 42820 + }, + { + "epoch": 2.274947368421053, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 42821 + }, + { + "epoch": 2.2750075187969925, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 42822 + }, + { + "epoch": 2.275067669172932, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 42823 + }, + { + "epoch": 2.275127819548872, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 42824 + }, + { + "epoch": 2.275187969924812, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 42825 + }, + { + "epoch": 2.275248120300752, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 42826 + }, + { + "epoch": 2.2753082706766916, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 42827 + }, + { + "epoch": 2.2753684210526317, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 42828 + }, + { + "epoch": 2.2754285714285714, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 42829 + }, + { + "epoch": 2.2754887218045114, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 42830 + }, + { + "epoch": 2.275548872180451, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 42831 + }, + { + "epoch": 2.2756090225563907, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9407, + "step": 42832 + }, + { + "epoch": 2.275669172932331, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 42833 + }, + { + "epoch": 2.2757293233082705, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 42834 + }, + { + "epoch": 2.2757894736842106, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 42835 + }, + { + "epoch": 2.2758496240601502, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 42836 + }, + { + "epoch": 2.2759097744360903, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 42837 + }, + { + "epoch": 2.27596992481203, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9432, + "step": 42838 + }, + { + "epoch": 2.27603007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 42839 + }, + { + "epoch": 2.2760902255639097, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 42840 + }, + { + "epoch": 2.27615037593985, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 42841 + }, + { + "epoch": 2.2762105263157895, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 42842 + }, + { + "epoch": 2.2762706766917296, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 42843 + }, + { + "epoch": 2.276330827067669, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 42844 + }, + { + "epoch": 2.276390977443609, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 42845 + }, + { + "epoch": 2.276451127819549, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 42846 + }, + { + "epoch": 2.2765112781954886, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 42847 + }, + { + "epoch": 2.2765714285714287, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 42848 + }, + { + "epoch": 2.2766315789473683, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 42849 + }, + { + "epoch": 2.2766917293233084, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 42850 + }, + { + "epoch": 2.276751879699248, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 42851 + }, + { + "epoch": 2.276812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 42852 + }, + { + "epoch": 2.276872180451128, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 42853 + }, + { + "epoch": 2.2769323308270675, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 42854 + }, + { + "epoch": 2.2769924812030076, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 42855 + }, + { + "epoch": 2.277052631578947, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 42856 + }, + { + "epoch": 2.2771127819548873, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 42857 + }, + { + "epoch": 2.277172932330827, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 42858 + }, + { + "epoch": 2.277233082706767, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 42859 + }, + { + "epoch": 2.2772932330827067, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1775, + "step": 42860 + }, + { + "epoch": 2.277353383458647, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.8944, + "step": 42861 + }, + { + "epoch": 2.2774135338345864, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 42862 + }, + { + "epoch": 2.277473684210526, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 42863 + }, + { + "epoch": 2.277533834586466, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 42864 + }, + { + "epoch": 2.277593984962406, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 42865 + }, + { + "epoch": 2.277654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 42866 + }, + { + "epoch": 2.2777142857142856, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 42867 + }, + { + "epoch": 2.2777744360902257, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 42868 + }, + { + "epoch": 2.2778345864661653, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 42869 + }, + { + "epoch": 2.2778947368421054, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 42870 + }, + { + "epoch": 2.277954887218045, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.956, + "step": 42871 + }, + { + "epoch": 2.2780150375939847, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 42872 + }, + { + "epoch": 2.278075187969925, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 42873 + }, + { + "epoch": 2.278135338345865, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 42874 + }, + { + "epoch": 2.2781954887218046, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 42875 + }, + { + "epoch": 2.278255639097744, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 42876 + }, + { + "epoch": 2.2783157894736843, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 42877 + }, + { + "epoch": 2.278375939849624, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 42878 + }, + { + "epoch": 2.278436090225564, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 42879 + }, + { + "epoch": 2.2784962406015037, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 42880 + }, + { + "epoch": 2.278556390977444, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 42881 + }, + { + "epoch": 2.2786165413533834, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1324, + "step": 42882 + }, + { + "epoch": 2.2786766917293235, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 42883 + }, + { + "epoch": 2.278736842105263, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 42884 + }, + { + "epoch": 2.278796992481203, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 42885 + }, + { + "epoch": 2.278857142857143, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 42886 + }, + { + "epoch": 2.2789172932330826, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 42887 + }, + { + "epoch": 2.2789774436090227, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 42888 + }, + { + "epoch": 2.2790375939849623, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 42889 + }, + { + "epoch": 2.2790977443609024, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 42890 + }, + { + "epoch": 2.279157894736842, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9277, + "step": 42891 + }, + { + "epoch": 2.279218045112782, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1146, + "step": 42892 + }, + { + "epoch": 2.279278195488722, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 42893 + }, + { + "epoch": 2.2793383458646614, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 42894 + }, + { + "epoch": 2.2793984962406015, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 42895 + }, + { + "epoch": 2.279458646616541, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 42896 + }, + { + "epoch": 2.2795187969924813, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 42897 + }, + { + "epoch": 2.279578947368421, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 42898 + }, + { + "epoch": 2.279639097744361, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 42899 + }, + { + "epoch": 2.2796992481203007, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 42900 + }, + { + "epoch": 2.2797593984962408, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 42901 + }, + { + "epoch": 2.2798195488721804, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 42902 + }, + { + "epoch": 2.27987969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 42903 + }, + { + "epoch": 2.27993984962406, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 42904 + }, + { + "epoch": 2.2800000000000002, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 42905 + }, + { + "epoch": 2.28006015037594, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 42906 + }, + { + "epoch": 2.2801203007518795, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 42907 + }, + { + "epoch": 2.2801804511278196, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 42908 + }, + { + "epoch": 2.2802406015037593, + "grad_norm": 1.40625, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 42909 + }, + { + "epoch": 2.2803007518796994, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 42910 + }, + { + "epoch": 2.280360902255639, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 42911 + }, + { + "epoch": 2.280421052631579, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 42912 + }, + { + "epoch": 2.2804812030075188, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 42913 + }, + { + "epoch": 2.280541353383459, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 42914 + }, + { + "epoch": 2.2806015037593985, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 42915 + }, + { + "epoch": 2.280661654135338, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 42916 + }, + { + "epoch": 2.2807218045112783, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 42917 + }, + { + "epoch": 2.280781954887218, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 42918 + }, + { + "epoch": 2.280842105263158, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 42919 + }, + { + "epoch": 2.2809022556390977, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 42920 + }, + { + "epoch": 2.2809624060150377, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 42921 + }, + { + "epoch": 2.2810225563909774, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 42922 + }, + { + "epoch": 2.2810827067669175, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 42923 + }, + { + "epoch": 2.281142857142857, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 42924 + }, + { + "epoch": 2.281203007518797, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 42925 + }, + { + "epoch": 2.281263157894737, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 42926 + }, + { + "epoch": 2.2813233082706765, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1622, + "step": 42927 + }, + { + "epoch": 2.2813834586466166, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 42928 + }, + { + "epoch": 2.2814436090225563, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 42929 + }, + { + "epoch": 2.2815037593984964, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 42930 + }, + { + "epoch": 2.281563909774436, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 42931 + }, + { + "epoch": 2.281624060150376, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 42932 + }, + { + "epoch": 2.2816842105263158, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 42933 + }, + { + "epoch": 2.2817443609022554, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 42934 + }, + { + "epoch": 2.2818045112781955, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 42935 + }, + { + "epoch": 2.281864661654135, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 42936 + }, + { + "epoch": 2.2819248120300752, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 42937 + }, + { + "epoch": 2.281984962406015, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 42938 + }, + { + "epoch": 2.282045112781955, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 42939 + }, + { + "epoch": 2.2821052631578946, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 42940 + }, + { + "epoch": 2.2821654135338347, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 42941 + }, + { + "epoch": 2.2822255639097744, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 42942 + }, + { + "epoch": 2.2822857142857145, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 42943 + }, + { + "epoch": 2.282345864661654, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 42944 + }, + { + "epoch": 2.282406015037594, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 42945 + }, + { + "epoch": 2.282466165413534, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 42946 + }, + { + "epoch": 2.2825263157894735, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 42947 + }, + { + "epoch": 2.2825864661654136, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 42948 + }, + { + "epoch": 2.2826466165413533, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 42949 + }, + { + "epoch": 2.2827067669172934, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 42950 + }, + { + "epoch": 2.282766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 42951 + }, + { + "epoch": 2.282827067669173, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9442, + "step": 42952 + }, + { + "epoch": 2.2828872180451127, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 42953 + }, + { + "epoch": 2.282947368421053, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 42954 + }, + { + "epoch": 2.2830075187969925, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 42955 + }, + { + "epoch": 2.283067669172932, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 42956 + }, + { + "epoch": 2.2831278195488722, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 42957 + }, + { + "epoch": 2.283187969924812, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 42958 + }, + { + "epoch": 2.283248120300752, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 42959 + }, + { + "epoch": 2.2833082706766916, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.94, + "step": 42960 + }, + { + "epoch": 2.2833684210526317, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.199, + "step": 42961 + }, + { + "epoch": 2.2834285714285714, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 42962 + }, + { + "epoch": 2.2834887218045115, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2311, + "step": 42963 + }, + { + "epoch": 2.283548872180451, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 42964 + }, + { + "epoch": 2.2836090225563908, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9347, + "step": 42965 + }, + { + "epoch": 2.283669172932331, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 42966 + }, + { + "epoch": 2.2837293233082705, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9362, + "step": 42967 + }, + { + "epoch": 2.2837894736842106, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 42968 + }, + { + "epoch": 2.2838496240601502, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 42969 + }, + { + "epoch": 2.2839097744360903, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 42970 + }, + { + "epoch": 2.28396992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 42971 + }, + { + "epoch": 2.28403007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 42972 + }, + { + "epoch": 2.2840902255639097, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 42973 + }, + { + "epoch": 2.2841503759398494, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 42974 + }, + { + "epoch": 2.2842105263157895, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 42975 + }, + { + "epoch": 2.2842706766917296, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 42976 + }, + { + "epoch": 2.284330827067669, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 42977 + }, + { + "epoch": 2.284390977443609, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 42978 + }, + { + "epoch": 2.284451127819549, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 42979 + }, + { + "epoch": 2.2845112781954886, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 42980 + }, + { + "epoch": 2.2845714285714287, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 42981 + }, + { + "epoch": 2.2846315789473683, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 42982 + }, + { + "epoch": 2.2846917293233084, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 42983 + }, + { + "epoch": 2.284751879699248, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 42984 + }, + { + "epoch": 2.284812030075188, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 42985 + }, + { + "epoch": 2.284872180451128, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 42986 + }, + { + "epoch": 2.2849323308270675, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 42987 + }, + { + "epoch": 2.2849924812030076, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 42988 + }, + { + "epoch": 2.2850526315789472, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 42989 + }, + { + "epoch": 2.2851127819548873, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 42990 + }, + { + "epoch": 2.285172932330827, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 42991 + }, + { + "epoch": 2.285233082706767, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 42992 + }, + { + "epoch": 2.2852932330827067, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 42993 + }, + { + "epoch": 2.285353383458647, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 42994 + }, + { + "epoch": 2.2854135338345865, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 42995 + }, + { + "epoch": 2.285473684210526, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 42996 + }, + { + "epoch": 2.285533834586466, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 42997 + }, + { + "epoch": 2.285593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1167, + "step": 42998 + }, + { + "epoch": 2.285654135338346, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9312, + "step": 42999 + }, + { + "epoch": 2.2857142857142856, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 43000 + }, + { + "epoch": 2.2857744360902257, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 43001 + }, + { + "epoch": 2.2858345864661653, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 43002 + }, + { + "epoch": 2.2858947368421054, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 43003 + }, + { + "epoch": 2.285954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 43004 + }, + { + "epoch": 2.2860150375939847, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9518, + "step": 43005 + }, + { + "epoch": 2.286075187969925, + "grad_norm": 1.53125, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 43006 + }, + { + "epoch": 2.286135338345865, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 43007 + }, + { + "epoch": 2.2861954887218046, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 43008 + }, + { + "epoch": 2.286255639097744, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 43009 + }, + { + "epoch": 2.2863157894736843, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 43010 + }, + { + "epoch": 2.286375939849624, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 43011 + }, + { + "epoch": 2.286436090225564, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 43012 + }, + { + "epoch": 2.2864962406015037, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 43013 + }, + { + "epoch": 2.286556390977444, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 43014 + }, + { + "epoch": 2.2866165413533834, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 43015 + }, + { + "epoch": 2.2866766917293235, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9151, + "step": 43016 + }, + { + "epoch": 2.286736842105263, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 43017 + }, + { + "epoch": 2.286796992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 43018 + }, + { + "epoch": 2.286857142857143, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 43019 + }, + { + "epoch": 2.2869172932330826, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 43020 + }, + { + "epoch": 2.2869774436090227, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 43021 + }, + { + "epoch": 2.2870375939849623, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 43022 + }, + { + "epoch": 2.2870977443609024, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 43023 + }, + { + "epoch": 2.287157894736842, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 43024 + }, + { + "epoch": 2.287218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 43025 + }, + { + "epoch": 2.287278195488722, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 43026 + }, + { + "epoch": 2.2873383458646614, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 43027 + }, + { + "epoch": 2.2873984962406015, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 43028 + }, + { + "epoch": 2.287458646616541, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 43029 + }, + { + "epoch": 2.2875187969924813, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 43030 + }, + { + "epoch": 2.287578947368421, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 43031 + }, + { + "epoch": 2.287639097744361, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 43032 + }, + { + "epoch": 2.2876992481203007, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9247, + "step": 43033 + }, + { + "epoch": 2.2877593984962408, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 43034 + }, + { + "epoch": 2.2878195488721804, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 43035 + }, + { + "epoch": 2.28787969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 43036 + }, + { + "epoch": 2.28793984962406, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9135, + "step": 43037 + }, + { + "epoch": 2.288, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 43038 + }, + { + "epoch": 2.28806015037594, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 43039 + }, + { + "epoch": 2.2881203007518796, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 43040 + }, + { + "epoch": 2.2881804511278196, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 43041 + }, + { + "epoch": 2.2882406015037593, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 43042 + }, + { + "epoch": 2.2883007518796994, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 43043 + }, + { + "epoch": 2.288360902255639, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 43044 + }, + { + "epoch": 2.288421052631579, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 43045 + }, + { + "epoch": 2.288481203007519, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 43046 + }, + { + "epoch": 2.288541353383459, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 43047 + }, + { + "epoch": 2.2886015037593985, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 43048 + }, + { + "epoch": 2.288661654135338, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 43049 + }, + { + "epoch": 2.2887218045112783, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 43050 + }, + { + "epoch": 2.288781954887218, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 43051 + }, + { + "epoch": 2.288842105263158, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 43052 + }, + { + "epoch": 2.2889022556390977, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 43053 + }, + { + "epoch": 2.2889624060150378, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.8822, + "step": 43054 + }, + { + "epoch": 2.2890225563909774, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 43055 + }, + { + "epoch": 2.2890827067669175, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 43056 + }, + { + "epoch": 2.289142857142857, + "grad_norm": 1.5859375, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 43057 + }, + { + "epoch": 2.289203007518797, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 43058 + }, + { + "epoch": 2.289263157894737, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 43059 + }, + { + "epoch": 2.2893233082706765, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1804, + "step": 43060 + }, + { + "epoch": 2.2893834586466166, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 43061 + }, + { + "epoch": 2.2894436090225563, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 43062 + }, + { + "epoch": 2.2895037593984964, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 43063 + }, + { + "epoch": 2.289563909774436, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 43064 + }, + { + "epoch": 2.289624060150376, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 43065 + }, + { + "epoch": 2.2896842105263158, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9499, + "step": 43066 + }, + { + "epoch": 2.2897443609022554, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 43067 + }, + { + "epoch": 2.2898045112781955, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 43068 + }, + { + "epoch": 2.289864661654135, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 43069 + }, + { + "epoch": 2.2899248120300753, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 43070 + }, + { + "epoch": 2.289984962406015, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 43071 + }, + { + "epoch": 2.290045112781955, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 43072 + }, + { + "epoch": 2.2901052631578946, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 43073 + }, + { + "epoch": 2.2901654135338347, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 43074 + }, + { + "epoch": 2.2902255639097744, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 43075 + }, + { + "epoch": 2.290285714285714, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 43076 + }, + { + "epoch": 2.290345864661654, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 43077 + }, + { + "epoch": 2.2904060150375942, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 43078 + }, + { + "epoch": 2.290466165413534, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 43079 + }, + { + "epoch": 2.2905263157894735, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 43080 + }, + { + "epoch": 2.2905864661654136, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 43081 + }, + { + "epoch": 2.2906466165413533, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 43082 + }, + { + "epoch": 2.2907067669172934, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 43083 + }, + { + "epoch": 2.290766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 43084 + }, + { + "epoch": 2.290827067669173, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 43085 + }, + { + "epoch": 2.2908872180451127, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 43086 + }, + { + "epoch": 2.290947368421053, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 43087 + }, + { + "epoch": 2.2910075187969925, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 43088 + }, + { + "epoch": 2.291067669172932, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 43089 + }, + { + "epoch": 2.2911278195488722, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 43090 + }, + { + "epoch": 2.291187969924812, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 43091 + }, + { + "epoch": 2.291248120300752, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 43092 + }, + { + "epoch": 2.2913082706766916, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 43093 + }, + { + "epoch": 2.2913684210526317, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 43094 + }, + { + "epoch": 2.2914285714285714, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 43095 + }, + { + "epoch": 2.2914887218045115, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 43096 + }, + { + "epoch": 2.291548872180451, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 43097 + }, + { + "epoch": 2.2916090225563908, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 43098 + }, + { + "epoch": 2.291669172932331, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 43099 + }, + { + "epoch": 2.2917293233082705, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 43100 + }, + { + "epoch": 2.2917894736842106, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 43101 + }, + { + "epoch": 2.2918496240601502, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 43102 + }, + { + "epoch": 2.2919097744360903, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 43103 + }, + { + "epoch": 2.29196992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 43104 + }, + { + "epoch": 2.29203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 43105 + }, + { + "epoch": 2.2920902255639097, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 43106 + }, + { + "epoch": 2.2921503759398494, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9175, + "step": 43107 + }, + { + "epoch": 2.2922105263157895, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 43108 + }, + { + "epoch": 2.2922706766917296, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 43109 + }, + { + "epoch": 2.292330827067669, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 43110 + }, + { + "epoch": 2.292390977443609, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 43111 + }, + { + "epoch": 2.292451127819549, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 43112 + }, + { + "epoch": 2.2925112781954886, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 43113 + }, + { + "epoch": 2.2925714285714287, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 43114 + }, + { + "epoch": 2.2926315789473684, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 43115 + }, + { + "epoch": 2.2926917293233084, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 43116 + }, + { + "epoch": 2.292751879699248, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 43117 + }, + { + "epoch": 2.292812030075188, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 43118 + }, + { + "epoch": 2.292872180451128, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 43119 + }, + { + "epoch": 2.2929323308270675, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 43120 + }, + { + "epoch": 2.2929924812030076, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 43121 + }, + { + "epoch": 2.2930526315789472, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 43122 + }, + { + "epoch": 2.2931127819548873, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 43123 + }, + { + "epoch": 2.293172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1568, + "step": 43124 + }, + { + "epoch": 2.293233082706767, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 43125 + }, + { + "epoch": 2.2932932330827067, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 43126 + }, + { + "epoch": 2.293353383458647, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 43127 + }, + { + "epoch": 2.2934135338345865, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 43128 + }, + { + "epoch": 2.293473684210526, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 43129 + }, + { + "epoch": 2.293533834586466, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 43130 + }, + { + "epoch": 2.293593984962406, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 43131 + }, + { + "epoch": 2.293654135338346, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 43132 + }, + { + "epoch": 2.2937142857142856, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 43133 + }, + { + "epoch": 2.2937744360902257, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 43134 + }, + { + "epoch": 2.2938345864661653, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 43135 + }, + { + "epoch": 2.2938947368421054, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 43136 + }, + { + "epoch": 2.293954887218045, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9412, + "step": 43137 + }, + { + "epoch": 2.2940150375939847, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 43138 + }, + { + "epoch": 2.294075187969925, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 43139 + }, + { + "epoch": 2.2941353383458645, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 43140 + }, + { + "epoch": 2.2941954887218046, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 43141 + }, + { + "epoch": 2.294255639097744, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1354, + "step": 43142 + }, + { + "epoch": 2.2943157894736843, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 43143 + }, + { + "epoch": 2.294375939849624, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 43144 + }, + { + "epoch": 2.294436090225564, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 43145 + }, + { + "epoch": 2.2944962406015037, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 43146 + }, + { + "epoch": 2.294556390977444, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 43147 + }, + { + "epoch": 2.2946165413533834, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2523, + "step": 43148 + }, + { + "epoch": 2.2946766917293235, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 43149 + }, + { + "epoch": 2.294736842105263, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.931, + "step": 43150 + }, + { + "epoch": 2.294796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.939, + "step": 43151 + }, + { + "epoch": 2.294857142857143, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 43152 + }, + { + "epoch": 2.2949172932330826, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 43153 + }, + { + "epoch": 2.2949774436090227, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 43154 + }, + { + "epoch": 2.2950375939849623, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 43155 + }, + { + "epoch": 2.2950977443609024, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 43156 + }, + { + "epoch": 2.295157894736842, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 43157 + }, + { + "epoch": 2.295218045112782, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 43158 + }, + { + "epoch": 2.295278195488722, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 43159 + }, + { + "epoch": 2.2953383458646615, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 43160 + }, + { + "epoch": 2.2953984962406015, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 43161 + }, + { + "epoch": 2.295458646616541, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 43162 + }, + { + "epoch": 2.2955187969924813, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 43163 + }, + { + "epoch": 2.295578947368421, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 43164 + }, + { + "epoch": 2.295639097744361, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 43165 + }, + { + "epoch": 2.2956992481203007, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 43166 + }, + { + "epoch": 2.2957593984962408, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 43167 + }, + { + "epoch": 2.2958195488721804, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 43168 + }, + { + "epoch": 2.29587969924812, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 43169 + }, + { + "epoch": 2.29593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 43170 + }, + { + "epoch": 2.296, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 43171 + }, + { + "epoch": 2.29606015037594, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 43172 + }, + { + "epoch": 2.2961203007518796, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 43173 + }, + { + "epoch": 2.2961804511278197, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1571, + "step": 43174 + }, + { + "epoch": 2.2962406015037593, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 43175 + }, + { + "epoch": 2.2963007518796994, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 43176 + }, + { + "epoch": 2.296360902255639, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9481, + "step": 43177 + }, + { + "epoch": 2.296421052631579, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 43178 + }, + { + "epoch": 2.296481203007519, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 43179 + }, + { + "epoch": 2.296541353383459, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 43180 + }, + { + "epoch": 2.2966015037593985, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9372, + "step": 43181 + }, + { + "epoch": 2.296661654135338, + "grad_norm": 1.4140625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 43182 + }, + { + "epoch": 2.2967218045112783, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 43183 + }, + { + "epoch": 2.296781954887218, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 43184 + }, + { + "epoch": 2.296842105263158, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 43185 + }, + { + "epoch": 2.2969022556390977, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 43186 + }, + { + "epoch": 2.2969624060150378, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 43187 + }, + { + "epoch": 2.2970225563909774, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 43188 + }, + { + "epoch": 2.2970827067669175, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 43189 + }, + { + "epoch": 2.297142857142857, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 43190 + }, + { + "epoch": 2.297203007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 43191 + }, + { + "epoch": 2.297263157894737, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.8708, + "step": 43192 + }, + { + "epoch": 2.2973233082706765, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 43193 + }, + { + "epoch": 2.2973834586466166, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 43194 + }, + { + "epoch": 2.2974436090225563, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 43195 + }, + { + "epoch": 2.2975037593984964, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 43196 + }, + { + "epoch": 2.297563909774436, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 43197 + }, + { + "epoch": 2.297624060150376, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 43198 + }, + { + "epoch": 2.2976842105263158, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 43199 + }, + { + "epoch": 2.2977443609022554, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 43200 + }, + { + "epoch": 2.2978045112781955, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 43201 + }, + { + "epoch": 2.297864661654135, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 43202 + }, + { + "epoch": 2.2979248120300753, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 43203 + }, + { + "epoch": 2.297984962406015, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 43204 + }, + { + "epoch": 2.298045112781955, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 43205 + }, + { + "epoch": 2.2981052631578947, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 43206 + }, + { + "epoch": 2.2981654135338347, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 43207 + }, + { + "epoch": 2.2982255639097744, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 43208 + }, + { + "epoch": 2.298285714285714, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 43209 + }, + { + "epoch": 2.298345864661654, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 43210 + }, + { + "epoch": 2.2984060150375942, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 43211 + }, + { + "epoch": 2.298466165413534, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 43212 + }, + { + "epoch": 2.2985263157894735, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 43213 + }, + { + "epoch": 2.2985864661654136, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 43214 + }, + { + "epoch": 2.2986466165413533, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9499, + "step": 43215 + }, + { + "epoch": 2.2987067669172934, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 43216 + }, + { + "epoch": 2.298766917293233, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 43217 + }, + { + "epoch": 2.298827067669173, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 43218 + }, + { + "epoch": 2.2988872180451128, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 43219 + }, + { + "epoch": 2.298947368421053, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 43220 + }, + { + "epoch": 2.2990075187969925, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 43221 + }, + { + "epoch": 2.299067669172932, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 43222 + }, + { + "epoch": 2.2991278195488722, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 43223 + }, + { + "epoch": 2.299187969924812, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 43224 + }, + { + "epoch": 2.299248120300752, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 43225 + }, + { + "epoch": 2.2993082706766916, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 43226 + }, + { + "epoch": 2.2993684210526317, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 43227 + }, + { + "epoch": 2.2994285714285714, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 43228 + }, + { + "epoch": 2.2994887218045115, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 43229 + }, + { + "epoch": 2.299548872180451, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 43230 + }, + { + "epoch": 2.2996090225563908, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 43231 + }, + { + "epoch": 2.299669172932331, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 43232 + }, + { + "epoch": 2.2997293233082705, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 43233 + }, + { + "epoch": 2.2997894736842106, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 43234 + }, + { + "epoch": 2.2998496240601503, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 43235 + }, + { + "epoch": 2.2999097744360903, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 43236 + }, + { + "epoch": 2.29996992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 43237 + }, + { + "epoch": 2.30003007518797, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 43238 + }, + { + "epoch": 2.3000902255639097, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 43239 + }, + { + "epoch": 2.3001503759398494, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 43240 + }, + { + "epoch": 2.3002105263157895, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 43241 + }, + { + "epoch": 2.300270676691729, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 43242 + }, + { + "epoch": 2.3003308270676692, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 43243 + }, + { + "epoch": 2.300390977443609, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 43244 + }, + { + "epoch": 2.300451127819549, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 43245 + }, + { + "epoch": 2.3005112781954886, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 43246 + }, + { + "epoch": 2.3005714285714287, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 43247 + }, + { + "epoch": 2.3006315789473684, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 43248 + }, + { + "epoch": 2.3006917293233085, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 43249 + }, + { + "epoch": 2.300751879699248, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 43250 + }, + { + "epoch": 2.300812030075188, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 43251 + }, + { + "epoch": 2.300872180451128, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 43252 + }, + { + "epoch": 2.3009323308270675, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 43253 + }, + { + "epoch": 2.3009924812030076, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 43254 + }, + { + "epoch": 2.3010526315789472, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 43255 + }, + { + "epoch": 2.3011127819548873, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 43256 + }, + { + "epoch": 2.301172932330827, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 43257 + }, + { + "epoch": 2.301233082706767, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 43258 + }, + { + "epoch": 2.3012932330827067, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 43259 + }, + { + "epoch": 2.301353383458647, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 43260 + }, + { + "epoch": 2.3014135338345865, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 43261 + }, + { + "epoch": 2.301473684210526, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 43262 + }, + { + "epoch": 2.301533834586466, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 43263 + }, + { + "epoch": 2.301593984962406, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 43264 + }, + { + "epoch": 2.301654135338346, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 43265 + }, + { + "epoch": 2.3017142857142856, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 43266 + }, + { + "epoch": 2.3017744360902257, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 43267 + }, + { + "epoch": 2.3018345864661653, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9349, + "step": 43268 + }, + { + "epoch": 2.3018947368421054, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 43269 + }, + { + "epoch": 2.301954887218045, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 43270 + }, + { + "epoch": 2.3020150375939847, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 43271 + }, + { + "epoch": 2.302075187969925, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 43272 + }, + { + "epoch": 2.3021353383458645, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 43273 + }, + { + "epoch": 2.3021954887218046, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 43274 + }, + { + "epoch": 2.302255639097744, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 43275 + }, + { + "epoch": 2.3023157894736843, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9469, + "step": 43276 + }, + { + "epoch": 2.302375939849624, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 43277 + }, + { + "epoch": 2.302436090225564, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9091, + "step": 43278 + }, + { + "epoch": 2.3024962406015037, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 43279 + }, + { + "epoch": 2.302556390977444, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 43280 + }, + { + "epoch": 2.3026165413533835, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 43281 + }, + { + "epoch": 2.3026766917293235, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 43282 + }, + { + "epoch": 2.302736842105263, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 43283 + }, + { + "epoch": 2.302796992481203, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 43284 + }, + { + "epoch": 2.302857142857143, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 43285 + }, + { + "epoch": 2.3029172932330826, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 43286 + }, + { + "epoch": 2.3029774436090227, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 43287 + }, + { + "epoch": 2.3030375939849623, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 43288 + }, + { + "epoch": 2.3030977443609024, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 43289 + }, + { + "epoch": 2.303157894736842, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 43290 + }, + { + "epoch": 2.303218045112782, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 43291 + }, + { + "epoch": 2.303278195488722, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1955, + "step": 43292 + }, + { + "epoch": 2.3033383458646615, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 43293 + }, + { + "epoch": 2.3033984962406016, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 43294 + }, + { + "epoch": 2.303458646616541, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 43295 + }, + { + "epoch": 2.3035187969924813, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 43296 + }, + { + "epoch": 2.303578947368421, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 43297 + }, + { + "epoch": 2.303639097744361, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 43298 + }, + { + "epoch": 2.3036992481203007, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 43299 + }, + { + "epoch": 2.303759398496241, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.2066, + "step": 43300 + }, + { + "epoch": 2.3038195488721804, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 43301 + }, + { + "epoch": 2.30387969924812, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 43302 + }, + { + "epoch": 2.30393984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 43303 + }, + { + "epoch": 2.304, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 43304 + }, + { + "epoch": 2.30406015037594, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 43305 + }, + { + "epoch": 2.3041203007518796, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 43306 + }, + { + "epoch": 2.3041804511278197, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 43307 + }, + { + "epoch": 2.3042406015037593, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 43308 + }, + { + "epoch": 2.3043007518796994, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 43309 + }, + { + "epoch": 2.304360902255639, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 43310 + }, + { + "epoch": 2.3044210526315787, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 43311 + }, + { + "epoch": 2.304481203007519, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.939, + "step": 43312 + }, + { + "epoch": 2.304541353383459, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 43313 + }, + { + "epoch": 2.3046015037593985, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 43314 + }, + { + "epoch": 2.304661654135338, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 43315 + }, + { + "epoch": 2.3047218045112783, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 43316 + }, + { + "epoch": 2.304781954887218, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 43317 + }, + { + "epoch": 2.304842105263158, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 43318 + }, + { + "epoch": 2.3049022556390977, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 43319 + }, + { + "epoch": 2.3049624060150378, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 43320 + }, + { + "epoch": 2.3050225563909774, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 43321 + }, + { + "epoch": 2.3050827067669175, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 43322 + }, + { + "epoch": 2.305142857142857, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 43323 + }, + { + "epoch": 2.305203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 43324 + }, + { + "epoch": 2.305263157894737, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 43325 + }, + { + "epoch": 2.3053233082706766, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 43326 + }, + { + "epoch": 2.3053834586466166, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 43327 + }, + { + "epoch": 2.3054436090225563, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 43328 + }, + { + "epoch": 2.3055037593984964, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 43329 + }, + { + "epoch": 2.305563909774436, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 43330 + }, + { + "epoch": 2.305624060150376, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9297, + "step": 43331 + }, + { + "epoch": 2.305684210526316, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 43332 + }, + { + "epoch": 2.3057443609022554, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 43333 + }, + { + "epoch": 2.3058045112781955, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 43334 + }, + { + "epoch": 2.305864661654135, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 43335 + }, + { + "epoch": 2.3059248120300753, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 43336 + }, + { + "epoch": 2.305984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 43337 + }, + { + "epoch": 2.306045112781955, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 43338 + }, + { + "epoch": 2.3061052631578947, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 43339 + }, + { + "epoch": 2.3061654135338348, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 43340 + }, + { + "epoch": 2.3062255639097744, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 43341 + }, + { + "epoch": 2.306285714285714, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 43342 + }, + { + "epoch": 2.306345864661654, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 43343 + }, + { + "epoch": 2.306406015037594, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 43344 + }, + { + "epoch": 2.306466165413534, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 43345 + }, + { + "epoch": 2.3065263157894735, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 43346 + }, + { + "epoch": 2.3065864661654136, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 43347 + }, + { + "epoch": 2.3066466165413533, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 43348 + }, + { + "epoch": 2.3067067669172934, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 43349 + }, + { + "epoch": 2.306766917293233, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 43350 + }, + { + "epoch": 2.306827067669173, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 43351 + }, + { + "epoch": 2.3068872180451128, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 43352 + }, + { + "epoch": 2.306947368421053, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 43353 + }, + { + "epoch": 2.3070075187969925, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.3206, + "step": 43354 + }, + { + "epoch": 2.307067669172932, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.3637, + "step": 43355 + }, + { + "epoch": 2.3071278195488722, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 43356 + }, + { + "epoch": 2.307187969924812, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 43357 + }, + { + "epoch": 2.307248120300752, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 43358 + }, + { + "epoch": 2.3073082706766916, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 43359 + }, + { + "epoch": 2.3073684210526317, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 43360 + }, + { + "epoch": 2.3074285714285714, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 43361 + }, + { + "epoch": 2.3074887218045115, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 43362 + }, + { + "epoch": 2.307548872180451, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 43363 + }, + { + "epoch": 2.3076090225563908, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 43364 + }, + { + "epoch": 2.307669172932331, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 43365 + }, + { + "epoch": 2.3077293233082705, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 43366 + }, + { + "epoch": 2.3077894736842106, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 43367 + }, + { + "epoch": 2.3078496240601503, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 43368 + }, + { + "epoch": 2.3079097744360904, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 43369 + }, + { + "epoch": 2.30796992481203, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 43370 + }, + { + "epoch": 2.30803007518797, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 43371 + }, + { + "epoch": 2.3080902255639097, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 43372 + }, + { + "epoch": 2.3081503759398494, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 43373 + }, + { + "epoch": 2.3082105263157895, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 43374 + }, + { + "epoch": 2.308270676691729, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 43375 + }, + { + "epoch": 2.3083308270676692, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 43376 + }, + { + "epoch": 2.308390977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 43377 + }, + { + "epoch": 2.308451127819549, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 43378 + }, + { + "epoch": 2.3085112781954886, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 43379 + }, + { + "epoch": 2.3085714285714287, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 43380 + }, + { + "epoch": 2.3086315789473684, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 43381 + }, + { + "epoch": 2.3086917293233085, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 43382 + }, + { + "epoch": 2.308751879699248, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 43383 + }, + { + "epoch": 2.308812030075188, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 43384 + }, + { + "epoch": 2.308872180451128, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 43385 + }, + { + "epoch": 2.3089323308270675, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 43386 + }, + { + "epoch": 2.3089924812030076, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 43387 + }, + { + "epoch": 2.3090526315789472, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 43388 + }, + { + "epoch": 2.3091127819548873, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 43389 + }, + { + "epoch": 2.309172932330827, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1843, + "step": 43390 + }, + { + "epoch": 2.309233082706767, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2087, + "step": 43391 + }, + { + "epoch": 2.3092932330827067, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 43392 + }, + { + "epoch": 2.309353383458647, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 43393 + }, + { + "epoch": 2.3094135338345865, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 43394 + }, + { + "epoch": 2.309473684210526, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 43395 + }, + { + "epoch": 2.309533834586466, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 43396 + }, + { + "epoch": 2.309593984962406, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 43397 + }, + { + "epoch": 2.309654135338346, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 43398 + }, + { + "epoch": 2.3097142857142856, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 43399 + }, + { + "epoch": 2.3097744360902257, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 43400 + }, + { + "epoch": 2.3098345864661654, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 43401 + }, + { + "epoch": 2.3098947368421054, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 43402 + }, + { + "epoch": 2.309954887218045, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 43403 + }, + { + "epoch": 2.3100150375939847, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 43404 + }, + { + "epoch": 2.310075187969925, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 43405 + }, + { + "epoch": 2.3101353383458645, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 43406 + }, + { + "epoch": 2.3101954887218046, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 43407 + }, + { + "epoch": 2.3102556390977442, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9451, + "step": 43408 + }, + { + "epoch": 2.3103157894736843, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 43409 + }, + { + "epoch": 2.310375939849624, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 43410 + }, + { + "epoch": 2.310436090225564, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 43411 + }, + { + "epoch": 2.3104962406015037, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 43412 + }, + { + "epoch": 2.3105563909774434, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 43413 + }, + { + "epoch": 2.3106165413533835, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 43414 + }, + { + "epoch": 2.3106766917293236, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 43415 + }, + { + "epoch": 2.310736842105263, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 43416 + }, + { + "epoch": 2.310796992481203, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 43417 + }, + { + "epoch": 2.310857142857143, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 43418 + }, + { + "epoch": 2.3109172932330826, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.8838, + "step": 43419 + }, + { + "epoch": 2.3109774436090227, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 43420 + }, + { + "epoch": 2.3110375939849623, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 43421 + }, + { + "epoch": 2.3110977443609024, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 43422 + }, + { + "epoch": 2.311157894736842, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 43423 + }, + { + "epoch": 2.311218045112782, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9101, + "step": 43424 + }, + { + "epoch": 2.311278195488722, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 43425 + }, + { + "epoch": 2.3113383458646615, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 43426 + }, + { + "epoch": 2.3113984962406016, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 43427 + }, + { + "epoch": 2.311458646616541, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 43428 + }, + { + "epoch": 2.3115187969924813, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 43429 + }, + { + "epoch": 2.311578947368421, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 43430 + }, + { + "epoch": 2.311639097744361, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 43431 + }, + { + "epoch": 2.3116992481203007, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 43432 + }, + { + "epoch": 2.311759398496241, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 43433 + }, + { + "epoch": 2.3118195488721804, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 43434 + }, + { + "epoch": 2.31187969924812, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 43435 + }, + { + "epoch": 2.31193984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 43436 + }, + { + "epoch": 2.312, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 43437 + }, + { + "epoch": 2.31206015037594, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 43438 + }, + { + "epoch": 2.3121203007518796, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 43439 + }, + { + "epoch": 2.3121804511278197, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 43440 + }, + { + "epoch": 2.3122406015037593, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9647, + "step": 43441 + }, + { + "epoch": 2.3123007518796994, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 43442 + }, + { + "epoch": 2.312360902255639, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 43443 + }, + { + "epoch": 2.3124210526315787, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 43444 + }, + { + "epoch": 2.312481203007519, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 43445 + }, + { + "epoch": 2.3125413533834585, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 43446 + }, + { + "epoch": 2.3126015037593985, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 43447 + }, + { + "epoch": 2.312661654135338, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 43448 + }, + { + "epoch": 2.3127218045112783, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 43449 + }, + { + "epoch": 2.312781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 43450 + }, + { + "epoch": 2.312842105263158, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 43451 + }, + { + "epoch": 2.3129022556390977, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 43452 + }, + { + "epoch": 2.3129624060150378, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 43453 + }, + { + "epoch": 2.3130225563909774, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 43454 + }, + { + "epoch": 2.3130827067669175, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 43455 + }, + { + "epoch": 2.313142857142857, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 43456 + }, + { + "epoch": 2.313203007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 43457 + }, + { + "epoch": 2.313263157894737, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 43458 + }, + { + "epoch": 2.3133233082706766, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 43459 + }, + { + "epoch": 2.3133834586466167, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 43460 + }, + { + "epoch": 2.3134436090225563, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 43461 + }, + { + "epoch": 2.3135037593984964, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 43462 + }, + { + "epoch": 2.313563909774436, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 43463 + }, + { + "epoch": 2.313624060150376, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.1876, + "step": 43464 + }, + { + "epoch": 2.313684210526316, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 43465 + }, + { + "epoch": 2.3137443609022554, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 43466 + }, + { + "epoch": 2.3138045112781955, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 43467 + }, + { + "epoch": 2.313864661654135, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 43468 + }, + { + "epoch": 2.3139248120300753, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 43469 + }, + { + "epoch": 2.313984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 43470 + }, + { + "epoch": 2.314045112781955, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 43471 + }, + { + "epoch": 2.3141052631578947, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 43472 + }, + { + "epoch": 2.3141654135338348, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 43473 + }, + { + "epoch": 2.3142255639097744, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9252, + "step": 43474 + }, + { + "epoch": 2.314285714285714, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 43475 + }, + { + "epoch": 2.314345864661654, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 43476 + }, + { + "epoch": 2.314406015037594, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 43477 + }, + { + "epoch": 2.314466165413534, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 43478 + }, + { + "epoch": 2.3145263157894735, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 43479 + }, + { + "epoch": 2.3145864661654136, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 43480 + }, + { + "epoch": 2.3146466165413533, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 43481 + }, + { + "epoch": 2.3147067669172934, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 43482 + }, + { + "epoch": 2.314766917293233, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 43483 + }, + { + "epoch": 2.314827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 43484 + }, + { + "epoch": 2.3148872180451128, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 43485 + }, + { + "epoch": 2.314947368421053, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 43486 + }, + { + "epoch": 2.3150075187969925, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 43487 + }, + { + "epoch": 2.315067669172932, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 43488 + }, + { + "epoch": 2.3151278195488723, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 43489 + }, + { + "epoch": 2.315187969924812, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 43490 + }, + { + "epoch": 2.315248120300752, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 43491 + }, + { + "epoch": 2.3153082706766916, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 43492 + }, + { + "epoch": 2.3153684210526317, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 43493 + }, + { + "epoch": 2.3154285714285714, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 43494 + }, + { + "epoch": 2.3154887218045115, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 43495 + }, + { + "epoch": 2.315548872180451, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 43496 + }, + { + "epoch": 2.315609022556391, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 43497 + }, + { + "epoch": 2.315669172932331, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 43498 + }, + { + "epoch": 2.3157293233082705, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 43499 + }, + { + "epoch": 2.3157894736842106, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 43500 + }, + { + "epoch": 2.3158496240601503, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 43501 + }, + { + "epoch": 2.3159097744360904, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 43502 + }, + { + "epoch": 2.31596992481203, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 43503 + }, + { + "epoch": 2.31603007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 43504 + }, + { + "epoch": 2.3160902255639098, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 43505 + }, + { + "epoch": 2.3161503759398494, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 43506 + }, + { + "epoch": 2.3162105263157895, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 43507 + }, + { + "epoch": 2.316270676691729, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 43508 + }, + { + "epoch": 2.3163308270676692, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 43509 + }, + { + "epoch": 2.316390977443609, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 43510 + }, + { + "epoch": 2.316451127819549, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 43511 + }, + { + "epoch": 2.3165112781954886, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 43512 + }, + { + "epoch": 2.3165714285714287, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.935, + "step": 43513 + }, + { + "epoch": 2.3166315789473684, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 43514 + }, + { + "epoch": 2.316691729323308, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 43515 + }, + { + "epoch": 2.316751879699248, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 43516 + }, + { + "epoch": 2.316812030075188, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 43517 + }, + { + "epoch": 2.316872180451128, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 43518 + }, + { + "epoch": 2.3169323308270675, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 43519 + }, + { + "epoch": 2.3169924812030076, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 43520 + }, + { + "epoch": 2.3170526315789473, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 43521 + }, + { + "epoch": 2.3171127819548873, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 43522 + }, + { + "epoch": 2.317172932330827, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 43523 + }, + { + "epoch": 2.317233082706767, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 43524 + }, + { + "epoch": 2.3172932330827067, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 43525 + }, + { + "epoch": 2.317353383458647, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 43526 + }, + { + "epoch": 2.3174135338345865, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 43527 + }, + { + "epoch": 2.317473684210526, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 43528 + }, + { + "epoch": 2.3175338345864662, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 43529 + }, + { + "epoch": 2.317593984962406, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 43530 + }, + { + "epoch": 2.317654135338346, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 43531 + }, + { + "epoch": 2.3177142857142856, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.8929, + "step": 43532 + }, + { + "epoch": 2.3177744360902257, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 43533 + }, + { + "epoch": 2.3178345864661654, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 43534 + }, + { + "epoch": 2.3178947368421055, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 43535 + }, + { + "epoch": 2.317954887218045, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 43536 + }, + { + "epoch": 2.3180150375939848, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 43537 + }, + { + "epoch": 2.318075187969925, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 43538 + }, + { + "epoch": 2.3181353383458645, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 43539 + }, + { + "epoch": 2.3181954887218046, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 43540 + }, + { + "epoch": 2.3182556390977442, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 43541 + }, + { + "epoch": 2.3183157894736843, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 43542 + }, + { + "epoch": 2.318375939849624, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 43543 + }, + { + "epoch": 2.318436090225564, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 43544 + }, + { + "epoch": 2.3184962406015037, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 43545 + }, + { + "epoch": 2.3185563909774434, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 43546 + }, + { + "epoch": 2.3186165413533835, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9244, + "step": 43547 + }, + { + "epoch": 2.3186766917293236, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 43548 + }, + { + "epoch": 2.318736842105263, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 43549 + }, + { + "epoch": 2.318796992481203, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 43550 + }, + { + "epoch": 2.318857142857143, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 43551 + }, + { + "epoch": 2.3189172932330826, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 43552 + }, + { + "epoch": 2.3189774436090227, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 43553 + }, + { + "epoch": 2.3190375939849623, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9307, + "step": 43554 + }, + { + "epoch": 2.3190977443609024, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 43555 + }, + { + "epoch": 2.319157894736842, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 43556 + }, + { + "epoch": 2.319218045112782, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 43557 + }, + { + "epoch": 2.319278195488722, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 43558 + }, + { + "epoch": 2.3193383458646615, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 43559 + }, + { + "epoch": 2.3193984962406016, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 43560 + }, + { + "epoch": 2.319458646616541, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 43561 + }, + { + "epoch": 2.3195187969924813, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 43562 + }, + { + "epoch": 2.319578947368421, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 43563 + }, + { + "epoch": 2.319639097744361, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 43564 + }, + { + "epoch": 2.3196992481203007, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 43565 + }, + { + "epoch": 2.319759398496241, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 43566 + }, + { + "epoch": 2.3198195488721804, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 43567 + }, + { + "epoch": 2.31987969924812, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9393, + "step": 43568 + }, + { + "epoch": 2.31993984962406, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 43569 + }, + { + "epoch": 2.32, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 43570 + }, + { + "epoch": 2.32006015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 43571 + }, + { + "epoch": 2.3201203007518796, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 43572 + }, + { + "epoch": 2.3201804511278197, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 43573 + }, + { + "epoch": 2.3202406015037593, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 43574 + }, + { + "epoch": 2.3203007518796994, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 43575 + }, + { + "epoch": 2.320360902255639, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 43576 + }, + { + "epoch": 2.3204210526315787, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 43577 + }, + { + "epoch": 2.320481203007519, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 43578 + }, + { + "epoch": 2.3205413533834585, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 43579 + }, + { + "epoch": 2.3206015037593986, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 43580 + }, + { + "epoch": 2.320661654135338, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 43581 + }, + { + "epoch": 2.3207218045112783, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 43582 + }, + { + "epoch": 2.320781954887218, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 43583 + }, + { + "epoch": 2.320842105263158, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 43584 + }, + { + "epoch": 2.3209022556390977, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 43585 + }, + { + "epoch": 2.320962406015038, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 43586 + }, + { + "epoch": 2.3210225563909774, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 43587 + }, + { + "epoch": 2.3210827067669175, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 43588 + }, + { + "epoch": 2.321142857142857, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 43589 + }, + { + "epoch": 2.321203007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 43590 + }, + { + "epoch": 2.321263157894737, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 43591 + }, + { + "epoch": 2.3213233082706766, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 43592 + }, + { + "epoch": 2.3213834586466167, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1333, + "step": 43593 + }, + { + "epoch": 2.3214436090225563, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2536, + "step": 43594 + }, + { + "epoch": 2.3215037593984964, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 43595 + }, + { + "epoch": 2.321563909774436, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 43596 + }, + { + "epoch": 2.321624060150376, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 43597 + }, + { + "epoch": 2.321684210526316, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 43598 + }, + { + "epoch": 2.3217443609022554, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 43599 + }, + { + "epoch": 2.3218045112781955, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 43600 + }, + { + "epoch": 2.321864661654135, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 43601 + }, + { + "epoch": 2.3219248120300753, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 43602 + }, + { + "epoch": 2.321984962406015, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 43603 + }, + { + "epoch": 2.322045112781955, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 43604 + }, + { + "epoch": 2.3221052631578947, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 43605 + }, + { + "epoch": 2.3221654135338348, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 43606 + }, + { + "epoch": 2.3222255639097744, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 43607 + }, + { + "epoch": 2.322285714285714, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9205, + "step": 43608 + }, + { + "epoch": 2.322345864661654, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 43609 + }, + { + "epoch": 2.322406015037594, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 43610 + }, + { + "epoch": 2.322466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 43611 + }, + { + "epoch": 2.3225263157894735, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 43612 + }, + { + "epoch": 2.3225864661654136, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 43613 + }, + { + "epoch": 2.3226466165413533, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 43614 + }, + { + "epoch": 2.3227067669172934, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 43615 + }, + { + "epoch": 2.322766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 43616 + }, + { + "epoch": 2.3228270676691727, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1778, + "step": 43617 + }, + { + "epoch": 2.322887218045113, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 43618 + }, + { + "epoch": 2.322947368421053, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 43619 + }, + { + "epoch": 2.3230075187969925, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 43620 + }, + { + "epoch": 2.323067669172932, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.936, + "step": 43621 + }, + { + "epoch": 2.3231278195488723, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 43622 + }, + { + "epoch": 2.323187969924812, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 43623 + }, + { + "epoch": 2.323248120300752, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 43624 + }, + { + "epoch": 2.3233082706766917, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 43625 + }, + { + "epoch": 2.3233684210526317, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1834, + "step": 43626 + }, + { + "epoch": 2.3234285714285714, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 43627 + }, + { + "epoch": 2.3234887218045115, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 43628 + }, + { + "epoch": 2.323548872180451, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 43629 + }, + { + "epoch": 2.323609022556391, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 43630 + }, + { + "epoch": 2.323669172932331, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 43631 + }, + { + "epoch": 2.3237293233082705, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 43632 + }, + { + "epoch": 2.3237894736842106, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 43633 + }, + { + "epoch": 2.3238496240601503, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9125, + "step": 43634 + }, + { + "epoch": 2.3239097744360904, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 43635 + }, + { + "epoch": 2.32396992481203, + "grad_norm": 1.5, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 43636 + }, + { + "epoch": 2.32403007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 43637 + }, + { + "epoch": 2.3240902255639098, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 43638 + }, + { + "epoch": 2.3241503759398494, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 43639 + }, + { + "epoch": 2.3242105263157895, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 43640 + }, + { + "epoch": 2.324270676691729, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 43641 + }, + { + "epoch": 2.3243308270676692, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 43642 + }, + { + "epoch": 2.324390977443609, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 43643 + }, + { + "epoch": 2.324451127819549, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 43644 + }, + { + "epoch": 2.3245112781954886, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 43645 + }, + { + "epoch": 2.3245714285714287, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 43646 + }, + { + "epoch": 2.3246315789473684, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 43647 + }, + { + "epoch": 2.324691729323308, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 43648 + }, + { + "epoch": 2.324751879699248, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 43649 + }, + { + "epoch": 2.324812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 43650 + }, + { + "epoch": 2.324872180451128, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 43651 + }, + { + "epoch": 2.3249323308270675, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 43652 + }, + { + "epoch": 2.3249924812030076, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 43653 + }, + { + "epoch": 2.3250526315789473, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 43654 + }, + { + "epoch": 2.3251127819548874, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 43655 + }, + { + "epoch": 2.325172932330827, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 43656 + }, + { + "epoch": 2.325233082706767, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 43657 + }, + { + "epoch": 2.3252932330827067, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 43658 + }, + { + "epoch": 2.325353383458647, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 43659 + }, + { + "epoch": 2.3254135338345865, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 43660 + }, + { + "epoch": 2.325473684210526, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9182, + "step": 43661 + }, + { + "epoch": 2.3255338345864662, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 43662 + }, + { + "epoch": 2.325593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9242, + "step": 43663 + }, + { + "epoch": 2.325654135338346, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 43664 + }, + { + "epoch": 2.3257142857142856, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 43665 + }, + { + "epoch": 2.3257744360902257, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 43666 + }, + { + "epoch": 2.3258345864661654, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 43667 + }, + { + "epoch": 2.3258947368421055, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 43668 + }, + { + "epoch": 2.325954887218045, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 43669 + }, + { + "epoch": 2.3260150375939848, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9378, + "step": 43670 + }, + { + "epoch": 2.326075187969925, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 43671 + }, + { + "epoch": 2.3261353383458645, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 43672 + }, + { + "epoch": 2.3261954887218046, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 43673 + }, + { + "epoch": 2.3262556390977442, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 43674 + }, + { + "epoch": 2.3263157894736843, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 43675 + }, + { + "epoch": 2.326375939849624, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 43676 + }, + { + "epoch": 2.326436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 43677 + }, + { + "epoch": 2.3264962406015037, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1718, + "step": 43678 + }, + { + "epoch": 2.3265563909774434, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 43679 + }, + { + "epoch": 2.3266165413533835, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 43680 + }, + { + "epoch": 2.326676691729323, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9245, + "step": 43681 + }, + { + "epoch": 2.326736842105263, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 43682 + }, + { + "epoch": 2.326796992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 43683 + }, + { + "epoch": 2.326857142857143, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 43684 + }, + { + "epoch": 2.3269172932330826, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 43685 + }, + { + "epoch": 2.3269774436090227, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 43686 + }, + { + "epoch": 2.3270375939849623, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 43687 + }, + { + "epoch": 2.3270977443609024, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 43688 + }, + { + "epoch": 2.327157894736842, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 43689 + }, + { + "epoch": 2.327218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 43690 + }, + { + "epoch": 2.327278195488722, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 43691 + }, + { + "epoch": 2.3273383458646615, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 43692 + }, + { + "epoch": 2.3273984962406016, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 43693 + }, + { + "epoch": 2.3274586466165412, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 43694 + }, + { + "epoch": 2.3275187969924813, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 43695 + }, + { + "epoch": 2.327578947368421, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 43696 + }, + { + "epoch": 2.327639097744361, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 43697 + }, + { + "epoch": 2.3276992481203007, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 43698 + }, + { + "epoch": 2.327759398496241, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 43699 + }, + { + "epoch": 2.3278195488721805, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 43700 + }, + { + "epoch": 2.32787969924812, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 43701 + }, + { + "epoch": 2.32793984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 43702 + }, + { + "epoch": 2.328, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 43703 + }, + { + "epoch": 2.32806015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 43704 + }, + { + "epoch": 2.3281203007518796, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 43705 + }, + { + "epoch": 2.3281804511278197, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 43706 + }, + { + "epoch": 2.3282406015037593, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9386, + "step": 43707 + }, + { + "epoch": 2.3283007518796994, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 43708 + }, + { + "epoch": 2.328360902255639, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 43709 + }, + { + "epoch": 2.3284210526315787, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 43710 + }, + { + "epoch": 2.328481203007519, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 43711 + }, + { + "epoch": 2.3285413533834585, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 43712 + }, + { + "epoch": 2.3286015037593986, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 43713 + }, + { + "epoch": 2.328661654135338, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 43714 + }, + { + "epoch": 2.3287218045112783, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 43715 + }, + { + "epoch": 2.328781954887218, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 43716 + }, + { + "epoch": 2.328842105263158, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9667, + "step": 43717 + }, + { + "epoch": 2.3289022556390977, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 43718 + }, + { + "epoch": 2.3289624060150373, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 43719 + }, + { + "epoch": 2.3290225563909774, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 43720 + }, + { + "epoch": 2.3290827067669175, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 43721 + }, + { + "epoch": 2.329142857142857, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 43722 + }, + { + "epoch": 2.329203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 43723 + }, + { + "epoch": 2.329263157894737, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 43724 + }, + { + "epoch": 2.3293233082706766, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 43725 + }, + { + "epoch": 2.3293834586466167, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 43726 + }, + { + "epoch": 2.3294436090225563, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 43727 + }, + { + "epoch": 2.3295037593984964, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 43728 + }, + { + "epoch": 2.329563909774436, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 43729 + }, + { + "epoch": 2.329624060150376, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 43730 + }, + { + "epoch": 2.329684210526316, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 43731 + }, + { + "epoch": 2.3297443609022555, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 43732 + }, + { + "epoch": 2.3298045112781955, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 43733 + }, + { + "epoch": 2.329864661654135, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 43734 + }, + { + "epoch": 2.3299248120300753, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 43735 + }, + { + "epoch": 2.329984962406015, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 43736 + }, + { + "epoch": 2.330045112781955, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 43737 + }, + { + "epoch": 2.3301052631578947, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 43738 + }, + { + "epoch": 2.3301654135338348, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 43739 + }, + { + "epoch": 2.3302255639097744, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 43740 + }, + { + "epoch": 2.330285714285714, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9472, + "step": 43741 + }, + { + "epoch": 2.330345864661654, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 43742 + }, + { + "epoch": 2.330406015037594, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 43743 + }, + { + "epoch": 2.330466165413534, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 43744 + }, + { + "epoch": 2.3305263157894736, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 43745 + }, + { + "epoch": 2.3305864661654137, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.2006, + "step": 43746 + }, + { + "epoch": 2.3306466165413533, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 43747 + }, + { + "epoch": 2.3307067669172934, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 43748 + }, + { + "epoch": 2.330766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 43749 + }, + { + "epoch": 2.3308270676691727, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 43750 + }, + { + "epoch": 2.330887218045113, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 43751 + }, + { + "epoch": 2.330947368421053, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 43752 + }, + { + "epoch": 2.3310075187969925, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 43753 + }, + { + "epoch": 2.331067669172932, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 43754 + }, + { + "epoch": 2.3311278195488723, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 43755 + }, + { + "epoch": 2.331187969924812, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 43756 + }, + { + "epoch": 2.331248120300752, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9264, + "step": 43757 + }, + { + "epoch": 2.3313082706766917, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 43758 + }, + { + "epoch": 2.3313684210526318, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 43759 + }, + { + "epoch": 2.3314285714285714, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 43760 + }, + { + "epoch": 2.3314887218045115, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 43761 + }, + { + "epoch": 2.331548872180451, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 43762 + }, + { + "epoch": 2.331609022556391, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 43763 + }, + { + "epoch": 2.331669172932331, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 43764 + }, + { + "epoch": 2.3317293233082705, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 43765 + }, + { + "epoch": 2.3317894736842106, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 43766 + }, + { + "epoch": 2.3318496240601503, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9079, + "step": 43767 + }, + { + "epoch": 2.3319097744360904, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 43768 + }, + { + "epoch": 2.33196992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 43769 + }, + { + "epoch": 2.33203007518797, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 43770 + }, + { + "epoch": 2.3320902255639098, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9416, + "step": 43771 + }, + { + "epoch": 2.3321503759398494, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 43772 + }, + { + "epoch": 2.3322105263157895, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 43773 + }, + { + "epoch": 2.332270676691729, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 43774 + }, + { + "epoch": 2.3323308270676693, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 43775 + }, + { + "epoch": 2.332390977443609, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 43776 + }, + { + "epoch": 2.332451127819549, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 43777 + }, + { + "epoch": 2.3325112781954886, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 43778 + }, + { + "epoch": 2.3325714285714287, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 43779 + }, + { + "epoch": 2.3326315789473684, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9383, + "step": 43780 + }, + { + "epoch": 2.332691729323308, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 43781 + }, + { + "epoch": 2.332751879699248, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 43782 + }, + { + "epoch": 2.332812030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 43783 + }, + { + "epoch": 2.332872180451128, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 43784 + }, + { + "epoch": 2.3329323308270675, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 43785 + }, + { + "epoch": 2.3329924812030076, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 43786 + }, + { + "epoch": 2.3330526315789473, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 43787 + }, + { + "epoch": 2.3331127819548874, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 43788 + }, + { + "epoch": 2.333172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 43789 + }, + { + "epoch": 2.333233082706767, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 43790 + }, + { + "epoch": 2.3332932330827068, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 43791 + }, + { + "epoch": 2.333353383458647, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 43792 + }, + { + "epoch": 2.3334135338345865, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 43793 + }, + { + "epoch": 2.333473684210526, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 43794 + }, + { + "epoch": 2.3335338345864662, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 43795 + }, + { + "epoch": 2.333593984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 43796 + }, + { + "epoch": 2.333654135338346, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 43797 + }, + { + "epoch": 2.3337142857142856, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 43798 + }, + { + "epoch": 2.3337744360902257, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 43799 + }, + { + "epoch": 2.3338345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 43800 + }, + { + "epoch": 2.3338947368421055, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 43801 + }, + { + "epoch": 2.333954887218045, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 43802 + }, + { + "epoch": 2.3340150375939848, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 43803 + }, + { + "epoch": 2.334075187969925, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 43804 + }, + { + "epoch": 2.3341353383458645, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 43805 + }, + { + "epoch": 2.3341954887218046, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 43806 + }, + { + "epoch": 2.3342556390977443, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 43807 + }, + { + "epoch": 2.3343157894736843, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 43808 + }, + { + "epoch": 2.334375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 43809 + }, + { + "epoch": 2.334436090225564, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 43810 + }, + { + "epoch": 2.3344962406015037, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 43811 + }, + { + "epoch": 2.3345563909774434, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 43812 + }, + { + "epoch": 2.3346165413533835, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 43813 + }, + { + "epoch": 2.334676691729323, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 43814 + }, + { + "epoch": 2.334736842105263, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 43815 + }, + { + "epoch": 2.334796992481203, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 43816 + }, + { + "epoch": 2.334857142857143, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 43817 + }, + { + "epoch": 2.3349172932330826, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 43818 + }, + { + "epoch": 2.3349774436090227, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 43819 + }, + { + "epoch": 2.3350375939849624, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 43820 + }, + { + "epoch": 2.3350977443609024, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 43821 + }, + { + "epoch": 2.335157894736842, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 43822 + }, + { + "epoch": 2.335218045112782, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 43823 + }, + { + "epoch": 2.335278195488722, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 43824 + }, + { + "epoch": 2.3353383458646615, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 43825 + }, + { + "epoch": 2.3353984962406016, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9382, + "step": 43826 + }, + { + "epoch": 2.3354586466165412, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 43827 + }, + { + "epoch": 2.3355187969924813, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 43828 + }, + { + "epoch": 2.335578947368421, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 43829 + }, + { + "epoch": 2.335639097744361, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 43830 + }, + { + "epoch": 2.3356992481203007, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 43831 + }, + { + "epoch": 2.335759398496241, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 43832 + }, + { + "epoch": 2.3358195488721805, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 43833 + }, + { + "epoch": 2.33587969924812, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 43834 + }, + { + "epoch": 2.33593984962406, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 43835 + }, + { + "epoch": 2.336, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 43836 + }, + { + "epoch": 2.33606015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 43837 + }, + { + "epoch": 2.3361203007518796, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 43838 + }, + { + "epoch": 2.3361804511278197, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 43839 + }, + { + "epoch": 2.3362406015037593, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 43840 + }, + { + "epoch": 2.3363007518796994, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 43841 + }, + { + "epoch": 2.336360902255639, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 43842 + }, + { + "epoch": 2.3364210526315787, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 43843 + }, + { + "epoch": 2.336481203007519, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 43844 + }, + { + "epoch": 2.3365413533834585, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 43845 + }, + { + "epoch": 2.3366015037593986, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9497, + "step": 43846 + }, + { + "epoch": 2.336661654135338, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 43847 + }, + { + "epoch": 2.3367218045112783, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 43848 + }, + { + "epoch": 2.336781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 43849 + }, + { + "epoch": 2.336842105263158, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 43850 + }, + { + "epoch": 2.3369022556390977, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 43851 + }, + { + "epoch": 2.3369624060150374, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 43852 + }, + { + "epoch": 2.3370225563909774, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 43853 + }, + { + "epoch": 2.3370827067669175, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 43854 + }, + { + "epoch": 2.337142857142857, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 43855 + }, + { + "epoch": 2.337203007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 43856 + }, + { + "epoch": 2.337263157894737, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 43857 + }, + { + "epoch": 2.3373233082706766, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 43858 + }, + { + "epoch": 2.3373834586466167, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 43859 + }, + { + "epoch": 2.3374436090225563, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 43860 + }, + { + "epoch": 2.3375037593984964, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 43861 + }, + { + "epoch": 2.337563909774436, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 43862 + }, + { + "epoch": 2.337624060150376, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 43863 + }, + { + "epoch": 2.337684210526316, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 43864 + }, + { + "epoch": 2.3377443609022555, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 43865 + }, + { + "epoch": 2.3378045112781956, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 43866 + }, + { + "epoch": 2.337864661654135, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9226, + "step": 43867 + }, + { + "epoch": 2.3379248120300753, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 43868 + }, + { + "epoch": 2.337984962406015, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 43869 + }, + { + "epoch": 2.338045112781955, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 43870 + }, + { + "epoch": 2.3381052631578947, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 43871 + }, + { + "epoch": 2.338165413533835, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 43872 + }, + { + "epoch": 2.3382255639097744, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 43873 + }, + { + "epoch": 2.338285714285714, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 43874 + }, + { + "epoch": 2.338345864661654, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 43875 + }, + { + "epoch": 2.338406015037594, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 43876 + }, + { + "epoch": 2.338466165413534, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 43877 + }, + { + "epoch": 2.3385263157894736, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 43878 + }, + { + "epoch": 2.3385864661654137, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 43879 + }, + { + "epoch": 2.3386466165413533, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 43880 + }, + { + "epoch": 2.3387067669172934, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 43881 + }, + { + "epoch": 2.338766917293233, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 43882 + }, + { + "epoch": 2.3388270676691727, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.8942, + "step": 43883 + }, + { + "epoch": 2.338887218045113, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 43884 + }, + { + "epoch": 2.3389473684210524, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 43885 + }, + { + "epoch": 2.3390075187969925, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 43886 + }, + { + "epoch": 2.339067669172932, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 43887 + }, + { + "epoch": 2.3391278195488723, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9245, + "step": 43888 + }, + { + "epoch": 2.339187969924812, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 43889 + }, + { + "epoch": 2.339248120300752, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 43890 + }, + { + "epoch": 2.3393082706766917, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 43891 + }, + { + "epoch": 2.3393684210526318, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 43892 + }, + { + "epoch": 2.3394285714285714, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 43893 + }, + { + "epoch": 2.3394887218045115, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 43894 + }, + { + "epoch": 2.339548872180451, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 43895 + }, + { + "epoch": 2.339609022556391, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1822, + "step": 43896 + }, + { + "epoch": 2.339669172932331, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 43897 + }, + { + "epoch": 2.3397293233082705, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 43898 + }, + { + "epoch": 2.3397894736842106, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 43899 + }, + { + "epoch": 2.3398496240601503, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 43900 + }, + { + "epoch": 2.3399097744360904, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 43901 + }, + { + "epoch": 2.33996992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 43902 + }, + { + "epoch": 2.34003007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9442, + "step": 43903 + }, + { + "epoch": 2.3400902255639098, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 43904 + }, + { + "epoch": 2.3401503759398494, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 43905 + }, + { + "epoch": 2.3402105263157895, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 43906 + }, + { + "epoch": 2.340270676691729, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 43907 + }, + { + "epoch": 2.3403308270676693, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9472, + "step": 43908 + }, + { + "epoch": 2.340390977443609, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 43909 + }, + { + "epoch": 2.340451127819549, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 43910 + }, + { + "epoch": 2.3405112781954887, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 43911 + }, + { + "epoch": 2.3405714285714287, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 43912 + }, + { + "epoch": 2.3406315789473684, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 43913 + }, + { + "epoch": 2.340691729323308, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 43914 + }, + { + "epoch": 2.340751879699248, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 43915 + }, + { + "epoch": 2.340812030075188, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 43916 + }, + { + "epoch": 2.340872180451128, + "grad_norm": 1.4765625, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 43917 + }, + { + "epoch": 2.3409323308270675, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 43918 + }, + { + "epoch": 2.3409924812030076, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 43919 + }, + { + "epoch": 2.3410526315789473, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 43920 + }, + { + "epoch": 2.3411127819548874, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 43921 + }, + { + "epoch": 2.341172932330827, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 43922 + }, + { + "epoch": 2.341233082706767, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 43923 + }, + { + "epoch": 2.3412932330827068, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 43924 + }, + { + "epoch": 2.341353383458647, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 43925 + }, + { + "epoch": 2.3414135338345865, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0994, + "step": 43926 + }, + { + "epoch": 2.341473684210526, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 43927 + }, + { + "epoch": 2.3415338345864662, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 43928 + }, + { + "epoch": 2.341593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 43929 + }, + { + "epoch": 2.341654135338346, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 43930 + }, + { + "epoch": 2.3417142857142856, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 43931 + }, + { + "epoch": 2.3417744360902257, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 43932 + }, + { + "epoch": 2.3418345864661654, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 43933 + }, + { + "epoch": 2.3418947368421055, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 43934 + }, + { + "epoch": 2.341954887218045, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 43935 + }, + { + "epoch": 2.3420150375939848, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 43936 + }, + { + "epoch": 2.342075187969925, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 43937 + }, + { + "epoch": 2.3421353383458645, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 43938 + }, + { + "epoch": 2.3421954887218046, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 43939 + }, + { + "epoch": 2.3422556390977443, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 43940 + }, + { + "epoch": 2.3423157894736844, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 43941 + }, + { + "epoch": 2.342375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 43942 + }, + { + "epoch": 2.342436090225564, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 43943 + }, + { + "epoch": 2.3424962406015037, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9086, + "step": 43944 + }, + { + "epoch": 2.3425563909774434, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 43945 + }, + { + "epoch": 2.3426165413533835, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 43946 + }, + { + "epoch": 2.342676691729323, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 43947 + }, + { + "epoch": 2.3427368421052632, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 43948 + }, + { + "epoch": 2.342796992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 43949 + }, + { + "epoch": 2.342857142857143, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9322, + "step": 43950 + }, + { + "epoch": 2.3429172932330826, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 43951 + }, + { + "epoch": 2.3429774436090227, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 43952 + }, + { + "epoch": 2.3430375939849624, + "grad_norm": 1.3671875, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 43953 + }, + { + "epoch": 2.343097744360902, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 43954 + }, + { + "epoch": 2.343157894736842, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 43955 + }, + { + "epoch": 2.343218045112782, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 43956 + }, + { + "epoch": 2.343278195488722, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 43957 + }, + { + "epoch": 2.3433383458646615, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 43958 + }, + { + "epoch": 2.3433984962406016, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 43959 + }, + { + "epoch": 2.3434586466165412, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 43960 + }, + { + "epoch": 2.3435187969924813, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 43961 + }, + { + "epoch": 2.343578947368421, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 43962 + }, + { + "epoch": 2.343639097744361, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 43963 + }, + { + "epoch": 2.3436992481203007, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 43964 + }, + { + "epoch": 2.343759398496241, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 43965 + }, + { + "epoch": 2.3438195488721805, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 43966 + }, + { + "epoch": 2.34387969924812, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 43967 + }, + { + "epoch": 2.34393984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 43968 + }, + { + "epoch": 2.344, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 43969 + }, + { + "epoch": 2.34406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 43970 + }, + { + "epoch": 2.3441203007518796, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 43971 + }, + { + "epoch": 2.3441804511278197, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 43972 + }, + { + "epoch": 2.3442406015037593, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 43973 + }, + { + "epoch": 2.3443007518796994, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 43974 + }, + { + "epoch": 2.344360902255639, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 43975 + }, + { + "epoch": 2.3444210526315787, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 43976 + }, + { + "epoch": 2.344481203007519, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 43977 + }, + { + "epoch": 2.3445413533834585, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 43978 + }, + { + "epoch": 2.3446015037593986, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 43979 + }, + { + "epoch": 2.3446616541353382, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 43980 + }, + { + "epoch": 2.3447218045112783, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 43981 + }, + { + "epoch": 2.344781954887218, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 43982 + }, + { + "epoch": 2.344842105263158, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 43983 + }, + { + "epoch": 2.3449022556390977, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 43984 + }, + { + "epoch": 2.3449624060150374, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 43985 + }, + { + "epoch": 2.3450225563909775, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 43986 + }, + { + "epoch": 2.345082706766917, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 43987 + }, + { + "epoch": 2.345142857142857, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 43988 + }, + { + "epoch": 2.345203007518797, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 43989 + }, + { + "epoch": 2.345263157894737, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 43990 + }, + { + "epoch": 2.3453233082706766, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 43991 + }, + { + "epoch": 2.3453834586466167, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 43992 + }, + { + "epoch": 2.3454436090225563, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 43993 + }, + { + "epoch": 2.3455037593984964, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 43994 + }, + { + "epoch": 2.345563909774436, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 43995 + }, + { + "epoch": 2.345624060150376, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 43996 + }, + { + "epoch": 2.345684210526316, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 43997 + }, + { + "epoch": 2.3457443609022555, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 43998 + }, + { + "epoch": 2.3458045112781956, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 43999 + }, + { + "epoch": 2.345864661654135, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.935, + "step": 44000 + }, + { + "epoch": 2.3459248120300753, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 44001 + }, + { + "epoch": 2.345984962406015, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 44002 + }, + { + "epoch": 2.346045112781955, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 44003 + }, + { + "epoch": 2.3461052631578947, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 44004 + }, + { + "epoch": 2.346165413533835, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 44005 + }, + { + "epoch": 2.3462255639097744, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 44006 + }, + { + "epoch": 2.346285714285714, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 44007 + }, + { + "epoch": 2.346345864661654, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 44008 + }, + { + "epoch": 2.346406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9362, + "step": 44009 + }, + { + "epoch": 2.346466165413534, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 44010 + }, + { + "epoch": 2.3465263157894736, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 44011 + }, + { + "epoch": 2.3465864661654137, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 44012 + }, + { + "epoch": 2.3466466165413533, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 44013 + }, + { + "epoch": 2.3467067669172934, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 44014 + }, + { + "epoch": 2.346766917293233, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 44015 + }, + { + "epoch": 2.3468270676691727, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.895, + "step": 44016 + }, + { + "epoch": 2.346887218045113, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 44017 + }, + { + "epoch": 2.3469473684210524, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 44018 + }, + { + "epoch": 2.3470075187969925, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9223, + "step": 44019 + }, + { + "epoch": 2.347067669172932, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 44020 + }, + { + "epoch": 2.3471278195488723, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 44021 + }, + { + "epoch": 2.347187969924812, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 44022 + }, + { + "epoch": 2.347248120300752, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2126, + "step": 44023 + }, + { + "epoch": 2.3473082706766917, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 44024 + }, + { + "epoch": 2.3473684210526318, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 44025 + }, + { + "epoch": 2.3474285714285714, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 44026 + }, + { + "epoch": 2.3474887218045115, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 44027 + }, + { + "epoch": 2.347548872180451, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 44028 + }, + { + "epoch": 2.347609022556391, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 44029 + }, + { + "epoch": 2.347669172932331, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 44030 + }, + { + "epoch": 2.3477293233082706, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 44031 + }, + { + "epoch": 2.3477894736842106, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 44032 + }, + { + "epoch": 2.3478496240601503, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 44033 + }, + { + "epoch": 2.3479097744360904, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 44034 + }, + { + "epoch": 2.34796992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 44035 + }, + { + "epoch": 2.34803007518797, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 44036 + }, + { + "epoch": 2.34809022556391, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 44037 + }, + { + "epoch": 2.3481503759398494, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 44038 + }, + { + "epoch": 2.3482105263157895, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 44039 + }, + { + "epoch": 2.348270676691729, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 44040 + }, + { + "epoch": 2.3483308270676693, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 44041 + }, + { + "epoch": 2.348390977443609, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 44042 + }, + { + "epoch": 2.348451127819549, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 44043 + }, + { + "epoch": 2.3485112781954887, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 44044 + }, + { + "epoch": 2.3485714285714288, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 44045 + }, + { + "epoch": 2.3486315789473684, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 44046 + }, + { + "epoch": 2.348691729323308, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 44047 + }, + { + "epoch": 2.348751879699248, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 44048 + }, + { + "epoch": 2.348812030075188, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 44049 + }, + { + "epoch": 2.348872180451128, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 44050 + }, + { + "epoch": 2.3489323308270675, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 44051 + }, + { + "epoch": 2.3489924812030076, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 44052 + }, + { + "epoch": 2.3490526315789473, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 44053 + }, + { + "epoch": 2.3491127819548874, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 44054 + }, + { + "epoch": 2.349172932330827, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 44055 + }, + { + "epoch": 2.3492330827067667, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 44056 + }, + { + "epoch": 2.3492932330827068, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 44057 + }, + { + "epoch": 2.349353383458647, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 44058 + }, + { + "epoch": 2.3494135338345865, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 44059 + }, + { + "epoch": 2.349473684210526, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 44060 + }, + { + "epoch": 2.3495338345864663, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 44061 + }, + { + "epoch": 2.349593984962406, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 44062 + }, + { + "epoch": 2.349654135338346, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 44063 + }, + { + "epoch": 2.3497142857142856, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 44064 + }, + { + "epoch": 2.3497744360902257, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 44065 + }, + { + "epoch": 2.3498345864661654, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 44066 + }, + { + "epoch": 2.3498947368421055, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 44067 + }, + { + "epoch": 2.349954887218045, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 44068 + }, + { + "epoch": 2.350015037593985, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 44069 + }, + { + "epoch": 2.350075187969925, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 44070 + }, + { + "epoch": 2.3501353383458645, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 44071 + }, + { + "epoch": 2.3501954887218046, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 44072 + }, + { + "epoch": 2.3502556390977443, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 44073 + }, + { + "epoch": 2.3503157894736844, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 44074 + }, + { + "epoch": 2.350375939849624, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 44075 + }, + { + "epoch": 2.350436090225564, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 44076 + }, + { + "epoch": 2.3504962406015038, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 44077 + }, + { + "epoch": 2.3505563909774434, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 44078 + }, + { + "epoch": 2.3506165413533835, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 44079 + }, + { + "epoch": 2.350676691729323, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 44080 + }, + { + "epoch": 2.3507368421052632, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 44081 + }, + { + "epoch": 2.350796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9168, + "step": 44082 + }, + { + "epoch": 2.350857142857143, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 44083 + }, + { + "epoch": 2.3509172932330826, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 44084 + }, + { + "epoch": 2.3509774436090227, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 44085 + }, + { + "epoch": 2.3510375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 44086 + }, + { + "epoch": 2.351097744360902, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 44087 + }, + { + "epoch": 2.351157894736842, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 44088 + }, + { + "epoch": 2.3512180451127818, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 44089 + }, + { + "epoch": 2.351278195488722, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 44090 + }, + { + "epoch": 2.3513383458646615, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 44091 + }, + { + "epoch": 2.3513984962406016, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 44092 + }, + { + "epoch": 2.3514586466165412, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 44093 + }, + { + "epoch": 2.3515187969924813, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 44094 + }, + { + "epoch": 2.351578947368421, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 44095 + }, + { + "epoch": 2.351639097744361, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 44096 + }, + { + "epoch": 2.3516992481203007, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 44097 + }, + { + "epoch": 2.351759398496241, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 44098 + }, + { + "epoch": 2.3518195488721805, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.336, + "step": 44099 + }, + { + "epoch": 2.35187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 44100 + }, + { + "epoch": 2.35193984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 44101 + }, + { + "epoch": 2.352, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.2063, + "step": 44102 + }, + { + "epoch": 2.35206015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9245, + "step": 44103 + }, + { + "epoch": 2.3521203007518796, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 44104 + }, + { + "epoch": 2.3521804511278197, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 44105 + }, + { + "epoch": 2.3522406015037594, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 44106 + }, + { + "epoch": 2.3523007518796994, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 44107 + }, + { + "epoch": 2.352360902255639, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 44108 + }, + { + "epoch": 2.3524210526315787, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 44109 + }, + { + "epoch": 2.352481203007519, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 44110 + }, + { + "epoch": 2.3525413533834585, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 44111 + }, + { + "epoch": 2.3526015037593986, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 44112 + }, + { + "epoch": 2.3526616541353382, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9345, + "step": 44113 + }, + { + "epoch": 2.3527218045112783, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 44114 + }, + { + "epoch": 2.352781954887218, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 44115 + }, + { + "epoch": 2.352842105263158, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 44116 + }, + { + "epoch": 2.3529022556390977, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 44117 + }, + { + "epoch": 2.3529624060150374, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 44118 + }, + { + "epoch": 2.3530225563909775, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 44119 + }, + { + "epoch": 2.353082706766917, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9444, + "step": 44120 + }, + { + "epoch": 2.353142857142857, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 44121 + }, + { + "epoch": 2.353203007518797, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 44122 + }, + { + "epoch": 2.353263157894737, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 44123 + }, + { + "epoch": 2.3533233082706766, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 44124 + }, + { + "epoch": 2.3533834586466167, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 44125 + }, + { + "epoch": 2.3534436090225563, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 44126 + }, + { + "epoch": 2.3535037593984964, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 44127 + }, + { + "epoch": 2.353563909774436, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 44128 + }, + { + "epoch": 2.353624060150376, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 44129 + }, + { + "epoch": 2.353684210526316, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 44130 + }, + { + "epoch": 2.3537443609022555, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 44131 + }, + { + "epoch": 2.3538045112781956, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 44132 + }, + { + "epoch": 2.353864661654135, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 44133 + }, + { + "epoch": 2.3539248120300753, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 44134 + }, + { + "epoch": 2.353984962406015, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 44135 + }, + { + "epoch": 2.354045112781955, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 44136 + }, + { + "epoch": 2.3541052631578947, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 44137 + }, + { + "epoch": 2.354165413533835, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 44138 + }, + { + "epoch": 2.3542255639097744, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 44139 + }, + { + "epoch": 2.354285714285714, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 44140 + }, + { + "epoch": 2.354345864661654, + "grad_norm": 1.5234375, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 44141 + }, + { + "epoch": 2.354406015037594, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 44142 + }, + { + "epoch": 2.354466165413534, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 44143 + }, + { + "epoch": 2.3545263157894736, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 44144 + }, + { + "epoch": 2.3545864661654137, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9499, + "step": 44145 + }, + { + "epoch": 2.3546466165413533, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 44146 + }, + { + "epoch": 2.3547067669172934, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 44147 + }, + { + "epoch": 2.354766917293233, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 44148 + }, + { + "epoch": 2.3548270676691727, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 44149 + }, + { + "epoch": 2.354887218045113, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 44150 + }, + { + "epoch": 2.3549473684210525, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 44151 + }, + { + "epoch": 2.3550075187969925, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 44152 + }, + { + "epoch": 2.355067669172932, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1504, + "step": 44153 + }, + { + "epoch": 2.3551278195488723, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 44154 + }, + { + "epoch": 2.355187969924812, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 44155 + }, + { + "epoch": 2.355248120300752, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 44156 + }, + { + "epoch": 2.3553082706766917, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 44157 + }, + { + "epoch": 2.3553684210526313, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 44158 + }, + { + "epoch": 2.3554285714285714, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 44159 + }, + { + "epoch": 2.3554887218045115, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1682, + "step": 44160 + }, + { + "epoch": 2.355548872180451, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 44161 + }, + { + "epoch": 2.355609022556391, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 44162 + }, + { + "epoch": 2.355669172932331, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 44163 + }, + { + "epoch": 2.3557293233082706, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 44164 + }, + { + "epoch": 2.3557894736842107, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 44165 + }, + { + "epoch": 2.3558496240601503, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 44166 + }, + { + "epoch": 2.3559097744360904, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 44167 + }, + { + "epoch": 2.35596992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 44168 + }, + { + "epoch": 2.35603007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 44169 + }, + { + "epoch": 2.35609022556391, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 44170 + }, + { + "epoch": 2.3561503759398494, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 44171 + }, + { + "epoch": 2.3562105263157895, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 44172 + }, + { + "epoch": 2.356270676691729, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 44173 + }, + { + "epoch": 2.3563308270676693, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9284, + "step": 44174 + }, + { + "epoch": 2.356390977443609, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 44175 + }, + { + "epoch": 2.356451127819549, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 44176 + }, + { + "epoch": 2.3565112781954887, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 44177 + }, + { + "epoch": 2.3565714285714288, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 44178 + }, + { + "epoch": 2.3566315789473684, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 44179 + }, + { + "epoch": 2.356691729323308, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 44180 + }, + { + "epoch": 2.356751879699248, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 44181 + }, + { + "epoch": 2.356812030075188, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 44182 + }, + { + "epoch": 2.356872180451128, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9421, + "step": 44183 + }, + { + "epoch": 2.3569323308270675, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 44184 + }, + { + "epoch": 2.3569924812030076, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 44185 + }, + { + "epoch": 2.3570526315789473, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 44186 + }, + { + "epoch": 2.3571127819548874, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 44187 + }, + { + "epoch": 2.357172932330827, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 44188 + }, + { + "epoch": 2.3572330827067667, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 44189 + }, + { + "epoch": 2.3572932330827068, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 44190 + }, + { + "epoch": 2.3573533834586464, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 44191 + }, + { + "epoch": 2.3574135338345865, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 44192 + }, + { + "epoch": 2.357473684210526, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 44193 + }, + { + "epoch": 2.3575338345864663, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 44194 + }, + { + "epoch": 2.357593984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 44195 + }, + { + "epoch": 2.357654135338346, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1342, + "step": 44196 + }, + { + "epoch": 2.3577142857142857, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 44197 + }, + { + "epoch": 2.3577744360902257, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 44198 + }, + { + "epoch": 2.3578345864661654, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 44199 + }, + { + "epoch": 2.3578947368421055, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 44200 + }, + { + "epoch": 2.357954887218045, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 44201 + }, + { + "epoch": 2.358015037593985, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1652, + "step": 44202 + }, + { + "epoch": 2.358075187969925, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 44203 + }, + { + "epoch": 2.3581353383458645, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 44204 + }, + { + "epoch": 2.3581954887218046, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 44205 + }, + { + "epoch": 2.3582556390977443, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9516, + "step": 44206 + }, + { + "epoch": 2.3583157894736844, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 44207 + }, + { + "epoch": 2.358375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 44208 + }, + { + "epoch": 2.358436090225564, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 44209 + }, + { + "epoch": 2.3584962406015038, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 44210 + }, + { + "epoch": 2.3585563909774434, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 44211 + }, + { + "epoch": 2.3586165413533835, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 44212 + }, + { + "epoch": 2.358676691729323, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 44213 + }, + { + "epoch": 2.3587368421052632, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 44214 + }, + { + "epoch": 2.358796992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 44215 + }, + { + "epoch": 2.358857142857143, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 44216 + }, + { + "epoch": 2.3589172932330826, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 44217 + }, + { + "epoch": 2.3589774436090227, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 44218 + }, + { + "epoch": 2.3590375939849624, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9248, + "step": 44219 + }, + { + "epoch": 2.359097744360902, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 44220 + }, + { + "epoch": 2.359157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 44221 + }, + { + "epoch": 2.3592180451127818, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 44222 + }, + { + "epoch": 2.359278195488722, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1271, + "step": 44223 + }, + { + "epoch": 2.3593383458646615, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1769, + "step": 44224 + }, + { + "epoch": 2.3593984962406016, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 44225 + }, + { + "epoch": 2.3594586466165413, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 44226 + }, + { + "epoch": 2.3595187969924813, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 44227 + }, + { + "epoch": 2.359578947368421, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 44228 + }, + { + "epoch": 2.359639097744361, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 44229 + }, + { + "epoch": 2.3596992481203007, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 44230 + }, + { + "epoch": 2.359759398496241, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 44231 + }, + { + "epoch": 2.3598195488721805, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 44232 + }, + { + "epoch": 2.35987969924812, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 44233 + }, + { + "epoch": 2.3599398496240602, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 44234 + }, + { + "epoch": 2.36, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 44235 + }, + { + "epoch": 2.36006015037594, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 44236 + }, + { + "epoch": 2.3601203007518796, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 44237 + }, + { + "epoch": 2.3601804511278197, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.2009, + "step": 44238 + }, + { + "epoch": 2.3602406015037594, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 44239 + }, + { + "epoch": 2.3603007518796995, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 44240 + }, + { + "epoch": 2.360360902255639, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 44241 + }, + { + "epoch": 2.3604210526315788, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 44242 + }, + { + "epoch": 2.360481203007519, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 44243 + }, + { + "epoch": 2.3605413533834585, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 44244 + }, + { + "epoch": 2.3606015037593986, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 44245 + }, + { + "epoch": 2.3606616541353382, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 44246 + }, + { + "epoch": 2.3607218045112783, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 44247 + }, + { + "epoch": 2.360781954887218, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 44248 + }, + { + "epoch": 2.360842105263158, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 44249 + }, + { + "epoch": 2.3609022556390977, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 44250 + }, + { + "epoch": 2.3609624060150374, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 44251 + }, + { + "epoch": 2.3610225563909775, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.203, + "step": 44252 + }, + { + "epoch": 2.361082706766917, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 44253 + }, + { + "epoch": 2.361142857142857, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 44254 + }, + { + "epoch": 2.361203007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 44255 + }, + { + "epoch": 2.361263157894737, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 44256 + }, + { + "epoch": 2.3613233082706766, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 44257 + }, + { + "epoch": 2.3613834586466167, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 44258 + }, + { + "epoch": 2.3614436090225563, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 44259 + }, + { + "epoch": 2.361503759398496, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 44260 + }, + { + "epoch": 2.361563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 44261 + }, + { + "epoch": 2.361624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 44262 + }, + { + "epoch": 2.361684210526316, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 44263 + }, + { + "epoch": 2.3617443609022555, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 44264 + }, + { + "epoch": 2.3618045112781956, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 44265 + }, + { + "epoch": 2.3618646616541352, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 44266 + }, + { + "epoch": 2.3619248120300753, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 44267 + }, + { + "epoch": 2.361984962406015, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 44268 + }, + { + "epoch": 2.362045112781955, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9277, + "step": 44269 + }, + { + "epoch": 2.3621052631578947, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 44270 + }, + { + "epoch": 2.362165413533835, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 44271 + }, + { + "epoch": 2.3622255639097745, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 44272 + }, + { + "epoch": 2.362285714285714, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 44273 + }, + { + "epoch": 2.362345864661654, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 44274 + }, + { + "epoch": 2.362406015037594, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.2028, + "step": 44275 + }, + { + "epoch": 2.362466165413534, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 44276 + }, + { + "epoch": 2.3625263157894736, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 44277 + }, + { + "epoch": 2.3625864661654137, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 44278 + }, + { + "epoch": 2.3626466165413533, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 44279 + }, + { + "epoch": 2.3627067669172934, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 44280 + }, + { + "epoch": 2.362766917293233, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 44281 + }, + { + "epoch": 2.3628270676691727, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 44282 + }, + { + "epoch": 2.362887218045113, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 44283 + }, + { + "epoch": 2.3629473684210525, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 44284 + }, + { + "epoch": 2.3630075187969926, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 44285 + }, + { + "epoch": 2.363067669172932, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 44286 + }, + { + "epoch": 2.3631278195488723, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 44287 + }, + { + "epoch": 2.363187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 44288 + }, + { + "epoch": 2.363248120300752, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 44289 + }, + { + "epoch": 2.3633082706766917, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 44290 + }, + { + "epoch": 2.3633684210526313, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 44291 + }, + { + "epoch": 2.3634285714285714, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 44292 + }, + { + "epoch": 2.3634887218045115, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 44293 + }, + { + "epoch": 2.363548872180451, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 44294 + }, + { + "epoch": 2.363609022556391, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 44295 + }, + { + "epoch": 2.363669172932331, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 44296 + }, + { + "epoch": 2.3637293233082706, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 44297 + }, + { + "epoch": 2.3637894736842107, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 44298 + }, + { + "epoch": 2.3638496240601503, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 44299 + }, + { + "epoch": 2.3639097744360904, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 44300 + }, + { + "epoch": 2.36396992481203, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 44301 + }, + { + "epoch": 2.36403007518797, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 44302 + }, + { + "epoch": 2.36409022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 44303 + }, + { + "epoch": 2.3641503759398494, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 44304 + }, + { + "epoch": 2.3642105263157895, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 44305 + }, + { + "epoch": 2.364270676691729, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 44306 + }, + { + "epoch": 2.3643308270676693, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 44307 + }, + { + "epoch": 2.364390977443609, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 44308 + }, + { + "epoch": 2.364451127819549, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 44309 + }, + { + "epoch": 2.3645112781954887, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 44310 + }, + { + "epoch": 2.3645714285714288, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 44311 + }, + { + "epoch": 2.3646315789473684, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 44312 + }, + { + "epoch": 2.364691729323308, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 44313 + }, + { + "epoch": 2.364751879699248, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 44314 + }, + { + "epoch": 2.364812030075188, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 44315 + }, + { + "epoch": 2.364872180451128, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 44316 + }, + { + "epoch": 2.3649323308270676, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 44317 + }, + { + "epoch": 2.3649924812030076, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 44318 + }, + { + "epoch": 2.3650526315789473, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 44319 + }, + { + "epoch": 2.3651127819548874, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9413, + "step": 44320 + }, + { + "epoch": 2.365172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 44321 + }, + { + "epoch": 2.3652330827067667, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.8713, + "step": 44322 + }, + { + "epoch": 2.365293233082707, + "grad_norm": 1.8671875, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 44323 + }, + { + "epoch": 2.3653533834586464, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 44324 + }, + { + "epoch": 2.3654135338345865, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 44325 + }, + { + "epoch": 2.365473684210526, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 44326 + }, + { + "epoch": 2.3655338345864663, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 44327 + }, + { + "epoch": 2.365593984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 44328 + }, + { + "epoch": 2.365654135338346, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 44329 + }, + { + "epoch": 2.3657142857142857, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 44330 + }, + { + "epoch": 2.3657744360902258, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 44331 + }, + { + "epoch": 2.3658345864661654, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 44332 + }, + { + "epoch": 2.3658947368421055, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 44333 + }, + { + "epoch": 2.365954887218045, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 44334 + }, + { + "epoch": 2.366015037593985, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 44335 + }, + { + "epoch": 2.366075187969925, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 44336 + }, + { + "epoch": 2.3661353383458645, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 44337 + }, + { + "epoch": 2.3661954887218046, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 44338 + }, + { + "epoch": 2.3662556390977443, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 44339 + }, + { + "epoch": 2.3663157894736844, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 44340 + }, + { + "epoch": 2.366375939849624, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 44341 + }, + { + "epoch": 2.366436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 44342 + }, + { + "epoch": 2.3664962406015038, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 44343 + }, + { + "epoch": 2.3665563909774434, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9151, + "step": 44344 + }, + { + "epoch": 2.3666165413533835, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 44345 + }, + { + "epoch": 2.366676691729323, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 44346 + }, + { + "epoch": 2.3667368421052632, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9376, + "step": 44347 + }, + { + "epoch": 2.366796992481203, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 44348 + }, + { + "epoch": 2.366857142857143, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 44349 + }, + { + "epoch": 2.3669172932330826, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 44350 + }, + { + "epoch": 2.3669774436090227, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 44351 + }, + { + "epoch": 2.3670375939849624, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 44352 + }, + { + "epoch": 2.367097744360902, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 44353 + }, + { + "epoch": 2.367157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 44354 + }, + { + "epoch": 2.3672180451127818, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 44355 + }, + { + "epoch": 2.367278195488722, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 44356 + }, + { + "epoch": 2.3673383458646615, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 44357 + }, + { + "epoch": 2.3673984962406016, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 44358 + }, + { + "epoch": 2.3674586466165413, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 44359 + }, + { + "epoch": 2.3675187969924814, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 44360 + }, + { + "epoch": 2.367578947368421, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 44361 + }, + { + "epoch": 2.3676390977443607, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 44362 + }, + { + "epoch": 2.3676992481203007, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 44363 + }, + { + "epoch": 2.367759398496241, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 44364 + }, + { + "epoch": 2.3678195488721805, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 44365 + }, + { + "epoch": 2.36787969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 44366 + }, + { + "epoch": 2.3679398496240602, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 44367 + }, + { + "epoch": 2.368, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1453, + "step": 44368 + }, + { + "epoch": 2.36806015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 44369 + }, + { + "epoch": 2.3681203007518796, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 44370 + }, + { + "epoch": 2.3681804511278197, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 44371 + }, + { + "epoch": 2.3682406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9372, + "step": 44372 + }, + { + "epoch": 2.3683007518796995, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 44373 + }, + { + "epoch": 2.368360902255639, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 44374 + }, + { + "epoch": 2.3684210526315788, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 44375 + }, + { + "epoch": 2.368481203007519, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9366, + "step": 44376 + }, + { + "epoch": 2.3685413533834585, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 44377 + }, + { + "epoch": 2.3686015037593986, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 44378 + }, + { + "epoch": 2.3686616541353382, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 44379 + }, + { + "epoch": 2.3687218045112783, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 44380 + }, + { + "epoch": 2.368781954887218, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 44381 + }, + { + "epoch": 2.368842105263158, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 44382 + }, + { + "epoch": 2.3689022556390977, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 44383 + }, + { + "epoch": 2.3689624060150374, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 44384 + }, + { + "epoch": 2.3690225563909775, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 44385 + }, + { + "epoch": 2.369082706766917, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 44386 + }, + { + "epoch": 2.369142857142857, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1623, + "step": 44387 + }, + { + "epoch": 2.369203007518797, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 44388 + }, + { + "epoch": 2.369263157894737, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 44389 + }, + { + "epoch": 2.3693233082706766, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 44390 + }, + { + "epoch": 2.3693834586466167, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 44391 + }, + { + "epoch": 2.3694436090225564, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 44392 + }, + { + "epoch": 2.369503759398496, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 44393 + }, + { + "epoch": 2.369563909774436, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 44394 + }, + { + "epoch": 2.369624060150376, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 44395 + }, + { + "epoch": 2.369684210526316, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 44396 + }, + { + "epoch": 2.3697443609022555, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 44397 + }, + { + "epoch": 2.3698045112781956, + "grad_norm": 2.09375, + "learning_rate": 3e-05, + "loss": 0.9296, + "step": 44398 + }, + { + "epoch": 2.3698646616541352, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 44399 + }, + { + "epoch": 2.3699248120300753, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 44400 + }, + { + "epoch": 2.369984962406015, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 44401 + }, + { + "epoch": 2.370045112781955, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 44402 + }, + { + "epoch": 2.3701052631578947, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 44403 + }, + { + "epoch": 2.370165413533835, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 44404 + }, + { + "epoch": 2.3702255639097745, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1424, + "step": 44405 + }, + { + "epoch": 2.370285714285714, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 44406 + }, + { + "epoch": 2.370345864661654, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 44407 + }, + { + "epoch": 2.370406015037594, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 44408 + }, + { + "epoch": 2.370466165413534, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 44409 + }, + { + "epoch": 2.3705263157894736, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 44410 + }, + { + "epoch": 2.3705864661654137, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 44411 + }, + { + "epoch": 2.3706466165413533, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9279, + "step": 44412 + }, + { + "epoch": 2.3707067669172934, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 44413 + }, + { + "epoch": 2.370766917293233, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.8951, + "step": 44414 + }, + { + "epoch": 2.3708270676691727, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 44415 + }, + { + "epoch": 2.370887218045113, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 44416 + }, + { + "epoch": 2.3709473684210525, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 44417 + }, + { + "epoch": 2.3710075187969926, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 44418 + }, + { + "epoch": 2.371067669172932, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 44419 + }, + { + "epoch": 2.3711278195488723, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 44420 + }, + { + "epoch": 2.371187969924812, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 44421 + }, + { + "epoch": 2.371248120300752, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 44422 + }, + { + "epoch": 2.3713082706766917, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 44423 + }, + { + "epoch": 2.3713684210526313, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 44424 + }, + { + "epoch": 2.3714285714285714, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 44425 + }, + { + "epoch": 2.371488721804511, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 44426 + }, + { + "epoch": 2.371548872180451, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 44427 + }, + { + "epoch": 2.371609022556391, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 44428 + }, + { + "epoch": 2.371669172932331, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 44429 + }, + { + "epoch": 2.3717293233082706, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 44430 + }, + { + "epoch": 2.3717894736842107, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9251, + "step": 44431 + }, + { + "epoch": 2.3718496240601503, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 44432 + }, + { + "epoch": 2.3719097744360904, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 44433 + }, + { + "epoch": 2.37196992481203, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1464, + "step": 44434 + }, + { + "epoch": 2.37203007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 44435 + }, + { + "epoch": 2.37209022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9447, + "step": 44436 + }, + { + "epoch": 2.3721503759398495, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 44437 + }, + { + "epoch": 2.3722105263157895, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 44438 + }, + { + "epoch": 2.372270676691729, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 44439 + }, + { + "epoch": 2.3723308270676693, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 44440 + }, + { + "epoch": 2.372390977443609, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 44441 + }, + { + "epoch": 2.372451127819549, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 44442 + }, + { + "epoch": 2.3725112781954887, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 44443 + }, + { + "epoch": 2.3725714285714288, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 44444 + }, + { + "epoch": 2.3726315789473684, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 44445 + }, + { + "epoch": 2.372691729323308, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 44446 + }, + { + "epoch": 2.372751879699248, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 44447 + }, + { + "epoch": 2.372812030075188, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.151, + "step": 44448 + }, + { + "epoch": 2.372872180451128, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 44449 + }, + { + "epoch": 2.3729323308270676, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 44450 + }, + { + "epoch": 2.3729924812030077, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 44451 + }, + { + "epoch": 2.3730526315789473, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 44452 + }, + { + "epoch": 2.3731127819548874, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 44453 + }, + { + "epoch": 2.373172932330827, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 44454 + }, + { + "epoch": 2.3732330827067667, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 44455 + }, + { + "epoch": 2.373293233082707, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 44456 + }, + { + "epoch": 2.3733533834586464, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 44457 + }, + { + "epoch": 2.3734135338345865, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 44458 + }, + { + "epoch": 2.373473684210526, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 44459 + }, + { + "epoch": 2.3735338345864663, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 44460 + }, + { + "epoch": 2.373593984962406, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 44461 + }, + { + "epoch": 2.373654135338346, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 44462 + }, + { + "epoch": 2.3737142857142857, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 44463 + }, + { + "epoch": 2.3737744360902253, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 44464 + }, + { + "epoch": 2.3738345864661654, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 44465 + }, + { + "epoch": 2.3738947368421055, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 44466 + }, + { + "epoch": 2.373954887218045, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1765, + "step": 44467 + }, + { + "epoch": 2.374015037593985, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 44468 + }, + { + "epoch": 2.374075187969925, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 44469 + }, + { + "epoch": 2.3741353383458645, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 44470 + }, + { + "epoch": 2.3741954887218046, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 44471 + }, + { + "epoch": 2.3742556390977443, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 44472 + }, + { + "epoch": 2.3743157894736844, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 44473 + }, + { + "epoch": 2.374375939849624, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 44474 + }, + { + "epoch": 2.374436090225564, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 44475 + }, + { + "epoch": 2.3744962406015038, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 44476 + }, + { + "epoch": 2.3745563909774434, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 44477 + }, + { + "epoch": 2.3746165413533835, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 44478 + }, + { + "epoch": 2.374676691729323, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 44479 + }, + { + "epoch": 2.3747368421052633, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 44480 + }, + { + "epoch": 2.374796992481203, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 44481 + }, + { + "epoch": 2.374857142857143, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 44482 + }, + { + "epoch": 2.3749172932330826, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 44483 + }, + { + "epoch": 2.3749774436090227, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 44484 + }, + { + "epoch": 2.3750375939849624, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 44485 + }, + { + "epoch": 2.375097744360902, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 44486 + }, + { + "epoch": 2.375157894736842, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 44487 + }, + { + "epoch": 2.375218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 44488 + }, + { + "epoch": 2.375278195488722, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 44489 + }, + { + "epoch": 2.3753383458646615, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 44490 + }, + { + "epoch": 2.3753984962406016, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 44491 + }, + { + "epoch": 2.3754586466165413, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 44492 + }, + { + "epoch": 2.3755187969924814, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 44493 + }, + { + "epoch": 2.375578947368421, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 44494 + }, + { + "epoch": 2.3756390977443607, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 44495 + }, + { + "epoch": 2.3756992481203008, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 44496 + }, + { + "epoch": 2.375759398496241, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 44497 + }, + { + "epoch": 2.3758195488721805, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 44498 + }, + { + "epoch": 2.37587969924812, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 44499 + }, + { + "epoch": 2.3759398496240602, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 44500 + }, + { + "epoch": 2.376, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 44501 + }, + { + "epoch": 2.37606015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 44502 + }, + { + "epoch": 2.3761203007518796, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 44503 + }, + { + "epoch": 2.3761804511278197, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 44504 + }, + { + "epoch": 2.3762406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 44505 + }, + { + "epoch": 2.3763007518796995, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 44506 + }, + { + "epoch": 2.376360902255639, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 44507 + }, + { + "epoch": 2.3764210526315788, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 44508 + }, + { + "epoch": 2.376481203007519, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.2011, + "step": 44509 + }, + { + "epoch": 2.3765413533834585, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 44510 + }, + { + "epoch": 2.3766015037593986, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 44511 + }, + { + "epoch": 2.3766616541353383, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 44512 + }, + { + "epoch": 2.3767218045112783, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 44513 + }, + { + "epoch": 2.376781954887218, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 44514 + }, + { + "epoch": 2.376842105263158, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 44515 + }, + { + "epoch": 2.3769022556390977, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 44516 + }, + { + "epoch": 2.3769624060150374, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 44517 + }, + { + "epoch": 2.3770225563909775, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 44518 + }, + { + "epoch": 2.377082706766917, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 44519 + }, + { + "epoch": 2.3771428571428572, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 44520 + }, + { + "epoch": 2.377203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 44521 + }, + { + "epoch": 2.377263157894737, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 44522 + }, + { + "epoch": 2.3773233082706766, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.2022, + "step": 44523 + }, + { + "epoch": 2.3773834586466167, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 44524 + }, + { + "epoch": 2.3774436090225564, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 44525 + }, + { + "epoch": 2.377503759398496, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 44526 + }, + { + "epoch": 2.377563909774436, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 44527 + }, + { + "epoch": 2.3776240601503758, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 44528 + }, + { + "epoch": 2.377684210526316, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 44529 + }, + { + "epoch": 2.3777443609022555, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 44530 + }, + { + "epoch": 2.3778045112781956, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 44531 + }, + { + "epoch": 2.3778646616541352, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 44532 + }, + { + "epoch": 2.3779248120300753, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 44533 + }, + { + "epoch": 2.377984962406015, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 44534 + }, + { + "epoch": 2.378045112781955, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 44535 + }, + { + "epoch": 2.3781052631578947, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 44536 + }, + { + "epoch": 2.378165413533835, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 44537 + }, + { + "epoch": 2.3782255639097745, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 44538 + }, + { + "epoch": 2.378285714285714, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 44539 + }, + { + "epoch": 2.378345864661654, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 44540 + }, + { + "epoch": 2.378406015037594, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 44541 + }, + { + "epoch": 2.378466165413534, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 44542 + }, + { + "epoch": 2.3785263157894736, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 44543 + }, + { + "epoch": 2.3785864661654137, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 44544 + }, + { + "epoch": 2.3786466165413533, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 44545 + }, + { + "epoch": 2.3787067669172934, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 44546 + }, + { + "epoch": 2.378766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 44547 + }, + { + "epoch": 2.3788270676691727, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 44548 + }, + { + "epoch": 2.378887218045113, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 44549 + }, + { + "epoch": 2.3789473684210525, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 44550 + }, + { + "epoch": 2.3790075187969926, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 44551 + }, + { + "epoch": 2.379067669172932, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 44552 + }, + { + "epoch": 2.3791278195488723, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 44553 + }, + { + "epoch": 2.379187969924812, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9201, + "step": 44554 + }, + { + "epoch": 2.379248120300752, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 44555 + }, + { + "epoch": 2.3793082706766917, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 44556 + }, + { + "epoch": 2.3793684210526314, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 44557 + }, + { + "epoch": 2.3794285714285714, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 44558 + }, + { + "epoch": 2.379488721804511, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 44559 + }, + { + "epoch": 2.379548872180451, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 44560 + }, + { + "epoch": 2.379609022556391, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9085, + "step": 44561 + }, + { + "epoch": 2.379669172932331, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 44562 + }, + { + "epoch": 2.3797293233082706, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 44563 + }, + { + "epoch": 2.3797894736842107, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 44564 + }, + { + "epoch": 2.3798496240601503, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 44565 + }, + { + "epoch": 2.3799097744360904, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.957, + "step": 44566 + }, + { + "epoch": 2.37996992481203, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 44567 + }, + { + "epoch": 2.38003007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 44568 + }, + { + "epoch": 2.38009022556391, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9323, + "step": 44569 + }, + { + "epoch": 2.3801503759398495, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 44570 + }, + { + "epoch": 2.3802105263157896, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 44571 + }, + { + "epoch": 2.380270676691729, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.8993, + "step": 44572 + }, + { + "epoch": 2.3803308270676693, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 44573 + }, + { + "epoch": 2.380390977443609, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1762, + "step": 44574 + }, + { + "epoch": 2.380451127819549, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 44575 + }, + { + "epoch": 2.3805112781954887, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 44576 + }, + { + "epoch": 2.380571428571429, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 44577 + }, + { + "epoch": 2.3806315789473684, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 44578 + }, + { + "epoch": 2.380691729323308, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 44579 + }, + { + "epoch": 2.380751879699248, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9343, + "step": 44580 + }, + { + "epoch": 2.380812030075188, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 44581 + }, + { + "epoch": 2.380872180451128, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 44582 + }, + { + "epoch": 2.3809323308270676, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 44583 + }, + { + "epoch": 2.3809924812030077, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 44584 + }, + { + "epoch": 2.3810526315789473, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 44585 + }, + { + "epoch": 2.3811127819548874, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 44586 + }, + { + "epoch": 2.381172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 44587 + }, + { + "epoch": 2.3812330827067667, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 44588 + }, + { + "epoch": 2.381293233082707, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 44589 + }, + { + "epoch": 2.3813533834586464, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 44590 + }, + { + "epoch": 2.3814135338345865, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 44591 + }, + { + "epoch": 2.381473684210526, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 44592 + }, + { + "epoch": 2.3815338345864663, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 44593 + }, + { + "epoch": 2.381593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 44594 + }, + { + "epoch": 2.381654135338346, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 44595 + }, + { + "epoch": 2.3817142857142857, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 44596 + }, + { + "epoch": 2.3817744360902253, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9292, + "step": 44597 + }, + { + "epoch": 2.3818345864661654, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 44598 + }, + { + "epoch": 2.3818947368421055, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 44599 + }, + { + "epoch": 2.381954887218045, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 44600 + }, + { + "epoch": 2.382015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 44601 + }, + { + "epoch": 2.382075187969925, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 44602 + }, + { + "epoch": 2.3821353383458646, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 44603 + }, + { + "epoch": 2.3821954887218046, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 44604 + }, + { + "epoch": 2.3822556390977443, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 44605 + }, + { + "epoch": 2.3823157894736844, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 44606 + }, + { + "epoch": 2.382375939849624, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 44607 + }, + { + "epoch": 2.382436090225564, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 44608 + }, + { + "epoch": 2.382496240601504, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 44609 + }, + { + "epoch": 2.3825563909774434, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 44610 + }, + { + "epoch": 2.3826165413533835, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 44611 + }, + { + "epoch": 2.382676691729323, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 44612 + }, + { + "epoch": 2.3827368421052633, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 44613 + }, + { + "epoch": 2.382796992481203, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.1699, + "step": 44614 + }, + { + "epoch": 2.382857142857143, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 44615 + }, + { + "epoch": 2.3829172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 44616 + }, + { + "epoch": 2.3829774436090227, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 44617 + }, + { + "epoch": 2.3830375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 44618 + }, + { + "epoch": 2.383097744360902, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 44619 + }, + { + "epoch": 2.383157894736842, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 44620 + }, + { + "epoch": 2.383218045112782, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 44621 + }, + { + "epoch": 2.383278195488722, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 44622 + }, + { + "epoch": 2.3833383458646615, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 44623 + }, + { + "epoch": 2.3833984962406016, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 44624 + }, + { + "epoch": 2.3834586466165413, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 44625 + }, + { + "epoch": 2.3835187969924814, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 44626 + }, + { + "epoch": 2.383578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 44627 + }, + { + "epoch": 2.3836390977443607, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 44628 + }, + { + "epoch": 2.3836992481203008, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 44629 + }, + { + "epoch": 2.3837593984962404, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 44630 + }, + { + "epoch": 2.3838195488721805, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 44631 + }, + { + "epoch": 2.38387969924812, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 44632 + }, + { + "epoch": 2.3839398496240602, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 44633 + }, + { + "epoch": 2.384, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 44634 + }, + { + "epoch": 2.38406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 44635 + }, + { + "epoch": 2.3841203007518796, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 44636 + }, + { + "epoch": 2.3841804511278197, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 44637 + }, + { + "epoch": 2.3842406015037594, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 44638 + }, + { + "epoch": 2.3843007518796995, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 44639 + }, + { + "epoch": 2.384360902255639, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 44640 + }, + { + "epoch": 2.3844210526315788, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 44641 + }, + { + "epoch": 2.384481203007519, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 44642 + }, + { + "epoch": 2.3845413533834585, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 44643 + }, + { + "epoch": 2.3846015037593986, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 44644 + }, + { + "epoch": 2.3846616541353383, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 44645 + }, + { + "epoch": 2.3847218045112784, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 44646 + }, + { + "epoch": 2.384781954887218, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 44647 + }, + { + "epoch": 2.384842105263158, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 44648 + }, + { + "epoch": 2.3849022556390977, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 44649 + }, + { + "epoch": 2.3849624060150374, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 44650 + }, + { + "epoch": 2.3850225563909775, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 44651 + }, + { + "epoch": 2.385082706766917, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 44652 + }, + { + "epoch": 2.3851428571428572, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 44653 + }, + { + "epoch": 2.385203007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9479, + "step": 44654 + }, + { + "epoch": 2.385263157894737, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 44655 + }, + { + "epoch": 2.3853233082706766, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 44656 + }, + { + "epoch": 2.3853834586466167, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 44657 + }, + { + "epoch": 2.3854436090225564, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 44658 + }, + { + "epoch": 2.385503759398496, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 44659 + }, + { + "epoch": 2.385563909774436, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 44660 + }, + { + "epoch": 2.3856240601503758, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 44661 + }, + { + "epoch": 2.385684210526316, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 44662 + }, + { + "epoch": 2.3857443609022555, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 44663 + }, + { + "epoch": 2.3858045112781956, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 44664 + }, + { + "epoch": 2.3858646616541352, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 44665 + }, + { + "epoch": 2.3859248120300753, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 44666 + }, + { + "epoch": 2.385984962406015, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 44667 + }, + { + "epoch": 2.386045112781955, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 44668 + }, + { + "epoch": 2.3861052631578947, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 44669 + }, + { + "epoch": 2.386165413533835, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 44670 + }, + { + "epoch": 2.3862255639097745, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 44671 + }, + { + "epoch": 2.386285714285714, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 44672 + }, + { + "epoch": 2.386345864661654, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 44673 + }, + { + "epoch": 2.386406015037594, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 44674 + }, + { + "epoch": 2.386466165413534, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9438, + "step": 44675 + }, + { + "epoch": 2.3865263157894736, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 44676 + }, + { + "epoch": 2.3865864661654137, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 44677 + }, + { + "epoch": 2.3866466165413533, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 44678 + }, + { + "epoch": 2.3867067669172934, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 44679 + }, + { + "epoch": 2.386766917293233, + "grad_norm": 1.5390625, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 44680 + }, + { + "epoch": 2.3868270676691727, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 44681 + }, + { + "epoch": 2.386887218045113, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 44682 + }, + { + "epoch": 2.3869473684210525, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 44683 + }, + { + "epoch": 2.3870075187969926, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 44684 + }, + { + "epoch": 2.3870676691729322, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 44685 + }, + { + "epoch": 2.3871278195488723, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 44686 + }, + { + "epoch": 2.387187969924812, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 44687 + }, + { + "epoch": 2.387248120300752, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 44688 + }, + { + "epoch": 2.3873082706766917, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 44689 + }, + { + "epoch": 2.3873684210526314, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.3344, + "step": 44690 + }, + { + "epoch": 2.3874285714285715, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9275, + "step": 44691 + }, + { + "epoch": 2.387488721804511, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 44692 + }, + { + "epoch": 2.387548872180451, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 44693 + }, + { + "epoch": 2.387609022556391, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 44694 + }, + { + "epoch": 2.387669172932331, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9451, + "step": 44695 + }, + { + "epoch": 2.3877293233082706, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 44696 + }, + { + "epoch": 2.3877894736842107, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 44697 + }, + { + "epoch": 2.3878496240601503, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 44698 + }, + { + "epoch": 2.38790977443609, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 44699 + }, + { + "epoch": 2.38796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 44700 + }, + { + "epoch": 2.38803007518797, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.131, + "step": 44701 + }, + { + "epoch": 2.38809022556391, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 44702 + }, + { + "epoch": 2.3881503759398495, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 44703 + }, + { + "epoch": 2.3882105263157896, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 44704 + }, + { + "epoch": 2.388270676691729, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 44705 + }, + { + "epoch": 2.3883308270676693, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 44706 + }, + { + "epoch": 2.388390977443609, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 44707 + }, + { + "epoch": 2.388451127819549, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 44708 + }, + { + "epoch": 2.3885112781954887, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 44709 + }, + { + "epoch": 2.388571428571429, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 44710 + }, + { + "epoch": 2.3886315789473684, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 44711 + }, + { + "epoch": 2.388691729323308, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 44712 + }, + { + "epoch": 2.388751879699248, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 44713 + }, + { + "epoch": 2.388812030075188, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 44714 + }, + { + "epoch": 2.388872180451128, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 44715 + }, + { + "epoch": 2.3889323308270676, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 44716 + }, + { + "epoch": 2.3889924812030077, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 44717 + }, + { + "epoch": 2.3890526315789473, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 44718 + }, + { + "epoch": 2.3891127819548874, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 44719 + }, + { + "epoch": 2.389172932330827, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 44720 + }, + { + "epoch": 2.3892330827067667, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 44721 + }, + { + "epoch": 2.389293233082707, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 44722 + }, + { + "epoch": 2.3893533834586465, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 44723 + }, + { + "epoch": 2.3894135338345865, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 44724 + }, + { + "epoch": 2.389473684210526, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 44725 + }, + { + "epoch": 2.3895338345864663, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 44726 + }, + { + "epoch": 2.389593984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 44727 + }, + { + "epoch": 2.389654135338346, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 44728 + }, + { + "epoch": 2.3897142857142857, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 44729 + }, + { + "epoch": 2.3897744360902253, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 44730 + }, + { + "epoch": 2.3898345864661654, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 44731 + }, + { + "epoch": 2.389894736842105, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 44732 + }, + { + "epoch": 2.389954887218045, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 44733 + }, + { + "epoch": 2.390015037593985, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 44734 + }, + { + "epoch": 2.390075187969925, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 44735 + }, + { + "epoch": 2.3901353383458646, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 44736 + }, + { + "epoch": 2.3901954887218047, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 44737 + }, + { + "epoch": 2.3902556390977443, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 44738 + }, + { + "epoch": 2.3903157894736844, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 44739 + }, + { + "epoch": 2.390375939849624, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 44740 + }, + { + "epoch": 2.390436090225564, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1534, + "step": 44741 + }, + { + "epoch": 2.390496240601504, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 44742 + }, + { + "epoch": 2.3905563909774434, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 44743 + }, + { + "epoch": 2.3906165413533835, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 44744 + }, + { + "epoch": 2.390676691729323, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 44745 + }, + { + "epoch": 2.3907368421052633, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 44746 + }, + { + "epoch": 2.390796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 44747 + }, + { + "epoch": 2.390857142857143, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1979, + "step": 44748 + }, + { + "epoch": 2.3909172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 44749 + }, + { + "epoch": 2.3909774436090228, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 44750 + }, + { + "epoch": 2.3910375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 44751 + }, + { + "epoch": 2.391097744360902, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 44752 + }, + { + "epoch": 2.391157894736842, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 44753 + }, + { + "epoch": 2.391218045112782, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 44754 + }, + { + "epoch": 2.391278195488722, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 44755 + }, + { + "epoch": 2.3913383458646615, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 44756 + }, + { + "epoch": 2.3913984962406016, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 44757 + }, + { + "epoch": 2.3914586466165413, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 44758 + }, + { + "epoch": 2.3915187969924814, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 44759 + }, + { + "epoch": 2.391578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 44760 + }, + { + "epoch": 2.3916390977443607, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 44761 + }, + { + "epoch": 2.3916992481203008, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9111, + "step": 44762 + }, + { + "epoch": 2.3917593984962404, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 44763 + }, + { + "epoch": 2.3918195488721805, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 44764 + }, + { + "epoch": 2.39187969924812, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 44765 + }, + { + "epoch": 2.3919398496240603, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 44766 + }, + { + "epoch": 2.392, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 44767 + }, + { + "epoch": 2.39206015037594, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 44768 + }, + { + "epoch": 2.3921203007518796, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 44769 + }, + { + "epoch": 2.3921804511278197, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 44770 + }, + { + "epoch": 2.3922406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 44771 + }, + { + "epoch": 2.3923007518796995, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 44772 + }, + { + "epoch": 2.392360902255639, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 44773 + }, + { + "epoch": 2.392421052631579, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 44774 + }, + { + "epoch": 2.392481203007519, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 44775 + }, + { + "epoch": 2.3925413533834585, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 44776 + }, + { + "epoch": 2.3926015037593986, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 44777 + }, + { + "epoch": 2.3926616541353383, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 44778 + }, + { + "epoch": 2.3927218045112784, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 44779 + }, + { + "epoch": 2.392781954887218, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 44780 + }, + { + "epoch": 2.392842105263158, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 44781 + }, + { + "epoch": 2.3929022556390978, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 44782 + }, + { + "epoch": 2.3929624060150374, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 44783 + }, + { + "epoch": 2.3930225563909775, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 44784 + }, + { + "epoch": 2.393082706766917, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 44785 + }, + { + "epoch": 2.3931428571428572, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 44786 + }, + { + "epoch": 2.393203007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 44787 + }, + { + "epoch": 2.393263157894737, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 44788 + }, + { + "epoch": 2.3933233082706766, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 44789 + }, + { + "epoch": 2.3933834586466167, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 44790 + }, + { + "epoch": 2.3934436090225564, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 44791 + }, + { + "epoch": 2.393503759398496, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 44792 + }, + { + "epoch": 2.393563909774436, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 44793 + }, + { + "epoch": 2.3936240601503758, + "grad_norm": 2.515625, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 44794 + }, + { + "epoch": 2.393684210526316, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 44795 + }, + { + "epoch": 2.3937443609022555, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 44796 + }, + { + "epoch": 2.3938045112781956, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 44797 + }, + { + "epoch": 2.3938646616541353, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 44798 + }, + { + "epoch": 2.3939248120300753, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1415, + "step": 44799 + }, + { + "epoch": 2.393984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 44800 + }, + { + "epoch": 2.3940451127819546, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 44801 + }, + { + "epoch": 2.3941052631578947, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 44802 + }, + { + "epoch": 2.394165413533835, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 44803 + }, + { + "epoch": 2.3942255639097745, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 44804 + }, + { + "epoch": 2.394285714285714, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 44805 + }, + { + "epoch": 2.394345864661654, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 44806 + }, + { + "epoch": 2.394406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 44807 + }, + { + "epoch": 2.394466165413534, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 44808 + }, + { + "epoch": 2.3945263157894736, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1084, + "step": 44809 + }, + { + "epoch": 2.3945864661654137, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 44810 + }, + { + "epoch": 2.3946466165413534, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 44811 + }, + { + "epoch": 2.3947067669172935, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 44812 + }, + { + "epoch": 2.394766917293233, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 44813 + }, + { + "epoch": 2.3948270676691727, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 44814 + }, + { + "epoch": 2.394887218045113, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 44815 + }, + { + "epoch": 2.3949473684210525, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.212, + "step": 44816 + }, + { + "epoch": 2.3950075187969926, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 44817 + }, + { + "epoch": 2.3950676691729322, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 44818 + }, + { + "epoch": 2.3951278195488723, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 44819 + }, + { + "epoch": 2.395187969924812, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 44820 + }, + { + "epoch": 2.395248120300752, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 44821 + }, + { + "epoch": 2.3953082706766917, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 44822 + }, + { + "epoch": 2.3953684210526314, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 44823 + }, + { + "epoch": 2.3954285714285715, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 44824 + }, + { + "epoch": 2.395488721804511, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 44825 + }, + { + "epoch": 2.395548872180451, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 44826 + }, + { + "epoch": 2.395609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 44827 + }, + { + "epoch": 2.395669172932331, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9438, + "step": 44828 + }, + { + "epoch": 2.3957293233082706, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 44829 + }, + { + "epoch": 2.3957894736842107, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 44830 + }, + { + "epoch": 2.3958496240601503, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 44831 + }, + { + "epoch": 2.39590977443609, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 44832 + }, + { + "epoch": 2.39596992481203, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 44833 + }, + { + "epoch": 2.3960300751879697, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 44834 + }, + { + "epoch": 2.39609022556391, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 44835 + }, + { + "epoch": 2.3961503759398495, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 44836 + }, + { + "epoch": 2.3962105263157896, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 44837 + }, + { + "epoch": 2.396270676691729, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 44838 + }, + { + "epoch": 2.3963308270676693, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.909, + "step": 44839 + }, + { + "epoch": 2.396390977443609, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 44840 + }, + { + "epoch": 2.396451127819549, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 44841 + }, + { + "epoch": 2.3965112781954887, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 44842 + }, + { + "epoch": 2.396571428571429, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 44843 + }, + { + "epoch": 2.3966315789473684, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9497, + "step": 44844 + }, + { + "epoch": 2.396691729323308, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 44845 + }, + { + "epoch": 2.396751879699248, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 44846 + }, + { + "epoch": 2.396812030075188, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 44847 + }, + { + "epoch": 2.396872180451128, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 44848 + }, + { + "epoch": 2.3969323308270676, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 44849 + }, + { + "epoch": 2.3969924812030077, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 44850 + }, + { + "epoch": 2.3970526315789473, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 44851 + }, + { + "epoch": 2.3971127819548874, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 44852 + }, + { + "epoch": 2.397172932330827, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 44853 + }, + { + "epoch": 2.3972330827067667, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 44854 + }, + { + "epoch": 2.397293233082707, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 44855 + }, + { + "epoch": 2.3973533834586465, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.3166, + "step": 44856 + }, + { + "epoch": 2.3974135338345866, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 44857 + }, + { + "epoch": 2.397473684210526, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 44858 + }, + { + "epoch": 2.3975338345864663, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 44859 + }, + { + "epoch": 2.397593984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 44860 + }, + { + "epoch": 2.397654135338346, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9354, + "step": 44861 + }, + { + "epoch": 2.3977142857142857, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 44862 + }, + { + "epoch": 2.3977744360902253, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 44863 + }, + { + "epoch": 2.3978345864661654, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 44864 + }, + { + "epoch": 2.397894736842105, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 44865 + }, + { + "epoch": 2.397954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.146, + "step": 44866 + }, + { + "epoch": 2.398015037593985, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 44867 + }, + { + "epoch": 2.398075187969925, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 44868 + }, + { + "epoch": 2.3981353383458646, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 44869 + }, + { + "epoch": 2.3981954887218047, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 44870 + }, + { + "epoch": 2.3982556390977443, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 44871 + }, + { + "epoch": 2.3983157894736844, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 44872 + }, + { + "epoch": 2.398375939849624, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 44873 + }, + { + "epoch": 2.398436090225564, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 44874 + }, + { + "epoch": 2.398496240601504, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 44875 + }, + { + "epoch": 2.3985563909774434, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 44876 + }, + { + "epoch": 2.3986165413533835, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 44877 + }, + { + "epoch": 2.398676691729323, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 44878 + }, + { + "epoch": 2.3987368421052633, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 44879 + }, + { + "epoch": 2.398796992481203, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 44880 + }, + { + "epoch": 2.398857142857143, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 44881 + }, + { + "epoch": 2.3989172932330827, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 44882 + }, + { + "epoch": 2.3989774436090228, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 44883 + }, + { + "epoch": 2.3990375939849624, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 44884 + }, + { + "epoch": 2.399097744360902, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 44885 + }, + { + "epoch": 2.399157894736842, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 44886 + }, + { + "epoch": 2.399218045112782, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 44887 + }, + { + "epoch": 2.399278195488722, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 44888 + }, + { + "epoch": 2.3993383458646615, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 44889 + }, + { + "epoch": 2.3993984962406016, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 44890 + }, + { + "epoch": 2.3994586466165413, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 44891 + }, + { + "epoch": 2.3995187969924814, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 44892 + }, + { + "epoch": 2.399578947368421, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 44893 + }, + { + "epoch": 2.3996390977443607, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 44894 + }, + { + "epoch": 2.3996992481203008, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 44895 + }, + { + "epoch": 2.3997593984962404, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 44896 + }, + { + "epoch": 2.3998195488721805, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 44897 + }, + { + "epoch": 2.39987969924812, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 44898 + }, + { + "epoch": 2.3999398496240603, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 44899 + }, + { + "epoch": 2.4, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 44900 + }, + { + "epoch": 2.40006015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 44901 + }, + { + "epoch": 2.4001203007518797, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 44902 + }, + { + "epoch": 2.4001804511278193, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 44903 + }, + { + "epoch": 2.4002406015037594, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 44904 + }, + { + "epoch": 2.4003007518796995, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 44905 + }, + { + "epoch": 2.400360902255639, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 44906 + }, + { + "epoch": 2.400421052631579, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 44907 + }, + { + "epoch": 2.400481203007519, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 44908 + }, + { + "epoch": 2.4005413533834585, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 44909 + }, + { + "epoch": 2.4006015037593986, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 44910 + }, + { + "epoch": 2.4006616541353383, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 44911 + }, + { + "epoch": 2.4007218045112784, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 44912 + }, + { + "epoch": 2.400781954887218, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 44913 + }, + { + "epoch": 2.400842105263158, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 44914 + }, + { + "epoch": 2.4009022556390978, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 44915 + }, + { + "epoch": 2.4009624060150374, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 44916 + }, + { + "epoch": 2.4010225563909775, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2389, + "step": 44917 + }, + { + "epoch": 2.401082706766917, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 44918 + }, + { + "epoch": 2.4011428571428572, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 44919 + }, + { + "epoch": 2.401203007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 44920 + }, + { + "epoch": 2.401263157894737, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 44921 + }, + { + "epoch": 2.4013233082706766, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 44922 + }, + { + "epoch": 2.4013834586466167, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 44923 + }, + { + "epoch": 2.4014436090225564, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1257, + "step": 44924 + }, + { + "epoch": 2.401503759398496, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 44925 + }, + { + "epoch": 2.401563909774436, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 44926 + }, + { + "epoch": 2.4016240601503758, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 44927 + }, + { + "epoch": 2.401684210526316, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 44928 + }, + { + "epoch": 2.4017443609022555, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 44929 + }, + { + "epoch": 2.4018045112781956, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 44930 + }, + { + "epoch": 2.4018646616541353, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 44931 + }, + { + "epoch": 2.4019248120300754, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 44932 + }, + { + "epoch": 2.401984962406015, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 44933 + }, + { + "epoch": 2.4020451127819546, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 44934 + }, + { + "epoch": 2.4021052631578947, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 44935 + }, + { + "epoch": 2.4021654135338344, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 44936 + }, + { + "epoch": 2.4022255639097745, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 44937 + }, + { + "epoch": 2.402285714285714, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 44938 + }, + { + "epoch": 2.4023458646616542, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 44939 + }, + { + "epoch": 2.402406015037594, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 44940 + }, + { + "epoch": 2.402466165413534, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 44941 + }, + { + "epoch": 2.4025263157894736, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 44942 + }, + { + "epoch": 2.4025864661654137, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 44943 + }, + { + "epoch": 2.4026466165413534, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 44944 + }, + { + "epoch": 2.4027067669172935, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 44945 + }, + { + "epoch": 2.402766917293233, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 44946 + }, + { + "epoch": 2.4028270676691728, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 44947 + }, + { + "epoch": 2.402887218045113, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 44948 + }, + { + "epoch": 2.4029473684210525, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 44949 + }, + { + "epoch": 2.4030075187969926, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 44950 + }, + { + "epoch": 2.4030676691729322, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 44951 + }, + { + "epoch": 2.4031278195488723, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 44952 + }, + { + "epoch": 2.403187969924812, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 44953 + }, + { + "epoch": 2.403248120300752, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 44954 + }, + { + "epoch": 2.4033082706766917, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 44955 + }, + { + "epoch": 2.4033684210526314, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 44956 + }, + { + "epoch": 2.4034285714285715, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 44957 + }, + { + "epoch": 2.403488721804511, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9516, + "step": 44958 + }, + { + "epoch": 2.403548872180451, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 44959 + }, + { + "epoch": 2.403609022556391, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 44960 + }, + { + "epoch": 2.403669172932331, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 44961 + }, + { + "epoch": 2.4037293233082706, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 44962 + }, + { + "epoch": 2.4037894736842107, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 44963 + }, + { + "epoch": 2.4038496240601503, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 44964 + }, + { + "epoch": 2.40390977443609, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 44965 + }, + { + "epoch": 2.40396992481203, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 44966 + }, + { + "epoch": 2.4040300751879697, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 44967 + }, + { + "epoch": 2.40409022556391, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 44968 + }, + { + "epoch": 2.4041503759398495, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 44969 + }, + { + "epoch": 2.4042105263157896, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 44970 + }, + { + "epoch": 2.4042706766917292, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 44971 + }, + { + "epoch": 2.4043308270676693, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 44972 + }, + { + "epoch": 2.404390977443609, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 44973 + }, + { + "epoch": 2.404451127819549, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 44974 + }, + { + "epoch": 2.4045112781954887, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 44975 + }, + { + "epoch": 2.404571428571429, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 44976 + }, + { + "epoch": 2.4046315789473685, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 44977 + }, + { + "epoch": 2.404691729323308, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 44978 + }, + { + "epoch": 2.404751879699248, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 44979 + }, + { + "epoch": 2.404812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 44980 + }, + { + "epoch": 2.404872180451128, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 44981 + }, + { + "epoch": 2.4049323308270676, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 44982 + }, + { + "epoch": 2.4049924812030077, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 44983 + }, + { + "epoch": 2.4050526315789473, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 44984 + }, + { + "epoch": 2.4051127819548874, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 44985 + }, + { + "epoch": 2.405172932330827, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 44986 + }, + { + "epoch": 2.4052330827067667, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 44987 + }, + { + "epoch": 2.405293233082707, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 44988 + }, + { + "epoch": 2.4053533834586465, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 44989 + }, + { + "epoch": 2.4054135338345866, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 44990 + }, + { + "epoch": 2.405473684210526, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 44991 + }, + { + "epoch": 2.4055338345864663, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 44992 + }, + { + "epoch": 2.405593984962406, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 44993 + }, + { + "epoch": 2.405654135338346, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9248, + "step": 44994 + }, + { + "epoch": 2.4057142857142857, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 44995 + }, + { + "epoch": 2.4057744360902253, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 44996 + }, + { + "epoch": 2.4058345864661654, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 44997 + }, + { + "epoch": 2.405894736842105, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 44998 + }, + { + "epoch": 2.405954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9453, + "step": 44999 + }, + { + "epoch": 2.406015037593985, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 45000 + }, + { + "epoch": 2.406075187969925, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1223, + "step": 45001 + }, + { + "epoch": 2.4061353383458646, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.8378, + "step": 45002 + }, + { + "epoch": 2.4061954887218047, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 45003 + }, + { + "epoch": 2.4062556390977443, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 45004 + }, + { + "epoch": 2.406315789473684, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 45005 + }, + { + "epoch": 2.406375939849624, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 45006 + }, + { + "epoch": 2.406436090225564, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 45007 + }, + { + "epoch": 2.406496240601504, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 45008 + }, + { + "epoch": 2.4065563909774434, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 45009 + }, + { + "epoch": 2.4066165413533835, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2383, + "step": 45010 + }, + { + "epoch": 2.406676691729323, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 45011 + }, + { + "epoch": 2.4067368421052633, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 45012 + }, + { + "epoch": 2.406796992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9511, + "step": 45013 + }, + { + "epoch": 2.406857142857143, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 45014 + }, + { + "epoch": 2.4069172932330827, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 45015 + }, + { + "epoch": 2.4069774436090228, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 45016 + }, + { + "epoch": 2.4070375939849624, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 45017 + }, + { + "epoch": 2.407097744360902, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 45018 + }, + { + "epoch": 2.407157894736842, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 45019 + }, + { + "epoch": 2.407218045112782, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 45020 + }, + { + "epoch": 2.407278195488722, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 45021 + }, + { + "epoch": 2.4073383458646616, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 45022 + }, + { + "epoch": 2.4073984962406016, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 45023 + }, + { + "epoch": 2.4074586466165413, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 45024 + }, + { + "epoch": 2.4075187969924814, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 45025 + }, + { + "epoch": 2.407578947368421, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 45026 + }, + { + "epoch": 2.4076390977443607, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 45027 + }, + { + "epoch": 2.407699248120301, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 45028 + }, + { + "epoch": 2.4077593984962404, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 45029 + }, + { + "epoch": 2.4078195488721805, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 45030 + }, + { + "epoch": 2.40787969924812, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 45031 + }, + { + "epoch": 2.4079398496240603, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 45032 + }, + { + "epoch": 2.408, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 45033 + }, + { + "epoch": 2.40806015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 45034 + }, + { + "epoch": 2.4081203007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 45035 + }, + { + "epoch": 2.4081804511278193, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 45036 + }, + { + "epoch": 2.4082406015037594, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 45037 + }, + { + "epoch": 2.4083007518796995, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 45038 + }, + { + "epoch": 2.408360902255639, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 45039 + }, + { + "epoch": 2.408421052631579, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 45040 + }, + { + "epoch": 2.408481203007519, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 45041 + }, + { + "epoch": 2.4085413533834585, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 45042 + }, + { + "epoch": 2.4086015037593986, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 45043 + }, + { + "epoch": 2.4086616541353383, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 45044 + }, + { + "epoch": 2.4087218045112784, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 45045 + }, + { + "epoch": 2.408781954887218, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 45046 + }, + { + "epoch": 2.408842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 45047 + }, + { + "epoch": 2.4089022556390978, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 45048 + }, + { + "epoch": 2.4089624060150374, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1343, + "step": 45049 + }, + { + "epoch": 2.4090225563909775, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 45050 + }, + { + "epoch": 2.409082706766917, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 45051 + }, + { + "epoch": 2.4091428571428573, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 45052 + }, + { + "epoch": 2.409203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 45053 + }, + { + "epoch": 2.409263157894737, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 45054 + }, + { + "epoch": 2.4093233082706766, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 45055 + }, + { + "epoch": 2.4093834586466167, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 45056 + }, + { + "epoch": 2.4094436090225564, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 45057 + }, + { + "epoch": 2.409503759398496, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 45058 + }, + { + "epoch": 2.409563909774436, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 45059 + }, + { + "epoch": 2.409624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 45060 + }, + { + "epoch": 2.409684210526316, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 45061 + }, + { + "epoch": 2.4097443609022555, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9067, + "step": 45062 + }, + { + "epoch": 2.4098045112781956, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 45063 + }, + { + "epoch": 2.4098646616541353, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 45064 + }, + { + "epoch": 2.4099248120300754, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 45065 + }, + { + "epoch": 2.409984962406015, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9514, + "step": 45066 + }, + { + "epoch": 2.4100451127819547, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9256, + "step": 45067 + }, + { + "epoch": 2.4101052631578948, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 45068 + }, + { + "epoch": 2.4101654135338344, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 45069 + }, + { + "epoch": 2.4102255639097745, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 45070 + }, + { + "epoch": 2.410285714285714, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 45071 + }, + { + "epoch": 2.4103458646616542, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 45072 + }, + { + "epoch": 2.410406015037594, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 45073 + }, + { + "epoch": 2.410466165413534, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 45074 + }, + { + "epoch": 2.4105263157894736, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 45075 + }, + { + "epoch": 2.4105864661654137, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 45076 + }, + { + "epoch": 2.4106466165413534, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 45077 + }, + { + "epoch": 2.4107067669172935, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 45078 + }, + { + "epoch": 2.410766917293233, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 45079 + }, + { + "epoch": 2.4108270676691728, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 45080 + }, + { + "epoch": 2.410887218045113, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 45081 + }, + { + "epoch": 2.4109473684210525, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 45082 + }, + { + "epoch": 2.4110075187969926, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 45083 + }, + { + "epoch": 2.4110676691729322, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 45084 + }, + { + "epoch": 2.4111278195488723, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 45085 + }, + { + "epoch": 2.411187969924812, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9192, + "step": 45086 + }, + { + "epoch": 2.411248120300752, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 45087 + }, + { + "epoch": 2.4113082706766917, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 45088 + }, + { + "epoch": 2.4113684210526314, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 45089 + }, + { + "epoch": 2.4114285714285715, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 45090 + }, + { + "epoch": 2.411488721804511, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 45091 + }, + { + "epoch": 2.411548872180451, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 45092 + }, + { + "epoch": 2.411609022556391, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 45093 + }, + { + "epoch": 2.411669172932331, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 45094 + }, + { + "epoch": 2.4117293233082706, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 45095 + }, + { + "epoch": 2.4117894736842107, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 45096 + }, + { + "epoch": 2.4118496240601504, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 45097 + }, + { + "epoch": 2.41190977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 45098 + }, + { + "epoch": 2.41196992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 45099 + }, + { + "epoch": 2.4120300751879697, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 45100 + }, + { + "epoch": 2.41209022556391, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9427, + "step": 45101 + }, + { + "epoch": 2.4121503759398495, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 45102 + }, + { + "epoch": 2.4122105263157896, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 45103 + }, + { + "epoch": 2.4122706766917292, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 45104 + }, + { + "epoch": 2.4123308270676693, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 45105 + }, + { + "epoch": 2.412390977443609, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 45106 + }, + { + "epoch": 2.4124511278195486, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 45107 + }, + { + "epoch": 2.4125112781954887, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 45108 + }, + { + "epoch": 2.412571428571429, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 45109 + }, + { + "epoch": 2.4126315789473685, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 45110 + }, + { + "epoch": 2.412691729323308, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 45111 + }, + { + "epoch": 2.412751879699248, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 45112 + }, + { + "epoch": 2.412812030075188, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 45113 + }, + { + "epoch": 2.412872180451128, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 45114 + }, + { + "epoch": 2.4129323308270676, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 45115 + }, + { + "epoch": 2.4129924812030077, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 45116 + }, + { + "epoch": 2.4130526315789473, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 45117 + }, + { + "epoch": 2.4131127819548874, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 45118 + }, + { + "epoch": 2.413172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 45119 + }, + { + "epoch": 2.4132330827067667, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 45120 + }, + { + "epoch": 2.413293233082707, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 45121 + }, + { + "epoch": 2.4133533834586465, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 45122 + }, + { + "epoch": 2.4134135338345866, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 45123 + }, + { + "epoch": 2.413473684210526, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 45124 + }, + { + "epoch": 2.4135338345864663, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 45125 + }, + { + "epoch": 2.413593984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 45126 + }, + { + "epoch": 2.413654135338346, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 45127 + }, + { + "epoch": 2.4137142857142857, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 45128 + }, + { + "epoch": 2.4137744360902254, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 45129 + }, + { + "epoch": 2.4138345864661654, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 45130 + }, + { + "epoch": 2.413894736842105, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 45131 + }, + { + "epoch": 2.413954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 45132 + }, + { + "epoch": 2.414015037593985, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 45133 + }, + { + "epoch": 2.414075187969925, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 45134 + }, + { + "epoch": 2.4141353383458646, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 45135 + }, + { + "epoch": 2.4141954887218047, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 45136 + }, + { + "epoch": 2.4142556390977443, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 45137 + }, + { + "epoch": 2.414315789473684, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 45138 + }, + { + "epoch": 2.414375939849624, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 45139 + }, + { + "epoch": 2.414436090225564, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 45140 + }, + { + "epoch": 2.414496240601504, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1653, + "step": 45141 + }, + { + "epoch": 2.4145563909774435, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 45142 + }, + { + "epoch": 2.4146165413533835, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 45143 + }, + { + "epoch": 2.414676691729323, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 45144 + }, + { + "epoch": 2.4147368421052633, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.8998, + "step": 45145 + }, + { + "epoch": 2.414796992481203, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 45146 + }, + { + "epoch": 2.414857142857143, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1925, + "step": 45147 + }, + { + "epoch": 2.4149172932330827, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 45148 + }, + { + "epoch": 2.414977443609023, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 45149 + }, + { + "epoch": 2.4150375939849624, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 45150 + }, + { + "epoch": 2.415097744360902, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.142, + "step": 45151 + }, + { + "epoch": 2.415157894736842, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 45152 + }, + { + "epoch": 2.415218045112782, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 45153 + }, + { + "epoch": 2.415278195488722, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 45154 + }, + { + "epoch": 2.4153383458646616, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 45155 + }, + { + "epoch": 2.4153984962406017, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 45156 + }, + { + "epoch": 2.4154586466165413, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 45157 + }, + { + "epoch": 2.4155187969924814, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 45158 + }, + { + "epoch": 2.415578947368421, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 45159 + }, + { + "epoch": 2.4156390977443607, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.2117, + "step": 45160 + }, + { + "epoch": 2.415699248120301, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 45161 + }, + { + "epoch": 2.4157593984962404, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 45162 + }, + { + "epoch": 2.4158195488721805, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 45163 + }, + { + "epoch": 2.41587969924812, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 45164 + }, + { + "epoch": 2.4159398496240603, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 45165 + }, + { + "epoch": 2.416, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9163, + "step": 45166 + }, + { + "epoch": 2.41606015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 45167 + }, + { + "epoch": 2.4161203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 45168 + }, + { + "epoch": 2.4161804511278193, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.919, + "step": 45169 + }, + { + "epoch": 2.4162406015037594, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 45170 + }, + { + "epoch": 2.416300751879699, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 45171 + }, + { + "epoch": 2.416360902255639, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 45172 + }, + { + "epoch": 2.416421052631579, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 45173 + }, + { + "epoch": 2.416481203007519, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 45174 + }, + { + "epoch": 2.4165413533834585, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9308, + "step": 45175 + }, + { + "epoch": 2.4166015037593986, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 45176 + }, + { + "epoch": 2.4166616541353383, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 45177 + }, + { + "epoch": 2.4167218045112784, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 45178 + }, + { + "epoch": 2.416781954887218, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 45179 + }, + { + "epoch": 2.416842105263158, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9267, + "step": 45180 + }, + { + "epoch": 2.4169022556390978, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 45181 + }, + { + "epoch": 2.4169624060150374, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 45182 + }, + { + "epoch": 2.4170225563909775, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 45183 + }, + { + "epoch": 2.417082706766917, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 45184 + }, + { + "epoch": 2.4171428571428573, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 45185 + }, + { + "epoch": 2.417203007518797, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 45186 + }, + { + "epoch": 2.417263157894737, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 45187 + }, + { + "epoch": 2.4173233082706767, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 45188 + }, + { + "epoch": 2.4173834586466167, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 45189 + }, + { + "epoch": 2.4174436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 45190 + }, + { + "epoch": 2.417503759398496, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 45191 + }, + { + "epoch": 2.417563909774436, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 45192 + }, + { + "epoch": 2.417624060150376, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 45193 + }, + { + "epoch": 2.417684210526316, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 45194 + }, + { + "epoch": 2.4177443609022555, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9305, + "step": 45195 + }, + { + "epoch": 2.4178045112781956, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 45196 + }, + { + "epoch": 2.4178646616541353, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 45197 + }, + { + "epoch": 2.4179248120300754, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 45198 + }, + { + "epoch": 2.417984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 45199 + }, + { + "epoch": 2.4180451127819547, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 45200 + }, + { + "epoch": 2.4181052631578948, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 45201 + }, + { + "epoch": 2.4181654135338344, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 45202 + }, + { + "epoch": 2.4182255639097745, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 45203 + }, + { + "epoch": 2.418285714285714, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2069, + "step": 45204 + }, + { + "epoch": 2.4183458646616542, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 45205 + }, + { + "epoch": 2.418406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 45206 + }, + { + "epoch": 2.418466165413534, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 45207 + }, + { + "epoch": 2.4185263157894736, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9297, + "step": 45208 + }, + { + "epoch": 2.4185864661654133, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 45209 + }, + { + "epoch": 2.4186466165413534, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1287, + "step": 45210 + }, + { + "epoch": 2.4187067669172935, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 45211 + }, + { + "epoch": 2.418766917293233, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 45212 + }, + { + "epoch": 2.4188270676691728, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 45213 + }, + { + "epoch": 2.418887218045113, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 45214 + }, + { + "epoch": 2.4189473684210525, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 45215 + }, + { + "epoch": 2.4190075187969926, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 45216 + }, + { + "epoch": 2.4190676691729323, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 45217 + }, + { + "epoch": 2.4191278195488723, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 45218 + }, + { + "epoch": 2.419187969924812, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9371, + "step": 45219 + }, + { + "epoch": 2.419248120300752, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 45220 + }, + { + "epoch": 2.4193082706766917, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 45221 + }, + { + "epoch": 2.4193684210526314, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 45222 + }, + { + "epoch": 2.4194285714285715, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 45223 + }, + { + "epoch": 2.419488721804511, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 45224 + }, + { + "epoch": 2.4195488721804512, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 45225 + }, + { + "epoch": 2.419609022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 45226 + }, + { + "epoch": 2.419669172932331, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 45227 + }, + { + "epoch": 2.4197293233082706, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 45228 + }, + { + "epoch": 2.4197894736842107, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 45229 + }, + { + "epoch": 2.4198496240601504, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 45230 + }, + { + "epoch": 2.41990977443609, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 45231 + }, + { + "epoch": 2.41996992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 45232 + }, + { + "epoch": 2.4200300751879698, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 45233 + }, + { + "epoch": 2.42009022556391, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 45234 + }, + { + "epoch": 2.4201503759398495, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 45235 + }, + { + "epoch": 2.4202105263157896, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 45236 + }, + { + "epoch": 2.4202706766917292, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 45237 + }, + { + "epoch": 2.4203308270676693, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 45238 + }, + { + "epoch": 2.420390977443609, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1888, + "step": 45239 + }, + { + "epoch": 2.4204511278195486, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 45240 + }, + { + "epoch": 2.4205112781954887, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 45241 + }, + { + "epoch": 2.420571428571429, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 45242 + }, + { + "epoch": 2.4206315789473685, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 45243 + }, + { + "epoch": 2.420691729323308, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 45244 + }, + { + "epoch": 2.420751879699248, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 45245 + }, + { + "epoch": 2.420812030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 45246 + }, + { + "epoch": 2.420872180451128, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 45247 + }, + { + "epoch": 2.4209323308270676, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 45248 + }, + { + "epoch": 2.4209924812030077, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 45249 + }, + { + "epoch": 2.4210526315789473, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 45250 + }, + { + "epoch": 2.4211127819548874, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 45251 + }, + { + "epoch": 2.421172932330827, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 45252 + }, + { + "epoch": 2.4212330827067667, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 45253 + }, + { + "epoch": 2.421293233082707, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 45254 + }, + { + "epoch": 2.4213533834586465, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 45255 + }, + { + "epoch": 2.4214135338345866, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 45256 + }, + { + "epoch": 2.4214736842105262, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 45257 + }, + { + "epoch": 2.4215338345864663, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 45258 + }, + { + "epoch": 2.421593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 45259 + }, + { + "epoch": 2.421654135338346, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 45260 + }, + { + "epoch": 2.4217142857142857, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 45261 + }, + { + "epoch": 2.4217744360902254, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 45262 + }, + { + "epoch": 2.4218345864661655, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 45263 + }, + { + "epoch": 2.421894736842105, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 45264 + }, + { + "epoch": 2.421954887218045, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 45265 + }, + { + "epoch": 2.422015037593985, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 45266 + }, + { + "epoch": 2.422075187969925, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 45267 + }, + { + "epoch": 2.4221353383458646, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 45268 + }, + { + "epoch": 2.4221954887218047, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 45269 + }, + { + "epoch": 2.4222556390977443, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1881, + "step": 45270 + }, + { + "epoch": 2.422315789473684, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 45271 + }, + { + "epoch": 2.422375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 45272 + }, + { + "epoch": 2.4224360902255637, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 45273 + }, + { + "epoch": 2.422496240601504, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 45274 + }, + { + "epoch": 2.4225563909774435, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 45275 + }, + { + "epoch": 2.4226165413533836, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 45276 + }, + { + "epoch": 2.422676691729323, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 45277 + }, + { + "epoch": 2.4227368421052633, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 45278 + }, + { + "epoch": 2.422796992481203, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 45279 + }, + { + "epoch": 2.422857142857143, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 45280 + }, + { + "epoch": 2.4229172932330827, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 45281 + }, + { + "epoch": 2.422977443609023, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 45282 + }, + { + "epoch": 2.4230375939849624, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 45283 + }, + { + "epoch": 2.423097744360902, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 45284 + }, + { + "epoch": 2.423157894736842, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 45285 + }, + { + "epoch": 2.423218045112782, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 45286 + }, + { + "epoch": 2.423278195488722, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 45287 + }, + { + "epoch": 2.4233383458646616, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 45288 + }, + { + "epoch": 2.4233984962406017, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 45289 + }, + { + "epoch": 2.4234586466165413, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 45290 + }, + { + "epoch": 2.4235187969924814, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 45291 + }, + { + "epoch": 2.423578947368421, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 45292 + }, + { + "epoch": 2.4236390977443607, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 45293 + }, + { + "epoch": 2.423699248120301, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 45294 + }, + { + "epoch": 2.4237593984962404, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.934, + "step": 45295 + }, + { + "epoch": 2.4238195488721805, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2087, + "step": 45296 + }, + { + "epoch": 2.42387969924812, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 45297 + }, + { + "epoch": 2.4239398496240603, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 45298 + }, + { + "epoch": 2.424, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 45299 + }, + { + "epoch": 2.42406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 45300 + }, + { + "epoch": 2.4241203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 45301 + }, + { + "epoch": 2.4241804511278193, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 45302 + }, + { + "epoch": 2.4242406015037594, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 45303 + }, + { + "epoch": 2.424300751879699, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 45304 + }, + { + "epoch": 2.424360902255639, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 45305 + }, + { + "epoch": 2.424421052631579, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 45306 + }, + { + "epoch": 2.424481203007519, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 45307 + }, + { + "epoch": 2.4245413533834586, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 45308 + }, + { + "epoch": 2.4246015037593986, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 45309 + }, + { + "epoch": 2.4246616541353383, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 45310 + }, + { + "epoch": 2.4247218045112784, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 45311 + }, + { + "epoch": 2.424781954887218, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 45312 + }, + { + "epoch": 2.424842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 45313 + }, + { + "epoch": 2.424902255639098, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 45314 + }, + { + "epoch": 2.4249624060150374, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 45315 + }, + { + "epoch": 2.4250225563909775, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 45316 + }, + { + "epoch": 2.425082706766917, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 45317 + }, + { + "epoch": 2.4251428571428573, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 45318 + }, + { + "epoch": 2.425203007518797, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 45319 + }, + { + "epoch": 2.425263157894737, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9371, + "step": 45320 + }, + { + "epoch": 2.4253233082706767, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 45321 + }, + { + "epoch": 2.4253834586466168, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 45322 + }, + { + "epoch": 2.4254436090225564, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 45323 + }, + { + "epoch": 2.425503759398496, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 45324 + }, + { + "epoch": 2.425563909774436, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 45325 + }, + { + "epoch": 2.425624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 45326 + }, + { + "epoch": 2.425684210526316, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 45327 + }, + { + "epoch": 2.4257443609022555, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 45328 + }, + { + "epoch": 2.4258045112781956, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 45329 + }, + { + "epoch": 2.4258646616541353, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 45330 + }, + { + "epoch": 2.4259248120300754, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 45331 + }, + { + "epoch": 2.425984962406015, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 45332 + }, + { + "epoch": 2.4260451127819547, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 45333 + }, + { + "epoch": 2.4261052631578948, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 45334 + }, + { + "epoch": 2.4261654135338344, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 45335 + }, + { + "epoch": 2.4262255639097745, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 45336 + }, + { + "epoch": 2.426285714285714, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1022, + "step": 45337 + }, + { + "epoch": 2.4263458646616543, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 45338 + }, + { + "epoch": 2.426406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 45339 + }, + { + "epoch": 2.426466165413534, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 45340 + }, + { + "epoch": 2.4265263157894736, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9332, + "step": 45341 + }, + { + "epoch": 2.4265864661654133, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 45342 + }, + { + "epoch": 2.4266466165413534, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 45343 + }, + { + "epoch": 2.4267067669172935, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 45344 + }, + { + "epoch": 2.426766917293233, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 45345 + }, + { + "epoch": 2.4268270676691728, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 45346 + }, + { + "epoch": 2.426887218045113, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 45347 + }, + { + "epoch": 2.4269473684210525, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 45348 + }, + { + "epoch": 2.4270075187969926, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 45349 + }, + { + "epoch": 2.4270676691729323, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 45350 + }, + { + "epoch": 2.4271278195488724, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1787, + "step": 45351 + }, + { + "epoch": 2.427187969924812, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 45352 + }, + { + "epoch": 2.427248120300752, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 45353 + }, + { + "epoch": 2.4273082706766917, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9406, + "step": 45354 + }, + { + "epoch": 2.4273684210526314, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 45355 + }, + { + "epoch": 2.4274285714285715, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 45356 + }, + { + "epoch": 2.427488721804511, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 45357 + }, + { + "epoch": 2.4275488721804512, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 45358 + }, + { + "epoch": 2.427609022556391, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 45359 + }, + { + "epoch": 2.427669172932331, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 45360 + }, + { + "epoch": 2.4277293233082706, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 45361 + }, + { + "epoch": 2.4277894736842107, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 45362 + }, + { + "epoch": 2.4278496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 45363 + }, + { + "epoch": 2.42790977443609, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 45364 + }, + { + "epoch": 2.42796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 45365 + }, + { + "epoch": 2.4280300751879698, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 45366 + }, + { + "epoch": 2.42809022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 45367 + }, + { + "epoch": 2.4281503759398495, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 45368 + }, + { + "epoch": 2.4282105263157896, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 45369 + }, + { + "epoch": 2.4282706766917292, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 45370 + }, + { + "epoch": 2.4283308270676693, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 45371 + }, + { + "epoch": 2.428390977443609, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 45372 + }, + { + "epoch": 2.4284511278195486, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1292, + "step": 45373 + }, + { + "epoch": 2.4285112781954887, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 45374 + }, + { + "epoch": 2.4285714285714284, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 45375 + }, + { + "epoch": 2.4286315789473685, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 45376 + }, + { + "epoch": 2.428691729323308, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 45377 + }, + { + "epoch": 2.428751879699248, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 45378 + }, + { + "epoch": 2.428812030075188, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9456, + "step": 45379 + }, + { + "epoch": 2.428872180451128, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 45380 + }, + { + "epoch": 2.4289323308270676, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 45381 + }, + { + "epoch": 2.4289924812030077, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 45382 + }, + { + "epoch": 2.4290526315789474, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 45383 + }, + { + "epoch": 2.4291127819548874, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 45384 + }, + { + "epoch": 2.429172932330827, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 45385 + }, + { + "epoch": 2.4292330827067667, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 45386 + }, + { + "epoch": 2.429293233082707, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 45387 + }, + { + "epoch": 2.4293533834586465, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 45388 + }, + { + "epoch": 2.4294135338345866, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 45389 + }, + { + "epoch": 2.4294736842105262, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 45390 + }, + { + "epoch": 2.4295338345864663, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 45391 + }, + { + "epoch": 2.429593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 45392 + }, + { + "epoch": 2.429654135338346, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 45393 + }, + { + "epoch": 2.4297142857142857, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 45394 + }, + { + "epoch": 2.4297744360902254, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 45395 + }, + { + "epoch": 2.4298345864661655, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 45396 + }, + { + "epoch": 2.429894736842105, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 45397 + }, + { + "epoch": 2.429954887218045, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 45398 + }, + { + "epoch": 2.430015037593985, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 45399 + }, + { + "epoch": 2.430075187969925, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 45400 + }, + { + "epoch": 2.4301353383458646, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 45401 + }, + { + "epoch": 2.4301954887218047, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 45402 + }, + { + "epoch": 2.4302556390977443, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 45403 + }, + { + "epoch": 2.430315789473684, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.2219, + "step": 45404 + }, + { + "epoch": 2.430375939849624, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 45405 + }, + { + "epoch": 2.4304360902255637, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 45406 + }, + { + "epoch": 2.430496240601504, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 45407 + }, + { + "epoch": 2.4305563909774435, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1419, + "step": 45408 + }, + { + "epoch": 2.4306165413533836, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1821, + "step": 45409 + }, + { + "epoch": 2.430676691729323, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 45410 + }, + { + "epoch": 2.4307368421052633, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 45411 + }, + { + "epoch": 2.430796992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9389, + "step": 45412 + }, + { + "epoch": 2.430857142857143, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 45413 + }, + { + "epoch": 2.4309172932330827, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 45414 + }, + { + "epoch": 2.430977443609023, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 45415 + }, + { + "epoch": 2.4310375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 45416 + }, + { + "epoch": 2.431097744360902, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 45417 + }, + { + "epoch": 2.431157894736842, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 45418 + }, + { + "epoch": 2.431218045112782, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 45419 + }, + { + "epoch": 2.431278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 45420 + }, + { + "epoch": 2.4313383458646616, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9224, + "step": 45421 + }, + { + "epoch": 2.4313984962406017, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 45422 + }, + { + "epoch": 2.4314586466165413, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 45423 + }, + { + "epoch": 2.4315187969924814, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 45424 + }, + { + "epoch": 2.431578947368421, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 45425 + }, + { + "epoch": 2.4316390977443607, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 45426 + }, + { + "epoch": 2.431699248120301, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 45427 + }, + { + "epoch": 2.4317593984962405, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 45428 + }, + { + "epoch": 2.4318195488721805, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 45429 + }, + { + "epoch": 2.43187969924812, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 45430 + }, + { + "epoch": 2.4319398496240603, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 45431 + }, + { + "epoch": 2.432, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 45432 + }, + { + "epoch": 2.43206015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 45433 + }, + { + "epoch": 2.4321203007518797, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 45434 + }, + { + "epoch": 2.4321804511278193, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 45435 + }, + { + "epoch": 2.4322406015037594, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 45436 + }, + { + "epoch": 2.432300751879699, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 45437 + }, + { + "epoch": 2.432360902255639, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 45438 + }, + { + "epoch": 2.432421052631579, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 45439 + }, + { + "epoch": 2.432481203007519, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 45440 + }, + { + "epoch": 2.4325413533834586, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 45441 + }, + { + "epoch": 2.4326015037593987, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 45442 + }, + { + "epoch": 2.4326616541353383, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 45443 + }, + { + "epoch": 2.432721804511278, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1692, + "step": 45444 + }, + { + "epoch": 2.432781954887218, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 45445 + }, + { + "epoch": 2.432842105263158, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 45446 + }, + { + "epoch": 2.432902255639098, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 45447 + }, + { + "epoch": 2.4329624060150374, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 45448 + }, + { + "epoch": 2.4330225563909775, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 45449 + }, + { + "epoch": 2.433082706766917, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1314, + "step": 45450 + }, + { + "epoch": 2.4331428571428573, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 45451 + }, + { + "epoch": 2.433203007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 45452 + }, + { + "epoch": 2.433263157894737, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 45453 + }, + { + "epoch": 2.4333233082706767, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 45454 + }, + { + "epoch": 2.4333834586466168, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 45455 + }, + { + "epoch": 2.4334436090225564, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 45456 + }, + { + "epoch": 2.433503759398496, + "grad_norm": 1.484375, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 45457 + }, + { + "epoch": 2.433563909774436, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 45458 + }, + { + "epoch": 2.433624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 45459 + }, + { + "epoch": 2.433684210526316, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 45460 + }, + { + "epoch": 2.4337443609022555, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 45461 + }, + { + "epoch": 2.4338045112781956, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 45462 + }, + { + "epoch": 2.4338646616541353, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 45463 + }, + { + "epoch": 2.4339248120300754, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.2126, + "step": 45464 + }, + { + "epoch": 2.433984962406015, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 45465 + }, + { + "epoch": 2.4340451127819547, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 45466 + }, + { + "epoch": 2.4341052631578948, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 45467 + }, + { + "epoch": 2.4341654135338344, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 45468 + }, + { + "epoch": 2.4342255639097745, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 45469 + }, + { + "epoch": 2.434285714285714, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 45470 + }, + { + "epoch": 2.4343458646616543, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 45471 + }, + { + "epoch": 2.434406015037594, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 45472 + }, + { + "epoch": 2.434466165413534, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 45473 + }, + { + "epoch": 2.4345263157894736, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 45474 + }, + { + "epoch": 2.4345864661654133, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 45475 + }, + { + "epoch": 2.4346466165413534, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 45476 + }, + { + "epoch": 2.434706766917293, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 45477 + }, + { + "epoch": 2.434766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 45478 + }, + { + "epoch": 2.434827067669173, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 45479 + }, + { + "epoch": 2.434887218045113, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 45480 + }, + { + "epoch": 2.4349473684210525, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 45481 + }, + { + "epoch": 2.4350075187969926, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 45482 + }, + { + "epoch": 2.4350676691729323, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1431, + "step": 45483 + }, + { + "epoch": 2.4351278195488724, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 45484 + }, + { + "epoch": 2.435187969924812, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 45485 + }, + { + "epoch": 2.435248120300752, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 45486 + }, + { + "epoch": 2.4353082706766918, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 45487 + }, + { + "epoch": 2.4353684210526314, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 0.938, + "step": 45488 + }, + { + "epoch": 2.4354285714285715, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 45489 + }, + { + "epoch": 2.435488721804511, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 45490 + }, + { + "epoch": 2.4355488721804512, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 45491 + }, + { + "epoch": 2.435609022556391, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 45492 + }, + { + "epoch": 2.435669172932331, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 45493 + }, + { + "epoch": 2.4357293233082706, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 45494 + }, + { + "epoch": 2.4357894736842107, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 45495 + }, + { + "epoch": 2.4358496240601504, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 45496 + }, + { + "epoch": 2.43590977443609, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 45497 + }, + { + "epoch": 2.43596992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 45498 + }, + { + "epoch": 2.4360300751879698, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 45499 + }, + { + "epoch": 2.43609022556391, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 45500 + }, + { + "epoch": 2.4361503759398495, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 45501 + }, + { + "epoch": 2.4362105263157896, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 45502 + }, + { + "epoch": 2.4362706766917293, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 45503 + }, + { + "epoch": 2.4363308270676693, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 45504 + }, + { + "epoch": 2.436390977443609, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 45505 + }, + { + "epoch": 2.4364511278195486, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 45506 + }, + { + "epoch": 2.4365112781954887, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 45507 + }, + { + "epoch": 2.4365714285714284, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 45508 + }, + { + "epoch": 2.4366315789473685, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 45509 + }, + { + "epoch": 2.436691729323308, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 45510 + }, + { + "epoch": 2.4367518796992482, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 45511 + }, + { + "epoch": 2.436812030075188, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 45512 + }, + { + "epoch": 2.436872180451128, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 45513 + }, + { + "epoch": 2.4369323308270676, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 45514 + }, + { + "epoch": 2.4369924812030077, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 45515 + }, + { + "epoch": 2.4370526315789474, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 45516 + }, + { + "epoch": 2.4371127819548875, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 45517 + }, + { + "epoch": 2.437172932330827, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 45518 + }, + { + "epoch": 2.4372330827067668, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 45519 + }, + { + "epoch": 2.437293233082707, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9231, + "step": 45520 + }, + { + "epoch": 2.4373533834586465, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 45521 + }, + { + "epoch": 2.4374135338345866, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 45522 + }, + { + "epoch": 2.4374736842105262, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 45523 + }, + { + "epoch": 2.4375338345864663, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 45524 + }, + { + "epoch": 2.437593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 45525 + }, + { + "epoch": 2.437654135338346, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 45526 + }, + { + "epoch": 2.4377142857142857, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 45527 + }, + { + "epoch": 2.4377744360902254, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 45528 + }, + { + "epoch": 2.4378345864661655, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 45529 + }, + { + "epoch": 2.437894736842105, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 45530 + }, + { + "epoch": 2.437954887218045, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 45531 + }, + { + "epoch": 2.438015037593985, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 45532 + }, + { + "epoch": 2.438075187969925, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 45533 + }, + { + "epoch": 2.4381353383458646, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 45534 + }, + { + "epoch": 2.4381954887218047, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 45535 + }, + { + "epoch": 2.4382556390977443, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 45536 + }, + { + "epoch": 2.438315789473684, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 45537 + }, + { + "epoch": 2.438375939849624, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 45538 + }, + { + "epoch": 2.4384360902255637, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 45539 + }, + { + "epoch": 2.438496240601504, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 45540 + }, + { + "epoch": 2.4385563909774435, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 45541 + }, + { + "epoch": 2.4386165413533836, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 45542 + }, + { + "epoch": 2.438676691729323, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 45543 + }, + { + "epoch": 2.4387368421052633, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 45544 + }, + { + "epoch": 2.438796992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 45545 + }, + { + "epoch": 2.4388571428571426, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 45546 + }, + { + "epoch": 2.4389172932330827, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 45547 + }, + { + "epoch": 2.438977443609023, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 45548 + }, + { + "epoch": 2.4390375939849624, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 45549 + }, + { + "epoch": 2.439097744360902, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 45550 + }, + { + "epoch": 2.439157894736842, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 45551 + }, + { + "epoch": 2.439218045112782, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 45552 + }, + { + "epoch": 2.439278195488722, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 45553 + }, + { + "epoch": 2.4393383458646616, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 45554 + }, + { + "epoch": 2.4393984962406017, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 45555 + }, + { + "epoch": 2.4394586466165413, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 45556 + }, + { + "epoch": 2.4395187969924814, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 45557 + }, + { + "epoch": 2.439578947368421, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 45558 + }, + { + "epoch": 2.4396390977443607, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 45559 + }, + { + "epoch": 2.439699248120301, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 45560 + }, + { + "epoch": 2.4397593984962405, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 45561 + }, + { + "epoch": 2.4398195488721806, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 45562 + }, + { + "epoch": 2.43987969924812, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 45563 + }, + { + "epoch": 2.4399398496240603, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 45564 + }, + { + "epoch": 2.44, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 45565 + }, + { + "epoch": 2.44006015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 45566 + }, + { + "epoch": 2.4401203007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 45567 + }, + { + "epoch": 2.4401804511278193, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1613, + "step": 45568 + }, + { + "epoch": 2.4402406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 45569 + }, + { + "epoch": 2.440300751879699, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 45570 + }, + { + "epoch": 2.440360902255639, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 45571 + }, + { + "epoch": 2.440421052631579, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 45572 + }, + { + "epoch": 2.440481203007519, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 45573 + }, + { + "epoch": 2.4405413533834586, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 45574 + }, + { + "epoch": 2.4406015037593987, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9336, + "step": 45575 + }, + { + "epoch": 2.4406616541353383, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 45576 + }, + { + "epoch": 2.440721804511278, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 45577 + }, + { + "epoch": 2.440781954887218, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9595, + "step": 45578 + }, + { + "epoch": 2.4408421052631577, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 45579 + }, + { + "epoch": 2.440902255639098, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 45580 + }, + { + "epoch": 2.4409624060150374, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 45581 + }, + { + "epoch": 2.4410225563909775, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 45582 + }, + { + "epoch": 2.441082706766917, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 45583 + }, + { + "epoch": 2.4411428571428573, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 45584 + }, + { + "epoch": 2.441203007518797, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 45585 + }, + { + "epoch": 2.441263157894737, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.3157, + "step": 45586 + }, + { + "epoch": 2.4413233082706767, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 45587 + }, + { + "epoch": 2.4413834586466168, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 45588 + }, + { + "epoch": 2.4414436090225564, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 45589 + }, + { + "epoch": 2.441503759398496, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 45590 + }, + { + "epoch": 2.441563909774436, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 45591 + }, + { + "epoch": 2.441624060150376, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 45592 + }, + { + "epoch": 2.441684210526316, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 45593 + }, + { + "epoch": 2.4417443609022556, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 45594 + }, + { + "epoch": 2.4418045112781956, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 45595 + }, + { + "epoch": 2.4418646616541353, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 45596 + }, + { + "epoch": 2.4419248120300754, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 45597 + }, + { + "epoch": 2.441984962406015, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 45598 + }, + { + "epoch": 2.4420451127819547, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 45599 + }, + { + "epoch": 2.442105263157895, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.8788, + "step": 45600 + }, + { + "epoch": 2.4421654135338344, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 45601 + }, + { + "epoch": 2.4422255639097745, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 45602 + }, + { + "epoch": 2.442285714285714, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 45603 + }, + { + "epoch": 2.4423458646616543, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 45604 + }, + { + "epoch": 2.442406015037594, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 45605 + }, + { + "epoch": 2.442466165413534, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 45606 + }, + { + "epoch": 2.4425263157894737, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1803, + "step": 45607 + }, + { + "epoch": 2.4425864661654133, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 45608 + }, + { + "epoch": 2.4426466165413534, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 45609 + }, + { + "epoch": 2.442706766917293, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.8823, + "step": 45610 + }, + { + "epoch": 2.442766917293233, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 45611 + }, + { + "epoch": 2.442827067669173, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 45612 + }, + { + "epoch": 2.442887218045113, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 45613 + }, + { + "epoch": 2.4429473684210525, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 45614 + }, + { + "epoch": 2.4430075187969926, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 45615 + }, + { + "epoch": 2.4430676691729323, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 45616 + }, + { + "epoch": 2.4431278195488724, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 45617 + }, + { + "epoch": 2.443187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 45618 + }, + { + "epoch": 2.443248120300752, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 45619 + }, + { + "epoch": 2.4433082706766918, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 45620 + }, + { + "epoch": 2.4433684210526314, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 45621 + }, + { + "epoch": 2.4434285714285715, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 45622 + }, + { + "epoch": 2.443488721804511, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 45623 + }, + { + "epoch": 2.4435488721804512, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 45624 + }, + { + "epoch": 2.443609022556391, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 45625 + }, + { + "epoch": 2.443669172932331, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1517, + "step": 45626 + }, + { + "epoch": 2.4437293233082706, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 45627 + }, + { + "epoch": 2.4437894736842107, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 45628 + }, + { + "epoch": 2.4438496240601504, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 45629 + }, + { + "epoch": 2.44390977443609, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 45630 + }, + { + "epoch": 2.44396992481203, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 45631 + }, + { + "epoch": 2.4440300751879698, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 45632 + }, + { + "epoch": 2.44409022556391, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9243, + "step": 45633 + }, + { + "epoch": 2.4441503759398495, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 45634 + }, + { + "epoch": 2.4442105263157896, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 45635 + }, + { + "epoch": 2.4442706766917293, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 45636 + }, + { + "epoch": 2.4443308270676694, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 45637 + }, + { + "epoch": 2.444390977443609, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 45638 + }, + { + "epoch": 2.4444511278195487, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 45639 + }, + { + "epoch": 2.4445112781954887, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 45640 + }, + { + "epoch": 2.4445714285714284, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 45641 + }, + { + "epoch": 2.4446315789473685, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 45642 + }, + { + "epoch": 2.444691729323308, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 45643 + }, + { + "epoch": 2.4447518796992482, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 45644 + }, + { + "epoch": 2.444812030075188, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 45645 + }, + { + "epoch": 2.444872180451128, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 45646 + }, + { + "epoch": 2.4449323308270676, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 45647 + }, + { + "epoch": 2.4449924812030073, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 45648 + }, + { + "epoch": 2.4450526315789474, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 45649 + }, + { + "epoch": 2.4451127819548875, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 45650 + }, + { + "epoch": 2.445172932330827, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 45651 + }, + { + "epoch": 2.4452330827067668, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 45652 + }, + { + "epoch": 2.445293233082707, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 45653 + }, + { + "epoch": 2.4453533834586465, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 45654 + }, + { + "epoch": 2.4454135338345866, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 45655 + }, + { + "epoch": 2.4454736842105262, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 45656 + }, + { + "epoch": 2.4455338345864663, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 45657 + }, + { + "epoch": 2.445593984962406, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 45658 + }, + { + "epoch": 2.445654135338346, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 45659 + }, + { + "epoch": 2.4457142857142857, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 45660 + }, + { + "epoch": 2.4457744360902254, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 45661 + }, + { + "epoch": 2.4458345864661655, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 45662 + }, + { + "epoch": 2.445894736842105, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 45663 + }, + { + "epoch": 2.445954887218045, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 45664 + }, + { + "epoch": 2.446015037593985, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1106, + "step": 45665 + }, + { + "epoch": 2.446075187969925, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 45666 + }, + { + "epoch": 2.4461353383458646, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 45667 + }, + { + "epoch": 2.4461954887218047, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 45668 + }, + { + "epoch": 2.4462556390977443, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 45669 + }, + { + "epoch": 2.446315789473684, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 45670 + }, + { + "epoch": 2.446375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 45671 + }, + { + "epoch": 2.4464360902255637, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 45672 + }, + { + "epoch": 2.446496240601504, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 45673 + }, + { + "epoch": 2.4465563909774435, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 45674 + }, + { + "epoch": 2.4466165413533836, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 45675 + }, + { + "epoch": 2.4466766917293232, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 45676 + }, + { + "epoch": 2.4467368421052633, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 45677 + }, + { + "epoch": 2.446796992481203, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 45678 + }, + { + "epoch": 2.4468571428571426, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 45679 + }, + { + "epoch": 2.4469172932330827, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 45680 + }, + { + "epoch": 2.4469774436090224, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 45681 + }, + { + "epoch": 2.4470375939849625, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.8858, + "step": 45682 + }, + { + "epoch": 2.447097744360902, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9348, + "step": 45683 + }, + { + "epoch": 2.447157894736842, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 45684 + }, + { + "epoch": 2.447218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 45685 + }, + { + "epoch": 2.447278195488722, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 45686 + }, + { + "epoch": 2.4473383458646616, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 45687 + }, + { + "epoch": 2.4473984962406017, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 45688 + }, + { + "epoch": 2.4474586466165413, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 45689 + }, + { + "epoch": 2.4475187969924814, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 45690 + }, + { + "epoch": 2.447578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 45691 + }, + { + "epoch": 2.4476390977443607, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 45692 + }, + { + "epoch": 2.447699248120301, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 45693 + }, + { + "epoch": 2.4477593984962405, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 45694 + }, + { + "epoch": 2.4478195488721806, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 45695 + }, + { + "epoch": 2.44787969924812, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 45696 + }, + { + "epoch": 2.4479398496240603, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 45697 + }, + { + "epoch": 2.448, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 45698 + }, + { + "epoch": 2.44806015037594, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 45699 + }, + { + "epoch": 2.4481203007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 45700 + }, + { + "epoch": 2.4481804511278193, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 45701 + }, + { + "epoch": 2.4482406015037594, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 45702 + }, + { + "epoch": 2.448300751879699, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 45703 + }, + { + "epoch": 2.448360902255639, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 45704 + }, + { + "epoch": 2.448421052631579, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 45705 + }, + { + "epoch": 2.448481203007519, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 45706 + }, + { + "epoch": 2.4485413533834586, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 45707 + }, + { + "epoch": 2.4486015037593987, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 45708 + }, + { + "epoch": 2.4486616541353383, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 45709 + }, + { + "epoch": 2.448721804511278, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 45710 + }, + { + "epoch": 2.448781954887218, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 45711 + }, + { + "epoch": 2.4488421052631577, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 45712 + }, + { + "epoch": 2.448902255639098, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 45713 + }, + { + "epoch": 2.4489624060150375, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.8957, + "step": 45714 + }, + { + "epoch": 2.4490225563909775, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 45715 + }, + { + "epoch": 2.449082706766917, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 45716 + }, + { + "epoch": 2.4491428571428573, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 45717 + }, + { + "epoch": 2.449203007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 45718 + }, + { + "epoch": 2.449263157894737, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 45719 + }, + { + "epoch": 2.4493233082706767, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 45720 + }, + { + "epoch": 2.4493834586466168, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 45721 + }, + { + "epoch": 2.4494436090225564, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 45722 + }, + { + "epoch": 2.449503759398496, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 45723 + }, + { + "epoch": 2.449563909774436, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 45724 + }, + { + "epoch": 2.449624060150376, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 45725 + }, + { + "epoch": 2.449684210526316, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 45726 + }, + { + "epoch": 2.4497443609022556, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9106, + "step": 45727 + }, + { + "epoch": 2.4498045112781957, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 45728 + }, + { + "epoch": 2.4498646616541353, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 45729 + }, + { + "epoch": 2.4499248120300754, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 45730 + }, + { + "epoch": 2.449984962406015, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 45731 + }, + { + "epoch": 2.4500451127819547, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 45732 + }, + { + "epoch": 2.450105263157895, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 45733 + }, + { + "epoch": 2.4501654135338344, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 45734 + }, + { + "epoch": 2.4502255639097745, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 45735 + }, + { + "epoch": 2.450285714285714, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 45736 + }, + { + "epoch": 2.4503458646616543, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 45737 + }, + { + "epoch": 2.450406015037594, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 45738 + }, + { + "epoch": 2.450466165413534, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 45739 + }, + { + "epoch": 2.4505263157894737, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 45740 + }, + { + "epoch": 2.4505864661654133, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1443, + "step": 45741 + }, + { + "epoch": 2.4506466165413534, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 45742 + }, + { + "epoch": 2.450706766917293, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 45743 + }, + { + "epoch": 2.450766917293233, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.8931, + "step": 45744 + }, + { + "epoch": 2.450827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 45745 + }, + { + "epoch": 2.450887218045113, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 45746 + }, + { + "epoch": 2.4509473684210525, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 45747 + }, + { + "epoch": 2.4510075187969926, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 45748 + }, + { + "epoch": 2.4510676691729323, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 45749 + }, + { + "epoch": 2.451127819548872, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2274, + "step": 45750 + }, + { + "epoch": 2.451187969924812, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 45751 + }, + { + "epoch": 2.451248120300752, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 45752 + }, + { + "epoch": 2.4513082706766918, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 45753 + }, + { + "epoch": 2.4513684210526314, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 45754 + }, + { + "epoch": 2.4514285714285715, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 45755 + }, + { + "epoch": 2.451488721804511, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 45756 + }, + { + "epoch": 2.4515488721804513, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 45757 + }, + { + "epoch": 2.451609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 45758 + }, + { + "epoch": 2.451669172932331, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 45759 + }, + { + "epoch": 2.4517293233082706, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 45760 + }, + { + "epoch": 2.4517894736842107, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 45761 + }, + { + "epoch": 2.4518496240601504, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 45762 + }, + { + "epoch": 2.45190977443609, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 45763 + }, + { + "epoch": 2.45196992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 45764 + }, + { + "epoch": 2.45203007518797, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 45765 + }, + { + "epoch": 2.45209022556391, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 45766 + }, + { + "epoch": 2.4521503759398495, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 45767 + }, + { + "epoch": 2.4522105263157896, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 45768 + }, + { + "epoch": 2.4522706766917293, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9191, + "step": 45769 + }, + { + "epoch": 2.4523308270676694, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 45770 + }, + { + "epoch": 2.452390977443609, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 45771 + }, + { + "epoch": 2.4524511278195487, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 45772 + }, + { + "epoch": 2.4525112781954888, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 45773 + }, + { + "epoch": 2.4525714285714284, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 45774 + }, + { + "epoch": 2.4526315789473685, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 45775 + }, + { + "epoch": 2.452691729323308, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 45776 + }, + { + "epoch": 2.4527518796992482, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 45777 + }, + { + "epoch": 2.452812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 45778 + }, + { + "epoch": 2.452872180451128, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 45779 + }, + { + "epoch": 2.4529323308270676, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 45780 + }, + { + "epoch": 2.4529924812030073, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 45781 + }, + { + "epoch": 2.4530526315789474, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 45782 + }, + { + "epoch": 2.4531127819548875, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 45783 + }, + { + "epoch": 2.453172932330827, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 45784 + }, + { + "epoch": 2.4532330827067668, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 45785 + }, + { + "epoch": 2.453293233082707, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.112, + "step": 45786 + }, + { + "epoch": 2.4533533834586465, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 45787 + }, + { + "epoch": 2.4534135338345866, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 45788 + }, + { + "epoch": 2.4534736842105263, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 45789 + }, + { + "epoch": 2.4535338345864663, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 45790 + }, + { + "epoch": 2.453593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 45791 + }, + { + "epoch": 2.453654135338346, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 45792 + }, + { + "epoch": 2.4537142857142857, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 45793 + }, + { + "epoch": 2.4537744360902254, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.941, + "step": 45794 + }, + { + "epoch": 2.4538345864661655, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 45795 + }, + { + "epoch": 2.453894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 45796 + }, + { + "epoch": 2.4539548872180452, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 45797 + }, + { + "epoch": 2.454015037593985, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9072, + "step": 45798 + }, + { + "epoch": 2.454075187969925, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 45799 + }, + { + "epoch": 2.4541353383458646, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.1458, + "step": 45800 + }, + { + "epoch": 2.4541954887218047, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 45801 + }, + { + "epoch": 2.4542556390977444, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 45802 + }, + { + "epoch": 2.454315789473684, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 45803 + }, + { + "epoch": 2.454375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 45804 + }, + { + "epoch": 2.4544360902255637, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 45805 + }, + { + "epoch": 2.454496240601504, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 45806 + }, + { + "epoch": 2.4545563909774435, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 45807 + }, + { + "epoch": 2.4546165413533836, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 45808 + }, + { + "epoch": 2.4546766917293232, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 45809 + }, + { + "epoch": 2.4547368421052633, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 45810 + }, + { + "epoch": 2.454796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 45811 + }, + { + "epoch": 2.4548571428571426, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 45812 + }, + { + "epoch": 2.4549172932330827, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 45813 + }, + { + "epoch": 2.4549774436090224, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9487, + "step": 45814 + }, + { + "epoch": 2.4550375939849625, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 45815 + }, + { + "epoch": 2.455097744360902, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 45816 + }, + { + "epoch": 2.455157894736842, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 45817 + }, + { + "epoch": 2.455218045112782, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 45818 + }, + { + "epoch": 2.455278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 45819 + }, + { + "epoch": 2.4553383458646616, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 45820 + }, + { + "epoch": 2.4553984962406017, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 45821 + }, + { + "epoch": 2.4554586466165413, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 45822 + }, + { + "epoch": 2.4555187969924814, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 45823 + }, + { + "epoch": 2.455578947368421, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 45824 + }, + { + "epoch": 2.4556390977443607, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 45825 + }, + { + "epoch": 2.455699248120301, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 45826 + }, + { + "epoch": 2.4557593984962405, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 45827 + }, + { + "epoch": 2.4558195488721806, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 45828 + }, + { + "epoch": 2.45587969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 45829 + }, + { + "epoch": 2.4559398496240603, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 45830 + }, + { + "epoch": 2.456, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 45831 + }, + { + "epoch": 2.45606015037594, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 45832 + }, + { + "epoch": 2.4561203007518797, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 45833 + }, + { + "epoch": 2.4561804511278194, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 45834 + }, + { + "epoch": 2.4562406015037594, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 45835 + }, + { + "epoch": 2.456300751879699, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 45836 + }, + { + "epoch": 2.456360902255639, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 45837 + }, + { + "epoch": 2.456421052631579, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 45838 + }, + { + "epoch": 2.456481203007519, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 45839 + }, + { + "epoch": 2.4565413533834586, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9364, + "step": 45840 + }, + { + "epoch": 2.4566015037593987, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 45841 + }, + { + "epoch": 2.4566616541353383, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 45842 + }, + { + "epoch": 2.456721804511278, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 45843 + }, + { + "epoch": 2.456781954887218, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 45844 + }, + { + "epoch": 2.4568421052631577, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 45845 + }, + { + "epoch": 2.456902255639098, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 45846 + }, + { + "epoch": 2.4569624060150375, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 45847 + }, + { + "epoch": 2.4570225563909776, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 45848 + }, + { + "epoch": 2.457082706766917, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 45849 + }, + { + "epoch": 2.4571428571428573, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 45850 + }, + { + "epoch": 2.457203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 45851 + }, + { + "epoch": 2.4572631578947366, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 45852 + }, + { + "epoch": 2.4573233082706767, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 45853 + }, + { + "epoch": 2.457383458646617, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 45854 + }, + { + "epoch": 2.4574436090225564, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 45855 + }, + { + "epoch": 2.457503759398496, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 45856 + }, + { + "epoch": 2.457563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9442, + "step": 45857 + }, + { + "epoch": 2.457624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 45858 + }, + { + "epoch": 2.457684210526316, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 45859 + }, + { + "epoch": 2.4577443609022556, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 45860 + }, + { + "epoch": 2.4578045112781957, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 45861 + }, + { + "epoch": 2.4578646616541353, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 45862 + }, + { + "epoch": 2.4579248120300754, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 45863 + }, + { + "epoch": 2.457984962406015, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9531, + "step": 45864 + }, + { + "epoch": 2.4580451127819547, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 45865 + }, + { + "epoch": 2.458105263157895, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 45866 + }, + { + "epoch": 2.4581654135338344, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 45867 + }, + { + "epoch": 2.4582255639097745, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 45868 + }, + { + "epoch": 2.458285714285714, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 45869 + }, + { + "epoch": 2.4583458646616543, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 45870 + }, + { + "epoch": 2.458406015037594, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 45871 + }, + { + "epoch": 2.458466165413534, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 45872 + }, + { + "epoch": 2.4585263157894737, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 45873 + }, + { + "epoch": 2.4585864661654133, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 45874 + }, + { + "epoch": 2.4586466165413534, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 45875 + }, + { + "epoch": 2.458706766917293, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 45876 + }, + { + "epoch": 2.458766917293233, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 45877 + }, + { + "epoch": 2.458827067669173, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 45878 + }, + { + "epoch": 2.458887218045113, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 45879 + }, + { + "epoch": 2.4589473684210525, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 45880 + }, + { + "epoch": 2.4590075187969926, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1332, + "step": 45881 + }, + { + "epoch": 2.4590676691729323, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9346, + "step": 45882 + }, + { + "epoch": 2.459127819548872, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 45883 + }, + { + "epoch": 2.459187969924812, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 45884 + }, + { + "epoch": 2.459248120300752, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 45885 + }, + { + "epoch": 2.4593082706766918, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 45886 + }, + { + "epoch": 2.4593684210526314, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 45887 + }, + { + "epoch": 2.4594285714285715, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 45888 + }, + { + "epoch": 2.459488721804511, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 45889 + }, + { + "epoch": 2.4595488721804513, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 45890 + }, + { + "epoch": 2.459609022556391, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 45891 + }, + { + "epoch": 2.459669172932331, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 45892 + }, + { + "epoch": 2.4597293233082707, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 45893 + }, + { + "epoch": 2.4597894736842107, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 45894 + }, + { + "epoch": 2.4598496240601504, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 45895 + }, + { + "epoch": 2.45990977443609, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 45896 + }, + { + "epoch": 2.45996992481203, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 45897 + }, + { + "epoch": 2.46003007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 45898 + }, + { + "epoch": 2.46009022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 45899 + }, + { + "epoch": 2.4601503759398495, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 45900 + }, + { + "epoch": 2.4602105263157896, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 45901 + }, + { + "epoch": 2.4602706766917293, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 45902 + }, + { + "epoch": 2.4603308270676694, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 45903 + }, + { + "epoch": 2.460390977443609, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 45904 + }, + { + "epoch": 2.4604511278195487, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 45905 + }, + { + "epoch": 2.4605112781954888, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 45906 + }, + { + "epoch": 2.4605714285714284, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 45907 + }, + { + "epoch": 2.4606315789473685, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 45908 + }, + { + "epoch": 2.460691729323308, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 45909 + }, + { + "epoch": 2.4607518796992482, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 45910 + }, + { + "epoch": 2.460812030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.956, + "step": 45911 + }, + { + "epoch": 2.460872180451128, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 45912 + }, + { + "epoch": 2.4609323308270676, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 45913 + }, + { + "epoch": 2.4609924812030073, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 45914 + }, + { + "epoch": 2.4610526315789474, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 45915 + }, + { + "epoch": 2.461112781954887, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 45916 + }, + { + "epoch": 2.461172932330827, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 45917 + }, + { + "epoch": 2.4612330827067668, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 45918 + }, + { + "epoch": 2.461293233082707, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 45919 + }, + { + "epoch": 2.4613533834586465, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.927, + "step": 45920 + }, + { + "epoch": 2.4614135338345866, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9345, + "step": 45921 + }, + { + "epoch": 2.4614736842105263, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 45922 + }, + { + "epoch": 2.4615338345864664, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 45923 + }, + { + "epoch": 2.461593984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 45924 + }, + { + "epoch": 2.461654135338346, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 45925 + }, + { + "epoch": 2.4617142857142857, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 45926 + }, + { + "epoch": 2.4617744360902254, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 45927 + }, + { + "epoch": 2.4618345864661655, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 45928 + }, + { + "epoch": 2.461894736842105, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 45929 + }, + { + "epoch": 2.4619548872180452, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 45930 + }, + { + "epoch": 2.462015037593985, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 45931 + }, + { + "epoch": 2.462075187969925, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 45932 + }, + { + "epoch": 2.4621353383458646, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 45933 + }, + { + "epoch": 2.4621954887218047, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 45934 + }, + { + "epoch": 2.4622556390977444, + "grad_norm": 1.484375, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 45935 + }, + { + "epoch": 2.462315789473684, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 45936 + }, + { + "epoch": 2.462375939849624, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 45937 + }, + { + "epoch": 2.4624360902255638, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 45938 + }, + { + "epoch": 2.462496240601504, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 45939 + }, + { + "epoch": 2.4625563909774435, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.947, + "step": 45940 + }, + { + "epoch": 2.4626165413533836, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 45941 + }, + { + "epoch": 2.4626766917293232, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 45942 + }, + { + "epoch": 2.4627368421052633, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 45943 + }, + { + "epoch": 2.462796992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1476, + "step": 45944 + }, + { + "epoch": 2.4628571428571426, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 45945 + }, + { + "epoch": 2.4629172932330827, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 45946 + }, + { + "epoch": 2.4629774436090224, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 45947 + }, + { + "epoch": 2.4630375939849625, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 45948 + }, + { + "epoch": 2.463097744360902, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 45949 + }, + { + "epoch": 2.463157894736842, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 45950 + }, + { + "epoch": 2.463218045112782, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 45951 + }, + { + "epoch": 2.463278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 45952 + }, + { + "epoch": 2.4633383458646616, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9423, + "step": 45953 + }, + { + "epoch": 2.4633984962406013, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 45954 + }, + { + "epoch": 2.4634586466165413, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 45955 + }, + { + "epoch": 2.4635187969924814, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1423, + "step": 45956 + }, + { + "epoch": 2.463578947368421, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 45957 + }, + { + "epoch": 2.4636390977443607, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 45958 + }, + { + "epoch": 2.463699248120301, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 45959 + }, + { + "epoch": 2.4637593984962405, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 45960 + }, + { + "epoch": 2.4638195488721806, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 45961 + }, + { + "epoch": 2.4638796992481202, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 45962 + }, + { + "epoch": 2.4639398496240603, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 45963 + }, + { + "epoch": 2.464, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 45964 + }, + { + "epoch": 2.46406015037594, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 45965 + }, + { + "epoch": 2.4641203007518797, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 45966 + }, + { + "epoch": 2.4641804511278194, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 45967 + }, + { + "epoch": 2.4642406015037595, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 45968 + }, + { + "epoch": 2.464300751879699, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 45969 + }, + { + "epoch": 2.464360902255639, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 45970 + }, + { + "epoch": 2.464421052631579, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 45971 + }, + { + "epoch": 2.464481203007519, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 45972 + }, + { + "epoch": 2.4645413533834586, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 45973 + }, + { + "epoch": 2.4646015037593987, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 45974 + }, + { + "epoch": 2.4646616541353383, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 45975 + }, + { + "epoch": 2.464721804511278, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 45976 + }, + { + "epoch": 2.464781954887218, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1525, + "step": 45977 + }, + { + "epoch": 2.4648421052631577, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 45978 + }, + { + "epoch": 2.464902255639098, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 45979 + }, + { + "epoch": 2.4649624060150375, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 45980 + }, + { + "epoch": 2.4650225563909776, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 45981 + }, + { + "epoch": 2.465082706766917, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1389, + "step": 45982 + }, + { + "epoch": 2.4651428571428573, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 45983 + }, + { + "epoch": 2.465203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 45984 + }, + { + "epoch": 2.4652631578947366, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 45985 + }, + { + "epoch": 2.4653233082706767, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 45986 + }, + { + "epoch": 2.465383458646617, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 45987 + }, + { + "epoch": 2.4654436090225564, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 45988 + }, + { + "epoch": 2.465503759398496, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 45989 + }, + { + "epoch": 2.465563909774436, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 45990 + }, + { + "epoch": 2.465624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 45991 + }, + { + "epoch": 2.465684210526316, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9214, + "step": 45992 + }, + { + "epoch": 2.4657443609022556, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 45993 + }, + { + "epoch": 2.4658045112781957, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 45994 + }, + { + "epoch": 2.4658646616541353, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9251, + "step": 45995 + }, + { + "epoch": 2.4659248120300754, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 45996 + }, + { + "epoch": 2.465984962406015, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 45997 + }, + { + "epoch": 2.4660451127819547, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 45998 + }, + { + "epoch": 2.466105263157895, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 45999 + }, + { + "epoch": 2.4661654135338344, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 46000 + }, + { + "epoch": 2.4662255639097745, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 46001 + }, + { + "epoch": 2.466285714285714, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 46002 + }, + { + "epoch": 2.4663458646616543, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 46003 + }, + { + "epoch": 2.466406015037594, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 46004 + }, + { + "epoch": 2.466466165413534, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 46005 + }, + { + "epoch": 2.4665263157894737, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9283, + "step": 46006 + }, + { + "epoch": 2.4665864661654133, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 46007 + }, + { + "epoch": 2.4666466165413534, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 46008 + }, + { + "epoch": 2.466706766917293, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1124, + "step": 46009 + }, + { + "epoch": 2.466766917293233, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1607, + "step": 46010 + }, + { + "epoch": 2.466827067669173, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 46011 + }, + { + "epoch": 2.466887218045113, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 46012 + }, + { + "epoch": 2.4669473684210526, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 46013 + }, + { + "epoch": 2.4670075187969926, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 46014 + }, + { + "epoch": 2.4670676691729323, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 46015 + }, + { + "epoch": 2.467127819548872, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 46016 + }, + { + "epoch": 2.467187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 46017 + }, + { + "epoch": 2.4672481203007517, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 46018 + }, + { + "epoch": 2.467308270676692, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 46019 + }, + { + "epoch": 2.4673684210526314, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 46020 + }, + { + "epoch": 2.4674285714285715, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 46021 + }, + { + "epoch": 2.467488721804511, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 46022 + }, + { + "epoch": 2.4675488721804513, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 46023 + }, + { + "epoch": 2.467609022556391, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 46024 + }, + { + "epoch": 2.467669172932331, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 46025 + }, + { + "epoch": 2.4677293233082707, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 46026 + }, + { + "epoch": 2.4677894736842108, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 46027 + }, + { + "epoch": 2.4678496240601504, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 46028 + }, + { + "epoch": 2.46790977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 46029 + }, + { + "epoch": 2.46796992481203, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 46030 + }, + { + "epoch": 2.46803007518797, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 46031 + }, + { + "epoch": 2.46809022556391, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.115, + "step": 46032 + }, + { + "epoch": 2.4681503759398495, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 46033 + }, + { + "epoch": 2.4682105263157896, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 46034 + }, + { + "epoch": 2.4682706766917293, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 46035 + }, + { + "epoch": 2.4683308270676694, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 46036 + }, + { + "epoch": 2.468390977443609, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 46037 + }, + { + "epoch": 2.4684511278195487, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 46038 + }, + { + "epoch": 2.4685112781954888, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 46039 + }, + { + "epoch": 2.4685714285714284, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 46040 + }, + { + "epoch": 2.4686315789473685, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1479, + "step": 46041 + }, + { + "epoch": 2.468691729323308, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 46042 + }, + { + "epoch": 2.4687518796992483, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 46043 + }, + { + "epoch": 2.468812030075188, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 46044 + }, + { + "epoch": 2.468872180451128, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 46045 + }, + { + "epoch": 2.4689323308270676, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 46046 + }, + { + "epoch": 2.4689924812030073, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 46047 + }, + { + "epoch": 2.4690526315789474, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 46048 + }, + { + "epoch": 2.469112781954887, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.247, + "step": 46049 + }, + { + "epoch": 2.469172932330827, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 46050 + }, + { + "epoch": 2.469233082706767, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 0.9378, + "step": 46051 + }, + { + "epoch": 2.469293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.8921, + "step": 46052 + }, + { + "epoch": 2.4693533834586465, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9376, + "step": 46053 + }, + { + "epoch": 2.4694135338345866, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 46054 + }, + { + "epoch": 2.4694736842105263, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 46055 + }, + { + "epoch": 2.4695338345864664, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 46056 + }, + { + "epoch": 2.469593984962406, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 46057 + }, + { + "epoch": 2.469654135338346, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 46058 + }, + { + "epoch": 2.4697142857142858, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 46059 + }, + { + "epoch": 2.4697744360902254, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 46060 + }, + { + "epoch": 2.4698345864661655, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 46061 + }, + { + "epoch": 2.469894736842105, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 46062 + }, + { + "epoch": 2.4699548872180452, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 46063 + }, + { + "epoch": 2.470015037593985, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 46064 + }, + { + "epoch": 2.470075187969925, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 46065 + }, + { + "epoch": 2.4701353383458646, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 46066 + }, + { + "epoch": 2.4701954887218047, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 46067 + }, + { + "epoch": 2.4702556390977444, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0909, + "step": 46068 + }, + { + "epoch": 2.470315789473684, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 46069 + }, + { + "epoch": 2.470375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 46070 + }, + { + "epoch": 2.4704360902255638, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 46071 + }, + { + "epoch": 2.470496240601504, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 46072 + }, + { + "epoch": 2.4705563909774435, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 46073 + }, + { + "epoch": 2.4706165413533836, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 46074 + }, + { + "epoch": 2.4706766917293232, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 46075 + }, + { + "epoch": 2.4707368421052633, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9481, + "step": 46076 + }, + { + "epoch": 2.470796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1026, + "step": 46077 + }, + { + "epoch": 2.4708571428571426, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 46078 + }, + { + "epoch": 2.4709172932330827, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 46079 + }, + { + "epoch": 2.4709774436090224, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1706, + "step": 46080 + }, + { + "epoch": 2.4710375939849625, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.2048, + "step": 46081 + }, + { + "epoch": 2.471097744360902, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 46082 + }, + { + "epoch": 2.471157894736842, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 46083 + }, + { + "epoch": 2.471218045112782, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 46084 + }, + { + "epoch": 2.471278195488722, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.2232, + "step": 46085 + }, + { + "epoch": 2.4713383458646616, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 46086 + }, + { + "epoch": 2.4713984962406013, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 46087 + }, + { + "epoch": 2.4714586466165414, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 46088 + }, + { + "epoch": 2.4715187969924814, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 46089 + }, + { + "epoch": 2.471578947368421, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9428, + "step": 46090 + }, + { + "epoch": 2.4716390977443607, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 46091 + }, + { + "epoch": 2.471699248120301, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 46092 + }, + { + "epoch": 2.4717593984962405, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 46093 + }, + { + "epoch": 2.4718195488721806, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 46094 + }, + { + "epoch": 2.4718796992481202, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 46095 + }, + { + "epoch": 2.4719398496240603, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 46096 + }, + { + "epoch": 2.472, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 46097 + }, + { + "epoch": 2.47206015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 46098 + }, + { + "epoch": 2.4721203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 46099 + }, + { + "epoch": 2.4721804511278194, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 46100 + }, + { + "epoch": 2.4722406015037595, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 46101 + }, + { + "epoch": 2.472300751879699, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 46102 + }, + { + "epoch": 2.472360902255639, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 46103 + }, + { + "epoch": 2.472421052631579, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 46104 + }, + { + "epoch": 2.472481203007519, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 46105 + }, + { + "epoch": 2.4725413533834586, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 46106 + }, + { + "epoch": 2.4726015037593987, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 46107 + }, + { + "epoch": 2.4726616541353383, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 46108 + }, + { + "epoch": 2.472721804511278, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 46109 + }, + { + "epoch": 2.472781954887218, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 46110 + }, + { + "epoch": 2.4728421052631577, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 46111 + }, + { + "epoch": 2.472902255639098, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 46112 + }, + { + "epoch": 2.4729624060150375, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 46113 + }, + { + "epoch": 2.4730225563909776, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 46114 + }, + { + "epoch": 2.473082706766917, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 46115 + }, + { + "epoch": 2.4731428571428573, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 46116 + }, + { + "epoch": 2.473203007518797, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 46117 + }, + { + "epoch": 2.4732631578947366, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 46118 + }, + { + "epoch": 2.4733233082706767, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 46119 + }, + { + "epoch": 2.4733834586466164, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 46120 + }, + { + "epoch": 2.4734436090225564, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 46121 + }, + { + "epoch": 2.473503759398496, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 46122 + }, + { + "epoch": 2.473563909774436, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 46123 + }, + { + "epoch": 2.473624060150376, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 46124 + }, + { + "epoch": 2.473684210526316, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 46125 + }, + { + "epoch": 2.4737443609022556, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 46126 + }, + { + "epoch": 2.4738045112781957, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 46127 + }, + { + "epoch": 2.4738646616541353, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 46128 + }, + { + "epoch": 2.4739248120300754, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 46129 + }, + { + "epoch": 2.473984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 46130 + }, + { + "epoch": 2.4740451127819547, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 46131 + }, + { + "epoch": 2.474105263157895, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 46132 + }, + { + "epoch": 2.4741654135338345, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 46133 + }, + { + "epoch": 2.4742255639097746, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 46134 + }, + { + "epoch": 2.474285714285714, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 46135 + }, + { + "epoch": 2.4743458646616543, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 46136 + }, + { + "epoch": 2.474406015037594, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 46137 + }, + { + "epoch": 2.474466165413534, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 46138 + }, + { + "epoch": 2.4745263157894737, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.8921, + "step": 46139 + }, + { + "epoch": 2.4745864661654133, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 46140 + }, + { + "epoch": 2.4746466165413534, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 46141 + }, + { + "epoch": 2.474706766917293, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 46142 + }, + { + "epoch": 2.474766917293233, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 46143 + }, + { + "epoch": 2.474827067669173, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 46144 + }, + { + "epoch": 2.474887218045113, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 46145 + }, + { + "epoch": 2.4749473684210526, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 46146 + }, + { + "epoch": 2.4750075187969927, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 46147 + }, + { + "epoch": 2.4750676691729323, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.8974, + "step": 46148 + }, + { + "epoch": 2.475127819548872, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 46149 + }, + { + "epoch": 2.475187969924812, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 46150 + }, + { + "epoch": 2.4752481203007517, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1439, + "step": 46151 + }, + { + "epoch": 2.475308270676692, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 46152 + }, + { + "epoch": 2.4753684210526314, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 46153 + }, + { + "epoch": 2.4754285714285715, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 46154 + }, + { + "epoch": 2.475488721804511, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 46155 + }, + { + "epoch": 2.4755488721804513, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 46156 + }, + { + "epoch": 2.475609022556391, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 46157 + }, + { + "epoch": 2.475669172932331, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 46158 + }, + { + "epoch": 2.4757293233082707, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 46159 + }, + { + "epoch": 2.4757894736842108, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 46160 + }, + { + "epoch": 2.4758496240601504, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 46161 + }, + { + "epoch": 2.47590977443609, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 46162 + }, + { + "epoch": 2.47596992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 46163 + }, + { + "epoch": 2.47603007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 46164 + }, + { + "epoch": 2.47609022556391, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 46165 + }, + { + "epoch": 2.4761503759398495, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 46166 + }, + { + "epoch": 2.4762105263157896, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 46167 + }, + { + "epoch": 2.4762706766917293, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 46168 + }, + { + "epoch": 2.4763308270676694, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 46169 + }, + { + "epoch": 2.476390977443609, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 46170 + }, + { + "epoch": 2.4764511278195487, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 46171 + }, + { + "epoch": 2.4765112781954888, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 46172 + }, + { + "epoch": 2.4765714285714284, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 46173 + }, + { + "epoch": 2.4766315789473685, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 46174 + }, + { + "epoch": 2.476691729323308, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 46175 + }, + { + "epoch": 2.4767518796992483, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 46176 + }, + { + "epoch": 2.476812030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 46177 + }, + { + "epoch": 2.476872180451128, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 46178 + }, + { + "epoch": 2.4769323308270677, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 46179 + }, + { + "epoch": 2.4769924812030073, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 46180 + }, + { + "epoch": 2.4770526315789474, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 46181 + }, + { + "epoch": 2.477112781954887, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 46182 + }, + { + "epoch": 2.477172932330827, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 46183 + }, + { + "epoch": 2.477233082706767, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 46184 + }, + { + "epoch": 2.477293233082707, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 46185 + }, + { + "epoch": 2.4773533834586465, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 46186 + }, + { + "epoch": 2.4774135338345866, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 46187 + }, + { + "epoch": 2.4774736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 46188 + }, + { + "epoch": 2.477533834586466, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 46189 + }, + { + "epoch": 2.477593984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 46190 + }, + { + "epoch": 2.477654135338346, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 46191 + }, + { + "epoch": 2.4777142857142858, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 46192 + }, + { + "epoch": 2.4777744360902254, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9181, + "step": 46193 + }, + { + "epoch": 2.4778345864661655, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 46194 + }, + { + "epoch": 2.477894736842105, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 46195 + }, + { + "epoch": 2.4779548872180452, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 46196 + }, + { + "epoch": 2.478015037593985, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 46197 + }, + { + "epoch": 2.478075187969925, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 46198 + }, + { + "epoch": 2.4781353383458646, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 46199 + }, + { + "epoch": 2.4781954887218047, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 46200 + }, + { + "epoch": 2.4782556390977444, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 46201 + }, + { + "epoch": 2.478315789473684, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 46202 + }, + { + "epoch": 2.478375939849624, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 46203 + }, + { + "epoch": 2.4784360902255638, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 46204 + }, + { + "epoch": 2.478496240601504, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 46205 + }, + { + "epoch": 2.4785563909774435, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 46206 + }, + { + "epoch": 2.4786165413533836, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 46207 + }, + { + "epoch": 2.4786766917293233, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 46208 + }, + { + "epoch": 2.4787368421052633, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 46209 + }, + { + "epoch": 2.478796992481203, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 46210 + }, + { + "epoch": 2.4788571428571426, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9373, + "step": 46211 + }, + { + "epoch": 2.4789172932330827, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 46212 + }, + { + "epoch": 2.4789774436090224, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1796, + "step": 46213 + }, + { + "epoch": 2.4790375939849625, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 46214 + }, + { + "epoch": 2.479097744360902, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 46215 + }, + { + "epoch": 2.4791578947368422, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 46216 + }, + { + "epoch": 2.479218045112782, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 46217 + }, + { + "epoch": 2.479278195488722, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 46218 + }, + { + "epoch": 2.4793383458646616, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 46219 + }, + { + "epoch": 2.4793984962406013, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9312, + "step": 46220 + }, + { + "epoch": 2.4794586466165414, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 46221 + }, + { + "epoch": 2.479518796992481, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 46222 + }, + { + "epoch": 2.479578947368421, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 46223 + }, + { + "epoch": 2.4796390977443608, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 46224 + }, + { + "epoch": 2.479699248120301, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 46225 + }, + { + "epoch": 2.4797593984962405, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 46226 + }, + { + "epoch": 2.4798195488721806, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 46227 + }, + { + "epoch": 2.4798796992481202, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 46228 + }, + { + "epoch": 2.4799398496240603, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 46229 + }, + { + "epoch": 2.48, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 46230 + }, + { + "epoch": 2.48006015037594, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 46231 + }, + { + "epoch": 2.4801203007518797, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 46232 + }, + { + "epoch": 2.4801804511278194, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 46233 + }, + { + "epoch": 2.4802406015037595, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 46234 + }, + { + "epoch": 2.480300751879699, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 46235 + }, + { + "epoch": 2.480360902255639, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 46236 + }, + { + "epoch": 2.480421052631579, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 46237 + }, + { + "epoch": 2.480481203007519, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 46238 + }, + { + "epoch": 2.4805413533834586, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 46239 + }, + { + "epoch": 2.4806015037593987, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0975, + "step": 46240 + }, + { + "epoch": 2.4806616541353383, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 46241 + }, + { + "epoch": 2.480721804511278, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 46242 + }, + { + "epoch": 2.480781954887218, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 46243 + }, + { + "epoch": 2.4808421052631577, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 46244 + }, + { + "epoch": 2.480902255639098, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 46245 + }, + { + "epoch": 2.4809624060150375, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 46246 + }, + { + "epoch": 2.4810225563909776, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9143, + "step": 46247 + }, + { + "epoch": 2.4810827067669172, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 46248 + }, + { + "epoch": 2.4811428571428573, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 46249 + }, + { + "epoch": 2.481203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 46250 + }, + { + "epoch": 2.4812631578947366, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 46251 + }, + { + "epoch": 2.4813233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 46252 + }, + { + "epoch": 2.4813834586466164, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 46253 + }, + { + "epoch": 2.4814436090225565, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 46254 + }, + { + "epoch": 2.481503759398496, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 46255 + }, + { + "epoch": 2.481563909774436, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9635, + "step": 46256 + }, + { + "epoch": 2.481624060150376, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 46257 + }, + { + "epoch": 2.481684210526316, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1958, + "step": 46258 + }, + { + "epoch": 2.4817443609022556, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 46259 + }, + { + "epoch": 2.4818045112781957, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9077, + "step": 46260 + }, + { + "epoch": 2.4818646616541353, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 46261 + }, + { + "epoch": 2.4819248120300754, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9251, + "step": 46262 + }, + { + "epoch": 2.481984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.931, + "step": 46263 + }, + { + "epoch": 2.4820451127819547, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 46264 + }, + { + "epoch": 2.482105263157895, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 46265 + }, + { + "epoch": 2.4821654135338345, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 46266 + }, + { + "epoch": 2.4822255639097746, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 46267 + }, + { + "epoch": 2.482285714285714, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 46268 + }, + { + "epoch": 2.4823458646616543, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 46269 + }, + { + "epoch": 2.482406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 46270 + }, + { + "epoch": 2.482466165413534, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 46271 + }, + { + "epoch": 2.4825263157894737, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 46272 + }, + { + "epoch": 2.4825864661654133, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 46273 + }, + { + "epoch": 2.4826466165413534, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 46274 + }, + { + "epoch": 2.482706766917293, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 46275 + }, + { + "epoch": 2.482766917293233, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 46276 + }, + { + "epoch": 2.482827067669173, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 46277 + }, + { + "epoch": 2.482887218045113, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 46278 + }, + { + "epoch": 2.4829473684210526, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.2241, + "step": 46279 + }, + { + "epoch": 2.4830075187969927, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 46280 + }, + { + "epoch": 2.4830676691729323, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 46281 + }, + { + "epoch": 2.483127819548872, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 46282 + }, + { + "epoch": 2.483187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 46283 + }, + { + "epoch": 2.4832481203007517, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 46284 + }, + { + "epoch": 2.483308270676692, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 46285 + }, + { + "epoch": 2.4833684210526314, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 46286 + }, + { + "epoch": 2.4834285714285715, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 46287 + }, + { + "epoch": 2.483488721804511, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.134, + "step": 46288 + }, + { + "epoch": 2.4835488721804513, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 46289 + }, + { + "epoch": 2.483609022556391, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 46290 + }, + { + "epoch": 2.4836691729323306, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 46291 + }, + { + "epoch": 2.4837293233082707, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 46292 + }, + { + "epoch": 2.4837894736842108, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 46293 + }, + { + "epoch": 2.4838496240601504, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 46294 + }, + { + "epoch": 2.48390977443609, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 46295 + }, + { + "epoch": 2.48396992481203, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 46296 + }, + { + "epoch": 2.48403007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 46297 + }, + { + "epoch": 2.48409022556391, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 46298 + }, + { + "epoch": 2.4841503759398496, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 46299 + }, + { + "epoch": 2.4842105263157896, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 46300 + }, + { + "epoch": 2.4842706766917293, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 46301 + }, + { + "epoch": 2.4843308270676694, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 46302 + }, + { + "epoch": 2.484390977443609, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 46303 + }, + { + "epoch": 2.4844511278195487, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 46304 + }, + { + "epoch": 2.484511278195489, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 46305 + }, + { + "epoch": 2.4845714285714284, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 46306 + }, + { + "epoch": 2.4846315789473685, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 46307 + }, + { + "epoch": 2.484691729323308, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9205, + "step": 46308 + }, + { + "epoch": 2.4847518796992483, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 46309 + }, + { + "epoch": 2.484812030075188, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 46310 + }, + { + "epoch": 2.484872180451128, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 46311 + }, + { + "epoch": 2.4849323308270677, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 46312 + }, + { + "epoch": 2.4849924812030073, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.2235, + "step": 46313 + }, + { + "epoch": 2.4850526315789474, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1269, + "step": 46314 + }, + { + "epoch": 2.485112781954887, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 46315 + }, + { + "epoch": 2.485172932330827, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1813, + "step": 46316 + }, + { + "epoch": 2.485233082706767, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 46317 + }, + { + "epoch": 2.485293233082707, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 46318 + }, + { + "epoch": 2.4853533834586465, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 46319 + }, + { + "epoch": 2.4854135338345866, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 46320 + }, + { + "epoch": 2.4854736842105263, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 46321 + }, + { + "epoch": 2.485533834586466, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 46322 + }, + { + "epoch": 2.485593984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 46323 + }, + { + "epoch": 2.4856541353383457, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 46324 + }, + { + "epoch": 2.4857142857142858, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 46325 + }, + { + "epoch": 2.4857744360902254, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9244, + "step": 46326 + }, + { + "epoch": 2.4858345864661655, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 46327 + }, + { + "epoch": 2.485894736842105, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 46328 + }, + { + "epoch": 2.4859548872180453, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1436, + "step": 46329 + }, + { + "epoch": 2.486015037593985, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 46330 + }, + { + "epoch": 2.486075187969925, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 46331 + }, + { + "epoch": 2.4861353383458646, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9531, + "step": 46332 + }, + { + "epoch": 2.4861954887218047, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 46333 + }, + { + "epoch": 2.4862556390977444, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 46334 + }, + { + "epoch": 2.486315789473684, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 46335 + }, + { + "epoch": 2.486375939849624, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 46336 + }, + { + "epoch": 2.4864360902255638, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 46337 + }, + { + "epoch": 2.486496240601504, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 46338 + }, + { + "epoch": 2.4865563909774435, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9469, + "step": 46339 + }, + { + "epoch": 2.4866165413533836, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9354, + "step": 46340 + }, + { + "epoch": 2.4866766917293233, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 46341 + }, + { + "epoch": 2.4867368421052634, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 46342 + }, + { + "epoch": 2.486796992481203, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 46343 + }, + { + "epoch": 2.4868571428571427, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.926, + "step": 46344 + }, + { + "epoch": 2.4869172932330827, + "grad_norm": 1.640625, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 46345 + }, + { + "epoch": 2.4869774436090224, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 46346 + }, + { + "epoch": 2.4870375939849625, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.8483, + "step": 46347 + }, + { + "epoch": 2.487097744360902, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 46348 + }, + { + "epoch": 2.4871578947368422, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 46349 + }, + { + "epoch": 2.487218045112782, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 46350 + }, + { + "epoch": 2.487278195488722, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9509, + "step": 46351 + }, + { + "epoch": 2.4873383458646616, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 46352 + }, + { + "epoch": 2.4873984962406013, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 46353 + }, + { + "epoch": 2.4874586466165414, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 46354 + }, + { + "epoch": 2.487518796992481, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 46355 + }, + { + "epoch": 2.487578947368421, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 46356 + }, + { + "epoch": 2.4876390977443608, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 46357 + }, + { + "epoch": 2.487699248120301, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 46358 + }, + { + "epoch": 2.4877593984962405, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 46359 + }, + { + "epoch": 2.4878195488721806, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 46360 + }, + { + "epoch": 2.4878796992481202, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 46361 + }, + { + "epoch": 2.4879398496240603, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 46362 + }, + { + "epoch": 2.488, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 46363 + }, + { + "epoch": 2.48806015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 46364 + }, + { + "epoch": 2.4881203007518797, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 46365 + }, + { + "epoch": 2.4881804511278194, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 46366 + }, + { + "epoch": 2.4882406015037595, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 46367 + }, + { + "epoch": 2.488300751879699, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 46368 + }, + { + "epoch": 2.488360902255639, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9257, + "step": 46369 + }, + { + "epoch": 2.488421052631579, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 46370 + }, + { + "epoch": 2.488481203007519, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 46371 + }, + { + "epoch": 2.4885413533834586, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.3082, + "step": 46372 + }, + { + "epoch": 2.4886015037593987, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 46373 + }, + { + "epoch": 2.4886616541353384, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 46374 + }, + { + "epoch": 2.488721804511278, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9323, + "step": 46375 + }, + { + "epoch": 2.488781954887218, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 46376 + }, + { + "epoch": 2.4888421052631577, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 46377 + }, + { + "epoch": 2.488902255639098, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 46378 + }, + { + "epoch": 2.4889624060150375, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 46379 + }, + { + "epoch": 2.4890225563909776, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 46380 + }, + { + "epoch": 2.4890827067669172, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 46381 + }, + { + "epoch": 2.4891428571428573, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 46382 + }, + { + "epoch": 2.489203007518797, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 46383 + }, + { + "epoch": 2.4892631578947366, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 46384 + }, + { + "epoch": 2.4893233082706767, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 46385 + }, + { + "epoch": 2.4893834586466164, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 46386 + }, + { + "epoch": 2.4894436090225565, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 46387 + }, + { + "epoch": 2.489503759398496, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 46388 + }, + { + "epoch": 2.489563909774436, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 46389 + }, + { + "epoch": 2.489624060150376, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 46390 + }, + { + "epoch": 2.489684210526316, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 46391 + }, + { + "epoch": 2.4897443609022556, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 46392 + }, + { + "epoch": 2.4898045112781952, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 46393 + }, + { + "epoch": 2.4898646616541353, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 46394 + }, + { + "epoch": 2.4899248120300754, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 46395 + }, + { + "epoch": 2.489984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 46396 + }, + { + "epoch": 2.4900451127819547, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 46397 + }, + { + "epoch": 2.490105263157895, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 46398 + }, + { + "epoch": 2.4901654135338345, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 46399 + }, + { + "epoch": 2.4902255639097746, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 46400 + }, + { + "epoch": 2.490285714285714, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 46401 + }, + { + "epoch": 2.4903458646616543, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 46402 + }, + { + "epoch": 2.490406015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 46403 + }, + { + "epoch": 2.490466165413534, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 46404 + }, + { + "epoch": 2.4905263157894737, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 46405 + }, + { + "epoch": 2.4905864661654133, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 46406 + }, + { + "epoch": 2.4906466165413534, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 46407 + }, + { + "epoch": 2.490706766917293, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 46408 + }, + { + "epoch": 2.490766917293233, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 46409 + }, + { + "epoch": 2.490827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 46410 + }, + { + "epoch": 2.490887218045113, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 46411 + }, + { + "epoch": 2.4909473684210526, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 46412 + }, + { + "epoch": 2.4910075187969927, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 46413 + }, + { + "epoch": 2.4910676691729323, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 46414 + }, + { + "epoch": 2.491127819548872, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9456, + "step": 46415 + }, + { + "epoch": 2.491187969924812, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 46416 + }, + { + "epoch": 2.4912481203007517, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 46417 + }, + { + "epoch": 2.491308270676692, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9112, + "step": 46418 + }, + { + "epoch": 2.4913684210526315, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 46419 + }, + { + "epoch": 2.4914285714285715, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 46420 + }, + { + "epoch": 2.491488721804511, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 46421 + }, + { + "epoch": 2.4915488721804513, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 46422 + }, + { + "epoch": 2.491609022556391, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 46423 + }, + { + "epoch": 2.4916691729323306, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 46424 + }, + { + "epoch": 2.4917293233082707, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 46425 + }, + { + "epoch": 2.4917894736842103, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 46426 + }, + { + "epoch": 2.4918496240601504, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 46427 + }, + { + "epoch": 2.49190977443609, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 46428 + }, + { + "epoch": 2.49196992481203, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1741, + "step": 46429 + }, + { + "epoch": 2.49203007518797, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 46430 + }, + { + "epoch": 2.49209022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 46431 + }, + { + "epoch": 2.4921503759398496, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 46432 + }, + { + "epoch": 2.4922105263157897, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 46433 + }, + { + "epoch": 2.4922706766917293, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 46434 + }, + { + "epoch": 2.4923308270676694, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 46435 + }, + { + "epoch": 2.492390977443609, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 46436 + }, + { + "epoch": 2.4924511278195487, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 46437 + }, + { + "epoch": 2.492511278195489, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 46438 + }, + { + "epoch": 2.4925714285714284, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 46439 + }, + { + "epoch": 2.4926315789473685, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 46440 + }, + { + "epoch": 2.492691729323308, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 46441 + }, + { + "epoch": 2.4927518796992483, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 46442 + }, + { + "epoch": 2.492812030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 46443 + }, + { + "epoch": 2.492872180451128, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 46444 + }, + { + "epoch": 2.4929323308270677, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 46445 + }, + { + "epoch": 2.4929924812030073, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.904, + "step": 46446 + }, + { + "epoch": 2.4930526315789474, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 46447 + }, + { + "epoch": 2.493112781954887, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 46448 + }, + { + "epoch": 2.493172932330827, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 46449 + }, + { + "epoch": 2.493233082706767, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 46450 + }, + { + "epoch": 2.493293233082707, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 46451 + }, + { + "epoch": 2.4933533834586465, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 46452 + }, + { + "epoch": 2.4934135338345866, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 46453 + }, + { + "epoch": 2.4934736842105263, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 46454 + }, + { + "epoch": 2.493533834586466, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 46455 + }, + { + "epoch": 2.493593984962406, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 46456 + }, + { + "epoch": 2.4936541353383457, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 46457 + }, + { + "epoch": 2.4937142857142858, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 46458 + }, + { + "epoch": 2.4937744360902254, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9219, + "step": 46459 + }, + { + "epoch": 2.4938345864661655, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 46460 + }, + { + "epoch": 2.493894736842105, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 46461 + }, + { + "epoch": 2.4939548872180453, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 46462 + }, + { + "epoch": 2.494015037593985, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 46463 + }, + { + "epoch": 2.494075187969925, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9221, + "step": 46464 + }, + { + "epoch": 2.4941353383458646, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 46465 + }, + { + "epoch": 2.4941954887218047, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9287, + "step": 46466 + }, + { + "epoch": 2.4942556390977444, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 46467 + }, + { + "epoch": 2.494315789473684, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 46468 + }, + { + "epoch": 2.494375939849624, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 46469 + }, + { + "epoch": 2.494436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 46470 + }, + { + "epoch": 2.494496240601504, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 46471 + }, + { + "epoch": 2.4945563909774435, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 46472 + }, + { + "epoch": 2.4946165413533836, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9282, + "step": 46473 + }, + { + "epoch": 2.4946766917293233, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 46474 + }, + { + "epoch": 2.4947368421052634, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 46475 + }, + { + "epoch": 2.494796992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 46476 + }, + { + "epoch": 2.4948571428571427, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 46477 + }, + { + "epoch": 2.4949172932330828, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 46478 + }, + { + "epoch": 2.4949774436090224, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 46479 + }, + { + "epoch": 2.4950375939849625, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 46480 + }, + { + "epoch": 2.495097744360902, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 46481 + }, + { + "epoch": 2.4951578947368422, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 46482 + }, + { + "epoch": 2.495218045112782, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 46483 + }, + { + "epoch": 2.495278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 46484 + }, + { + "epoch": 2.4953383458646616, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 46485 + }, + { + "epoch": 2.4953984962406013, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 46486 + }, + { + "epoch": 2.4954586466165414, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 46487 + }, + { + "epoch": 2.495518796992481, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 46488 + }, + { + "epoch": 2.495578947368421, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 46489 + }, + { + "epoch": 2.4956390977443608, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 46490 + }, + { + "epoch": 2.495699248120301, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 46491 + }, + { + "epoch": 2.4957593984962405, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 46492 + }, + { + "epoch": 2.4958195488721806, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 46493 + }, + { + "epoch": 2.4958796992481203, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9321, + "step": 46494 + }, + { + "epoch": 2.49593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 46495 + }, + { + "epoch": 2.496, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 46496 + }, + { + "epoch": 2.49606015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 46497 + }, + { + "epoch": 2.4961203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 46498 + }, + { + "epoch": 2.4961804511278194, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 46499 + }, + { + "epoch": 2.4962406015037595, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 46500 + }, + { + "epoch": 2.496300751879699, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 46501 + }, + { + "epoch": 2.4963609022556392, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 46502 + }, + { + "epoch": 2.496421052631579, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 46503 + }, + { + "epoch": 2.496481203007519, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.126, + "step": 46504 + }, + { + "epoch": 2.4965413533834586, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 46505 + }, + { + "epoch": 2.4966015037593987, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 46506 + }, + { + "epoch": 2.4966616541353384, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 46507 + }, + { + "epoch": 2.496721804511278, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 46508 + }, + { + "epoch": 2.496781954887218, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 46509 + }, + { + "epoch": 2.4968421052631578, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 46510 + }, + { + "epoch": 2.496902255639098, + "grad_norm": 1.3984375, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 46511 + }, + { + "epoch": 2.4969624060150375, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 46512 + }, + { + "epoch": 2.4970225563909776, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 46513 + }, + { + "epoch": 2.4970827067669172, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 46514 + }, + { + "epoch": 2.4971428571428573, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.8793, + "step": 46515 + }, + { + "epoch": 2.497203007518797, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 46516 + }, + { + "epoch": 2.4972631578947366, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 46517 + }, + { + "epoch": 2.4973233082706767, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 46518 + }, + { + "epoch": 2.4973834586466164, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 46519 + }, + { + "epoch": 2.4974436090225565, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 46520 + }, + { + "epoch": 2.497503759398496, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 46521 + }, + { + "epoch": 2.497563909774436, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 46522 + }, + { + "epoch": 2.497624060150376, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 46523 + }, + { + "epoch": 2.497684210526316, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 46524 + }, + { + "epoch": 2.4977443609022556, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 46525 + }, + { + "epoch": 2.4978045112781952, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 46526 + }, + { + "epoch": 2.4978646616541353, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 46527 + }, + { + "epoch": 2.4979248120300754, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 46528 + }, + { + "epoch": 2.497984962406015, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 46529 + }, + { + "epoch": 2.4980451127819547, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 46530 + }, + { + "epoch": 2.498105263157895, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 46531 + }, + { + "epoch": 2.4981654135338345, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 46532 + }, + { + "epoch": 2.4982255639097746, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 46533 + }, + { + "epoch": 2.498285714285714, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 46534 + }, + { + "epoch": 2.4983458646616543, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 46535 + }, + { + "epoch": 2.498406015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 46536 + }, + { + "epoch": 2.498466165413534, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9269, + "step": 46537 + }, + { + "epoch": 2.4985263157894737, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1851, + "step": 46538 + }, + { + "epoch": 2.4985864661654134, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 46539 + }, + { + "epoch": 2.4986466165413534, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 46540 + }, + { + "epoch": 2.498706766917293, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 46541 + }, + { + "epoch": 2.498766917293233, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 46542 + }, + { + "epoch": 2.498827067669173, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 46543 + }, + { + "epoch": 2.498887218045113, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 46544 + }, + { + "epoch": 2.4989473684210526, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 46545 + }, + { + "epoch": 2.4990075187969927, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9234, + "step": 46546 + }, + { + "epoch": 2.4990676691729323, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 46547 + }, + { + "epoch": 2.499127819548872, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 46548 + }, + { + "epoch": 2.499187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 46549 + }, + { + "epoch": 2.4992481203007517, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 46550 + }, + { + "epoch": 2.499308270676692, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 46551 + }, + { + "epoch": 2.4993684210526315, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1226, + "step": 46552 + }, + { + "epoch": 2.4994285714285716, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 46553 + }, + { + "epoch": 2.499488721804511, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 46554 + }, + { + "epoch": 2.4995488721804513, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 46555 + }, + { + "epoch": 2.499609022556391, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 46556 + }, + { + "epoch": 2.4996691729323306, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 46557 + }, + { + "epoch": 2.4997293233082707, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 46558 + }, + { + "epoch": 2.4997894736842103, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 46559 + }, + { + "epoch": 2.4998496240601504, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 46560 + }, + { + "epoch": 2.49990977443609, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 46561 + }, + { + "epoch": 2.49996992481203, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 46562 + }, + { + "epoch": 2.50003007518797, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 46563 + }, + { + "epoch": 2.50009022556391, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 46564 + }, + { + "epoch": 2.5001503759398496, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 46565 + }, + { + "epoch": 2.500210526315789, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 46566 + }, + { + "epoch": 2.5002706766917293, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 46567 + }, + { + "epoch": 2.5003308270676694, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 46568 + }, + { + "epoch": 2.500390977443609, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 46569 + }, + { + "epoch": 2.5004511278195487, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.957, + "step": 46570 + }, + { + "epoch": 2.500511278195489, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 46571 + }, + { + "epoch": 2.5005714285714284, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 46572 + }, + { + "epoch": 2.5006315789473685, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 46573 + }, + { + "epoch": 2.500691729323308, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9321, + "step": 46574 + }, + { + "epoch": 2.5007518796992483, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9292, + "step": 46575 + }, + { + "epoch": 2.500812030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 46576 + }, + { + "epoch": 2.500872180451128, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 46577 + }, + { + "epoch": 2.5009323308270677, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 46578 + }, + { + "epoch": 2.5009924812030073, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 46579 + }, + { + "epoch": 2.5010526315789474, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 46580 + }, + { + "epoch": 2.501112781954887, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 46581 + }, + { + "epoch": 2.501172932330827, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 46582 + }, + { + "epoch": 2.501233082706767, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 46583 + }, + { + "epoch": 2.501293233082707, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 46584 + }, + { + "epoch": 2.5013533834586466, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 46585 + }, + { + "epoch": 2.5014135338345866, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 46586 + }, + { + "epoch": 2.5014736842105263, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 46587 + }, + { + "epoch": 2.501533834586466, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 46588 + }, + { + "epoch": 2.501593984962406, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 46589 + }, + { + "epoch": 2.501654135338346, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.099, + "step": 46590 + }, + { + "epoch": 2.501714285714286, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 46591 + }, + { + "epoch": 2.5017744360902254, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 46592 + }, + { + "epoch": 2.5018345864661655, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 46593 + }, + { + "epoch": 2.501894736842105, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 46594 + }, + { + "epoch": 2.5019548872180453, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 46595 + }, + { + "epoch": 2.502015037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 46596 + }, + { + "epoch": 2.5020751879699246, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 46597 + }, + { + "epoch": 2.5021353383458647, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 46598 + }, + { + "epoch": 2.5021954887218048, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.256, + "step": 46599 + }, + { + "epoch": 2.5022556390977444, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 46600 + }, + { + "epoch": 2.502315789473684, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 46601 + }, + { + "epoch": 2.502375939849624, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 46602 + }, + { + "epoch": 2.502436090225564, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 46603 + }, + { + "epoch": 2.502496240601504, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 46604 + }, + { + "epoch": 2.5025563909774435, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 46605 + }, + { + "epoch": 2.502616541353383, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 46606 + }, + { + "epoch": 2.5026766917293233, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 46607 + }, + { + "epoch": 2.5027368421052634, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 46608 + }, + { + "epoch": 2.502796992481203, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.1236, + "step": 46609 + }, + { + "epoch": 2.5028571428571427, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 46610 + }, + { + "epoch": 2.5029172932330828, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 46611 + }, + { + "epoch": 2.5029774436090224, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 46612 + }, + { + "epoch": 2.5030375939849625, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 46613 + }, + { + "epoch": 2.503097744360902, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 46614 + }, + { + "epoch": 2.5031578947368422, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 46615 + }, + { + "epoch": 2.503218045112782, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 46616 + }, + { + "epoch": 2.503278195488722, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 46617 + }, + { + "epoch": 2.5033383458646616, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.2739, + "step": 46618 + }, + { + "epoch": 2.5033984962406013, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 46619 + }, + { + "epoch": 2.5034586466165414, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 46620 + }, + { + "epoch": 2.5035187969924815, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 46621 + }, + { + "epoch": 2.503578947368421, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 46622 + }, + { + "epoch": 2.5036390977443608, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 46623 + }, + { + "epoch": 2.503699248120301, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 46624 + }, + { + "epoch": 2.5037593984962405, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9408, + "step": 46625 + }, + { + "epoch": 2.5038195488721806, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 46626 + }, + { + "epoch": 2.5038796992481203, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 46627 + }, + { + "epoch": 2.50393984962406, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9442, + "step": 46628 + }, + { + "epoch": 2.504, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9361, + "step": 46629 + }, + { + "epoch": 2.50406015037594, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 46630 + }, + { + "epoch": 2.5041203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 46631 + }, + { + "epoch": 2.5041804511278194, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 46632 + }, + { + "epoch": 2.5042406015037595, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 46633 + }, + { + "epoch": 2.504300751879699, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 46634 + }, + { + "epoch": 2.5043609022556392, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 46635 + }, + { + "epoch": 2.504421052631579, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 46636 + }, + { + "epoch": 2.5044812030075185, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 46637 + }, + { + "epoch": 2.5045413533834586, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 46638 + }, + { + "epoch": 2.5046015037593987, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 46639 + }, + { + "epoch": 2.5046616541353384, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 46640 + }, + { + "epoch": 2.504721804511278, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 46641 + }, + { + "epoch": 2.504781954887218, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 46642 + }, + { + "epoch": 2.5048421052631578, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 46643 + }, + { + "epoch": 2.504902255639098, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 46644 + }, + { + "epoch": 2.5049624060150375, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 46645 + }, + { + "epoch": 2.5050225563909776, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 46646 + }, + { + "epoch": 2.5050827067669172, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 46647 + }, + { + "epoch": 2.5051428571428573, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 46648 + }, + { + "epoch": 2.505203007518797, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 46649 + }, + { + "epoch": 2.5052631578947366, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 46650 + }, + { + "epoch": 2.5053233082706767, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1619, + "step": 46651 + }, + { + "epoch": 2.5053834586466164, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 46652 + }, + { + "epoch": 2.5054436090225565, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 46653 + }, + { + "epoch": 2.505503759398496, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 46654 + }, + { + "epoch": 2.505563909774436, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1222, + "step": 46655 + }, + { + "epoch": 2.505624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 46656 + }, + { + "epoch": 2.505684210526316, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 46657 + }, + { + "epoch": 2.5057443609022556, + "grad_norm": 1.53125, + "learning_rate": 3e-05, + "loss": 1.1492, + "step": 46658 + }, + { + "epoch": 2.5058045112781953, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 46659 + }, + { + "epoch": 2.5058646616541354, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9166, + "step": 46660 + }, + { + "epoch": 2.5059248120300754, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 46661 + }, + { + "epoch": 2.505984962406015, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 46662 + }, + { + "epoch": 2.5060451127819547, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 46663 + }, + { + "epoch": 2.506105263157895, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 46664 + }, + { + "epoch": 2.5061654135338345, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 46665 + }, + { + "epoch": 2.5062255639097746, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9333, + "step": 46666 + }, + { + "epoch": 2.5062857142857142, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 46667 + }, + { + "epoch": 2.506345864661654, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 46668 + }, + { + "epoch": 2.506406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 46669 + }, + { + "epoch": 2.506466165413534, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 46670 + }, + { + "epoch": 2.5065263157894737, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 46671 + }, + { + "epoch": 2.5065864661654134, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 46672 + }, + { + "epoch": 2.5066466165413535, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 46673 + }, + { + "epoch": 2.506706766917293, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 46674 + }, + { + "epoch": 2.506766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 46675 + }, + { + "epoch": 2.506827067669173, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 46676 + }, + { + "epoch": 2.506887218045113, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 46677 + }, + { + "epoch": 2.5069473684210526, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 46678 + }, + { + "epoch": 2.5070075187969927, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 46679 + }, + { + "epoch": 2.5070676691729323, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 46680 + }, + { + "epoch": 2.507127819548872, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 46681 + }, + { + "epoch": 2.507187969924812, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 46682 + }, + { + "epoch": 2.5072481203007517, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 46683 + }, + { + "epoch": 2.507308270676692, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 46684 + }, + { + "epoch": 2.5073684210526315, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 46685 + }, + { + "epoch": 2.5074285714285716, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 46686 + }, + { + "epoch": 2.507488721804511, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 46687 + }, + { + "epoch": 2.5075488721804513, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 46688 + }, + { + "epoch": 2.507609022556391, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 46689 + }, + { + "epoch": 2.5076691729323306, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 46690 + }, + { + "epoch": 2.5077293233082707, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 46691 + }, + { + "epoch": 2.507789473684211, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 46692 + }, + { + "epoch": 2.5078496240601504, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 46693 + }, + { + "epoch": 2.50790977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 46694 + }, + { + "epoch": 2.50796992481203, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 46695 + }, + { + "epoch": 2.50803007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 46696 + }, + { + "epoch": 2.50809022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 46697 + }, + { + "epoch": 2.5081503759398496, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 46698 + }, + { + "epoch": 2.5082105263157892, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 46699 + }, + { + "epoch": 2.5082706766917293, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 46700 + }, + { + "epoch": 2.5083308270676694, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 46701 + }, + { + "epoch": 2.508390977443609, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 46702 + }, + { + "epoch": 2.5084511278195487, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 46703 + }, + { + "epoch": 2.508511278195489, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 46704 + }, + { + "epoch": 2.5085714285714285, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 46705 + }, + { + "epoch": 2.5086315789473685, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 46706 + }, + { + "epoch": 2.508691729323308, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9516, + "step": 46707 + }, + { + "epoch": 2.508751879699248, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 46708 + }, + { + "epoch": 2.508812030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 46709 + }, + { + "epoch": 2.508872180451128, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 46710 + }, + { + "epoch": 2.5089323308270677, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 46711 + }, + { + "epoch": 2.5089924812030073, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 46712 + }, + { + "epoch": 2.5090526315789474, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 46713 + }, + { + "epoch": 2.509112781954887, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 46714 + }, + { + "epoch": 2.509172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 46715 + }, + { + "epoch": 2.509233082706767, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 46716 + }, + { + "epoch": 2.509293233082707, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 46717 + }, + { + "epoch": 2.5093533834586466, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 46718 + }, + { + "epoch": 2.5094135338345867, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 46719 + }, + { + "epoch": 2.5094736842105263, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 46720 + }, + { + "epoch": 2.509533834586466, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 46721 + }, + { + "epoch": 2.509593984962406, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 46722 + }, + { + "epoch": 2.509654135338346, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 46723 + }, + { + "epoch": 2.509714285714286, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 46724 + }, + { + "epoch": 2.5097744360902254, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 46725 + }, + { + "epoch": 2.5098345864661655, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 46726 + }, + { + "epoch": 2.509894736842105, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 46727 + }, + { + "epoch": 2.5099548872180453, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 46728 + }, + { + "epoch": 2.510015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 46729 + }, + { + "epoch": 2.5100751879699246, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 46730 + }, + { + "epoch": 2.5101353383458647, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 46731 + }, + { + "epoch": 2.5101954887218048, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 46732 + }, + { + "epoch": 2.5102556390977444, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 46733 + }, + { + "epoch": 2.510315789473684, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 46734 + }, + { + "epoch": 2.510375939849624, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 46735 + }, + { + "epoch": 2.510436090225564, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.8685, + "step": 46736 + }, + { + "epoch": 2.510496240601504, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 46737 + }, + { + "epoch": 2.5105563909774435, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 46738 + }, + { + "epoch": 2.510616541353383, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.2208, + "step": 46739 + }, + { + "epoch": 2.5106766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 46740 + }, + { + "epoch": 2.5107368421052634, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 46741 + }, + { + "epoch": 2.510796992481203, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 46742 + }, + { + "epoch": 2.5108571428571427, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9481, + "step": 46743 + }, + { + "epoch": 2.5109172932330828, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.924, + "step": 46744 + }, + { + "epoch": 2.5109774436090224, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9222, + "step": 46745 + }, + { + "epoch": 2.5110375939849625, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 46746 + }, + { + "epoch": 2.511097744360902, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 46747 + }, + { + "epoch": 2.5111578947368423, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 46748 + }, + { + "epoch": 2.511218045112782, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 46749 + }, + { + "epoch": 2.511278195488722, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 46750 + }, + { + "epoch": 2.5113383458646616, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 46751 + }, + { + "epoch": 2.5113984962406013, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 46752 + }, + { + "epoch": 2.5114586466165414, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9263, + "step": 46753 + }, + { + "epoch": 2.511518796992481, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 46754 + }, + { + "epoch": 2.511578947368421, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 46755 + }, + { + "epoch": 2.511639097744361, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.2035, + "step": 46756 + }, + { + "epoch": 2.511699248120301, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 46757 + }, + { + "epoch": 2.5117593984962405, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 46758 + }, + { + "epoch": 2.5118195488721806, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 46759 + }, + { + "epoch": 2.5118796992481203, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 46760 + }, + { + "epoch": 2.51193984962406, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 46761 + }, + { + "epoch": 2.512, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 46762 + }, + { + "epoch": 2.51206015037594, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 46763 + }, + { + "epoch": 2.5121203007518798, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 46764 + }, + { + "epoch": 2.5121804511278194, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 46765 + }, + { + "epoch": 2.5122406015037595, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 46766 + }, + { + "epoch": 2.512300751879699, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 46767 + }, + { + "epoch": 2.5123609022556392, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 46768 + }, + { + "epoch": 2.512421052631579, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 46769 + }, + { + "epoch": 2.5124812030075185, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 46770 + }, + { + "epoch": 2.5125413533834586, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 46771 + }, + { + "epoch": 2.5126015037593987, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 46772 + }, + { + "epoch": 2.5126616541353384, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 46773 + }, + { + "epoch": 2.512721804511278, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2708, + "step": 46774 + }, + { + "epoch": 2.512781954887218, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 46775 + }, + { + "epoch": 2.5128421052631578, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 46776 + }, + { + "epoch": 2.512902255639098, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 46777 + }, + { + "epoch": 2.5129624060150375, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 46778 + }, + { + "epoch": 2.5130225563909776, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 46779 + }, + { + "epoch": 2.5130827067669173, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 46780 + }, + { + "epoch": 2.5131428571428573, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 46781 + }, + { + "epoch": 2.513203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 46782 + }, + { + "epoch": 2.5132631578947366, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 46783 + }, + { + "epoch": 2.5133233082706767, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9088, + "step": 46784 + }, + { + "epoch": 2.5133834586466164, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 46785 + }, + { + "epoch": 2.5134436090225565, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 46786 + }, + { + "epoch": 2.513503759398496, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1298, + "step": 46787 + }, + { + "epoch": 2.5135639097744362, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 46788 + }, + { + "epoch": 2.513624060150376, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 46789 + }, + { + "epoch": 2.513684210526316, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9202, + "step": 46790 + }, + { + "epoch": 2.5137443609022556, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 46791 + }, + { + "epoch": 2.5138045112781953, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 46792 + }, + { + "epoch": 2.5138646616541354, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 46793 + }, + { + "epoch": 2.5139248120300755, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 46794 + }, + { + "epoch": 2.513984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 46795 + }, + { + "epoch": 2.5140451127819547, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 46796 + }, + { + "epoch": 2.514105263157895, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 46797 + }, + { + "epoch": 2.5141654135338345, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 46798 + }, + { + "epoch": 2.5142255639097746, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 46799 + }, + { + "epoch": 2.5142857142857142, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 46800 + }, + { + "epoch": 2.514345864661654, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 46801 + }, + { + "epoch": 2.514406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 46802 + }, + { + "epoch": 2.514466165413534, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 46803 + }, + { + "epoch": 2.5145263157894737, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 46804 + }, + { + "epoch": 2.5145864661654134, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 46805 + }, + { + "epoch": 2.5146466165413535, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 46806 + }, + { + "epoch": 2.514706766917293, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 46807 + }, + { + "epoch": 2.514766917293233, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 46808 + }, + { + "epoch": 2.514827067669173, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 46809 + }, + { + "epoch": 2.5148872180451125, + "grad_norm": 1.53125, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 46810 + }, + { + "epoch": 2.5149473684210526, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.8864, + "step": 46811 + }, + { + "epoch": 2.5150075187969927, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 46812 + }, + { + "epoch": 2.5150676691729323, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 46813 + }, + { + "epoch": 2.515127819548872, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 46814 + }, + { + "epoch": 2.515187969924812, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 46815 + }, + { + "epoch": 2.5152481203007517, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 46816 + }, + { + "epoch": 2.515308270676692, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 46817 + }, + { + "epoch": 2.5153684210526315, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 46818 + }, + { + "epoch": 2.5154285714285716, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 46819 + }, + { + "epoch": 2.515488721804511, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 46820 + }, + { + "epoch": 2.5155488721804513, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 46821 + }, + { + "epoch": 2.515609022556391, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 46822 + }, + { + "epoch": 2.5156691729323306, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 46823 + }, + { + "epoch": 2.5157293233082707, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 46824 + }, + { + "epoch": 2.515789473684211, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 46825 + }, + { + "epoch": 2.5158496240601504, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1421, + "step": 46826 + }, + { + "epoch": 2.51590977443609, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 46827 + }, + { + "epoch": 2.51596992481203, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 46828 + }, + { + "epoch": 2.51603007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 46829 + }, + { + "epoch": 2.51609022556391, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 46830 + }, + { + "epoch": 2.5161503759398496, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 46831 + }, + { + "epoch": 2.5162105263157892, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 46832 + }, + { + "epoch": 2.5162706766917293, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 46833 + }, + { + "epoch": 2.5163308270676694, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 46834 + }, + { + "epoch": 2.516390977443609, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 46835 + }, + { + "epoch": 2.5164511278195487, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 46836 + }, + { + "epoch": 2.516511278195489, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 46837 + }, + { + "epoch": 2.5165714285714285, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 46838 + }, + { + "epoch": 2.5166315789473686, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 46839 + }, + { + "epoch": 2.516691729323308, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 46840 + }, + { + "epoch": 2.516751879699248, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 46841 + }, + { + "epoch": 2.516812030075188, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 46842 + }, + { + "epoch": 2.516872180451128, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 46843 + }, + { + "epoch": 2.5169323308270677, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 46844 + }, + { + "epoch": 2.5169924812030073, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 46845 + }, + { + "epoch": 2.5170526315789474, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 46846 + }, + { + "epoch": 2.517112781954887, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 46847 + }, + { + "epoch": 2.517172932330827, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 46848 + }, + { + "epoch": 2.517233082706767, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 46849 + }, + { + "epoch": 2.517293233082707, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 46850 + }, + { + "epoch": 2.5173533834586466, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 46851 + }, + { + "epoch": 2.5174135338345867, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 46852 + }, + { + "epoch": 2.5174736842105263, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 46853 + }, + { + "epoch": 2.517533834586466, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 46854 + }, + { + "epoch": 2.517593984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 46855 + }, + { + "epoch": 2.5176541353383457, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 46856 + }, + { + "epoch": 2.517714285714286, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 46857 + }, + { + "epoch": 2.5177744360902254, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 46858 + }, + { + "epoch": 2.5178345864661655, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 46859 + }, + { + "epoch": 2.517894736842105, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 46860 + }, + { + "epoch": 2.5179548872180453, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1381, + "step": 46861 + }, + { + "epoch": 2.518015037593985, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 46862 + }, + { + "epoch": 2.5180751879699246, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 46863 + }, + { + "epoch": 2.5181353383458647, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 46864 + }, + { + "epoch": 2.5181954887218048, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 46865 + }, + { + "epoch": 2.5182556390977444, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 46866 + }, + { + "epoch": 2.518315789473684, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 46867 + }, + { + "epoch": 2.518375939849624, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 46868 + }, + { + "epoch": 2.518436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 46869 + }, + { + "epoch": 2.518496240601504, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 46870 + }, + { + "epoch": 2.5185563909774435, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 46871 + }, + { + "epoch": 2.518616541353383, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1914, + "step": 46872 + }, + { + "epoch": 2.5186766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 46873 + }, + { + "epoch": 2.5187368421052634, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1155, + "step": 46874 + }, + { + "epoch": 2.518796992481203, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 46875 + }, + { + "epoch": 2.5188571428571427, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 46876 + }, + { + "epoch": 2.5189172932330828, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 46877 + }, + { + "epoch": 2.5189774436090224, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 46878 + }, + { + "epoch": 2.5190375939849625, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 46879 + }, + { + "epoch": 2.519097744360902, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 46880 + }, + { + "epoch": 2.5191578947368423, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 46881 + }, + { + "epoch": 2.519218045112782, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 46882 + }, + { + "epoch": 2.519278195488722, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 46883 + }, + { + "epoch": 2.5193383458646617, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 46884 + }, + { + "epoch": 2.5193984962406013, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 46885 + }, + { + "epoch": 2.5194586466165414, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 46886 + }, + { + "epoch": 2.519518796992481, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 46887 + }, + { + "epoch": 2.519578947368421, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 46888 + }, + { + "epoch": 2.519639097744361, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 46889 + }, + { + "epoch": 2.519699248120301, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 46890 + }, + { + "epoch": 2.5197593984962405, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 46891 + }, + { + "epoch": 2.5198195488721806, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 46892 + }, + { + "epoch": 2.5198796992481203, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 46893 + }, + { + "epoch": 2.51993984962406, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 46894 + }, + { + "epoch": 2.52, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 46895 + }, + { + "epoch": 2.52006015037594, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 46896 + }, + { + "epoch": 2.5201203007518798, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 46897 + }, + { + "epoch": 2.5201804511278194, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 46898 + }, + { + "epoch": 2.5202406015037595, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 46899 + }, + { + "epoch": 2.520300751879699, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 46900 + }, + { + "epoch": 2.5203609022556392, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 46901 + }, + { + "epoch": 2.520421052631579, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 46902 + }, + { + "epoch": 2.5204812030075185, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 46903 + }, + { + "epoch": 2.5205413533834586, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 46904 + }, + { + "epoch": 2.5206015037593987, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1463, + "step": 46905 + }, + { + "epoch": 2.5206616541353384, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 46906 + }, + { + "epoch": 2.520721804511278, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 46907 + }, + { + "epoch": 2.520781954887218, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 46908 + }, + { + "epoch": 2.5208421052631578, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 46909 + }, + { + "epoch": 2.520902255639098, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 46910 + }, + { + "epoch": 2.5209624060150375, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 46911 + }, + { + "epoch": 2.5210225563909776, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 46912 + }, + { + "epoch": 2.5210827067669173, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 46913 + }, + { + "epoch": 2.5211428571428574, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 46914 + }, + { + "epoch": 2.521203007518797, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 46915 + }, + { + "epoch": 2.5212631578947367, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 46916 + }, + { + "epoch": 2.5213233082706767, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 46917 + }, + { + "epoch": 2.5213834586466164, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 46918 + }, + { + "epoch": 2.5214436090225565, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 46919 + }, + { + "epoch": 2.521503759398496, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 46920 + }, + { + "epoch": 2.5215639097744362, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 46921 + }, + { + "epoch": 2.521624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 46922 + }, + { + "epoch": 2.521684210526316, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 46923 + }, + { + "epoch": 2.5217443609022556, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 46924 + }, + { + "epoch": 2.5218045112781953, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 46925 + }, + { + "epoch": 2.5218646616541354, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 46926 + }, + { + "epoch": 2.5219248120300755, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 46927 + }, + { + "epoch": 2.521984962406015, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 46928 + }, + { + "epoch": 2.5220451127819548, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 46929 + }, + { + "epoch": 2.522105263157895, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 46930 + }, + { + "epoch": 2.5221654135338345, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 46931 + }, + { + "epoch": 2.5222255639097746, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 46932 + }, + { + "epoch": 2.5222857142857142, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 46933 + }, + { + "epoch": 2.522345864661654, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 46934 + }, + { + "epoch": 2.522406015037594, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 46935 + }, + { + "epoch": 2.522466165413534, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9031, + "step": 46936 + }, + { + "epoch": 2.5225263157894737, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 46937 + }, + { + "epoch": 2.5225864661654134, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 46938 + }, + { + "epoch": 2.5226466165413535, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 46939 + }, + { + "epoch": 2.522706766917293, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 46940 + }, + { + "epoch": 2.522766917293233, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9025, + "step": 46941 + }, + { + "epoch": 2.522827067669173, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0863, + "step": 46942 + }, + { + "epoch": 2.5228872180451125, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 46943 + }, + { + "epoch": 2.5229473684210526, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 46944 + }, + { + "epoch": 2.5230075187969927, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 46945 + }, + { + "epoch": 2.5230676691729323, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 46946 + }, + { + "epoch": 2.523127819548872, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 46947 + }, + { + "epoch": 2.523187969924812, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 46948 + }, + { + "epoch": 2.5232481203007517, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 46949 + }, + { + "epoch": 2.523308270676692, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 46950 + }, + { + "epoch": 2.5233684210526315, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 46951 + }, + { + "epoch": 2.5234285714285716, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 46952 + }, + { + "epoch": 2.5234887218045112, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 46953 + }, + { + "epoch": 2.5235488721804513, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 46954 + }, + { + "epoch": 2.523609022556391, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 46955 + }, + { + "epoch": 2.5236691729323306, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 46956 + }, + { + "epoch": 2.5237293233082707, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 46957 + }, + { + "epoch": 2.5237894736842104, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 46958 + }, + { + "epoch": 2.5238496240601505, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 46959 + }, + { + "epoch": 2.52390977443609, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 46960 + }, + { + "epoch": 2.52396992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 46961 + }, + { + "epoch": 2.52403007518797, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 46962 + }, + { + "epoch": 2.52409022556391, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 46963 + }, + { + "epoch": 2.5241503759398496, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 46964 + }, + { + "epoch": 2.5242105263157892, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 46965 + }, + { + "epoch": 2.5242706766917293, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 46966 + }, + { + "epoch": 2.5243308270676694, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 46967 + }, + { + "epoch": 2.524390977443609, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 46968 + }, + { + "epoch": 2.5244511278195487, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 46969 + }, + { + "epoch": 2.524511278195489, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9033, + "step": 46970 + }, + { + "epoch": 2.5245714285714285, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 46971 + }, + { + "epoch": 2.5246315789473686, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 46972 + }, + { + "epoch": 2.524691729323308, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 46973 + }, + { + "epoch": 2.524751879699248, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 46974 + }, + { + "epoch": 2.524812030075188, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 46975 + }, + { + "epoch": 2.524872180451128, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 46976 + }, + { + "epoch": 2.5249323308270677, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 46977 + }, + { + "epoch": 2.5249924812030073, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 46978 + }, + { + "epoch": 2.5250526315789474, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 46979 + }, + { + "epoch": 2.525112781954887, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 46980 + }, + { + "epoch": 2.525172932330827, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1377, + "step": 46981 + }, + { + "epoch": 2.525233082706767, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 46982 + }, + { + "epoch": 2.525293233082707, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 46983 + }, + { + "epoch": 2.5253533834586466, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 46984 + }, + { + "epoch": 2.5254135338345867, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 46985 + }, + { + "epoch": 2.5254736842105263, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 46986 + }, + { + "epoch": 2.525533834586466, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 46987 + }, + { + "epoch": 2.525593984962406, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 46988 + }, + { + "epoch": 2.5256541353383457, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 46989 + }, + { + "epoch": 2.525714285714286, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 46990 + }, + { + "epoch": 2.5257744360902254, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 46991 + }, + { + "epoch": 2.5258345864661655, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 46992 + }, + { + "epoch": 2.525894736842105, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 46993 + }, + { + "epoch": 2.5259548872180453, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 46994 + }, + { + "epoch": 2.526015037593985, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 46995 + }, + { + "epoch": 2.5260751879699246, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 46996 + }, + { + "epoch": 2.5261353383458647, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 46997 + }, + { + "epoch": 2.5261954887218048, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 46998 + }, + { + "epoch": 2.5262556390977444, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 46999 + }, + { + "epoch": 2.526315789473684, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 47000 + }, + { + "epoch": 2.526375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 47001 + }, + { + "epoch": 2.526436090225564, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 47002 + }, + { + "epoch": 2.526496240601504, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9312, + "step": 47003 + }, + { + "epoch": 2.5265563909774436, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 47004 + }, + { + "epoch": 2.526616541353383, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 47005 + }, + { + "epoch": 2.5266766917293233, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 47006 + }, + { + "epoch": 2.5267368421052634, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 47007 + }, + { + "epoch": 2.526796992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 47008 + }, + { + "epoch": 2.5268571428571427, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 47009 + }, + { + "epoch": 2.526917293233083, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 47010 + }, + { + "epoch": 2.5269774436090224, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 47011 + }, + { + "epoch": 2.5270375939849625, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 47012 + }, + { + "epoch": 2.527097744360902, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 47013 + }, + { + "epoch": 2.5271578947368423, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 47014 + }, + { + "epoch": 2.527218045112782, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 47015 + }, + { + "epoch": 2.527278195488722, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 47016 + }, + { + "epoch": 2.5273383458646617, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 47017 + }, + { + "epoch": 2.5273984962406013, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 47018 + }, + { + "epoch": 2.5274586466165414, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 47019 + }, + { + "epoch": 2.527518796992481, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 47020 + }, + { + "epoch": 2.527578947368421, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 47021 + }, + { + "epoch": 2.527639097744361, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 47022 + }, + { + "epoch": 2.527699248120301, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 47023 + }, + { + "epoch": 2.5277593984962405, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 47024 + }, + { + "epoch": 2.5278195488721806, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 47025 + }, + { + "epoch": 2.5278796992481203, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9431, + "step": 47026 + }, + { + "epoch": 2.52793984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 47027 + }, + { + "epoch": 2.528, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 47028 + }, + { + "epoch": 2.52806015037594, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 47029 + }, + { + "epoch": 2.5281203007518798, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 47030 + }, + { + "epoch": 2.5281804511278194, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 47031 + }, + { + "epoch": 2.5282406015037595, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 47032 + }, + { + "epoch": 2.528300751879699, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 47033 + }, + { + "epoch": 2.5283609022556393, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 47034 + }, + { + "epoch": 2.528421052631579, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 47035 + }, + { + "epoch": 2.5284812030075186, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 47036 + }, + { + "epoch": 2.5285413533834586, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 47037 + }, + { + "epoch": 2.5286015037593987, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 47038 + }, + { + "epoch": 2.5286616541353384, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 47039 + }, + { + "epoch": 2.528721804511278, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1993, + "step": 47040 + }, + { + "epoch": 2.528781954887218, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 47041 + }, + { + "epoch": 2.528842105263158, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 47042 + }, + { + "epoch": 2.528902255639098, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 47043 + }, + { + "epoch": 2.5289624060150375, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 47044 + }, + { + "epoch": 2.529022556390977, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 47045 + }, + { + "epoch": 2.5290827067669173, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 47046 + }, + { + "epoch": 2.5291428571428574, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9407, + "step": 47047 + }, + { + "epoch": 2.529203007518797, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 47048 + }, + { + "epoch": 2.5292631578947367, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1199, + "step": 47049 + }, + { + "epoch": 2.5293233082706768, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 47050 + }, + { + "epoch": 2.5293834586466164, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 47051 + }, + { + "epoch": 2.5294436090225565, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 47052 + }, + { + "epoch": 2.529503759398496, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 47053 + }, + { + "epoch": 2.5295639097744362, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 47054 + }, + { + "epoch": 2.529624060150376, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9188, + "step": 47055 + }, + { + "epoch": 2.529684210526316, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1584, + "step": 47056 + }, + { + "epoch": 2.5297443609022556, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 47057 + }, + { + "epoch": 2.5298045112781953, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 47058 + }, + { + "epoch": 2.5298646616541354, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 47059 + }, + { + "epoch": 2.5299248120300755, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 47060 + }, + { + "epoch": 2.529984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 47061 + }, + { + "epoch": 2.5300451127819548, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 47062 + }, + { + "epoch": 2.530105263157895, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 47063 + }, + { + "epoch": 2.5301654135338345, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 47064 + }, + { + "epoch": 2.5302255639097746, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 47065 + }, + { + "epoch": 2.5302857142857142, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 47066 + }, + { + "epoch": 2.530345864661654, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 47067 + }, + { + "epoch": 2.530406015037594, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 47068 + }, + { + "epoch": 2.530466165413534, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9246, + "step": 47069 + }, + { + "epoch": 2.5305263157894737, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 47070 + }, + { + "epoch": 2.5305864661654134, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 47071 + }, + { + "epoch": 2.5306466165413535, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 47072 + }, + { + "epoch": 2.530706766917293, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 47073 + }, + { + "epoch": 2.530766917293233, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 47074 + }, + { + "epoch": 2.530827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 47075 + }, + { + "epoch": 2.5308872180451125, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 47076 + }, + { + "epoch": 2.5309473684210526, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 47077 + }, + { + "epoch": 2.5310075187969927, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1138, + "step": 47078 + }, + { + "epoch": 2.5310676691729324, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 47079 + }, + { + "epoch": 2.531127819548872, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 47080 + }, + { + "epoch": 2.531187969924812, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.13, + "step": 47081 + }, + { + "epoch": 2.5312481203007517, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 47082 + }, + { + "epoch": 2.531308270676692, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 47083 + }, + { + "epoch": 2.5313684210526315, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 47084 + }, + { + "epoch": 2.5314285714285716, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 47085 + }, + { + "epoch": 2.5314887218045112, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 47086 + }, + { + "epoch": 2.5315488721804513, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 47087 + }, + { + "epoch": 2.531609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 47088 + }, + { + "epoch": 2.5316691729323306, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 47089 + }, + { + "epoch": 2.5317293233082707, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 47090 + }, + { + "epoch": 2.5317894736842104, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 47091 + }, + { + "epoch": 2.5318496240601505, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 47092 + }, + { + "epoch": 2.53190977443609, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 47093 + }, + { + "epoch": 2.53196992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 47094 + }, + { + "epoch": 2.53203007518797, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 47095 + }, + { + "epoch": 2.53209022556391, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 47096 + }, + { + "epoch": 2.5321503759398496, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 47097 + }, + { + "epoch": 2.5322105263157892, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 47098 + }, + { + "epoch": 2.5322706766917293, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 47099 + }, + { + "epoch": 2.5323308270676694, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 47100 + }, + { + "epoch": 2.532390977443609, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 47101 + }, + { + "epoch": 2.5324511278195487, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 47102 + }, + { + "epoch": 2.532511278195489, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 47103 + }, + { + "epoch": 2.5325714285714285, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1307, + "step": 47104 + }, + { + "epoch": 2.5326315789473686, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 47105 + }, + { + "epoch": 2.532691729323308, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 47106 + }, + { + "epoch": 2.532751879699248, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 47107 + }, + { + "epoch": 2.532812030075188, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 47108 + }, + { + "epoch": 2.532872180451128, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 47109 + }, + { + "epoch": 2.5329323308270677, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 47110 + }, + { + "epoch": 2.5329924812030074, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 47111 + }, + { + "epoch": 2.5330526315789474, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 47112 + }, + { + "epoch": 2.533112781954887, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 47113 + }, + { + "epoch": 2.533172932330827, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9296, + "step": 47114 + }, + { + "epoch": 2.533233082706767, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 47115 + }, + { + "epoch": 2.533293233082707, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 47116 + }, + { + "epoch": 2.5333533834586466, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 47117 + }, + { + "epoch": 2.5334135338345867, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 47118 + }, + { + "epoch": 2.5334736842105263, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 47119 + }, + { + "epoch": 2.533533834586466, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 47120 + }, + { + "epoch": 2.533593984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 47121 + }, + { + "epoch": 2.5336541353383457, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 47122 + }, + { + "epoch": 2.533714285714286, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 47123 + }, + { + "epoch": 2.5337744360902255, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 47124 + }, + { + "epoch": 2.5338345864661656, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 47125 + }, + { + "epoch": 2.533894736842105, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 47126 + }, + { + "epoch": 2.5339548872180453, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 47127 + }, + { + "epoch": 2.534015037593985, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 47128 + }, + { + "epoch": 2.5340751879699246, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 47129 + }, + { + "epoch": 2.5341353383458647, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 47130 + }, + { + "epoch": 2.534195488721805, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 47131 + }, + { + "epoch": 2.5342556390977444, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 47132 + }, + { + "epoch": 2.534315789473684, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 47133 + }, + { + "epoch": 2.534375939849624, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 47134 + }, + { + "epoch": 2.534436090225564, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 47135 + }, + { + "epoch": 2.534496240601504, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 47136 + }, + { + "epoch": 2.5345563909774436, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 47137 + }, + { + "epoch": 2.534616541353383, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1378, + "step": 47138 + }, + { + "epoch": 2.5346766917293233, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 47139 + }, + { + "epoch": 2.5347368421052634, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9366, + "step": 47140 + }, + { + "epoch": 2.534796992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 47141 + }, + { + "epoch": 2.5348571428571427, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 47142 + }, + { + "epoch": 2.534917293233083, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 47143 + }, + { + "epoch": 2.5349774436090224, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 47144 + }, + { + "epoch": 2.5350375939849625, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 47145 + }, + { + "epoch": 2.535097744360902, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 47146 + }, + { + "epoch": 2.535157894736842, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 47147 + }, + { + "epoch": 2.535218045112782, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 47148 + }, + { + "epoch": 2.535278195488722, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 47149 + }, + { + "epoch": 2.5353383458646617, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 47150 + }, + { + "epoch": 2.5353984962406013, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 47151 + }, + { + "epoch": 2.5354586466165414, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 47152 + }, + { + "epoch": 2.535518796992481, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 47153 + }, + { + "epoch": 2.535578947368421, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 47154 + }, + { + "epoch": 2.535639097744361, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 47155 + }, + { + "epoch": 2.535699248120301, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 47156 + }, + { + "epoch": 2.5357593984962405, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 47157 + }, + { + "epoch": 2.5358195488721806, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 47158 + }, + { + "epoch": 2.5358796992481203, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 47159 + }, + { + "epoch": 2.53593984962406, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 47160 + }, + { + "epoch": 2.536, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 47161 + }, + { + "epoch": 2.53606015037594, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 47162 + }, + { + "epoch": 2.5361203007518798, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 47163 + }, + { + "epoch": 2.5361804511278194, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 47164 + }, + { + "epoch": 2.5362406015037595, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 47165 + }, + { + "epoch": 2.536300751879699, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 47166 + }, + { + "epoch": 2.5363609022556393, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 47167 + }, + { + "epoch": 2.536421052631579, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 47168 + }, + { + "epoch": 2.5364812030075186, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 47169 + }, + { + "epoch": 2.5365413533834587, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 47170 + }, + { + "epoch": 2.5366015037593987, + "grad_norm": 1.6171875, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 47171 + }, + { + "epoch": 2.5366616541353384, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 47172 + }, + { + "epoch": 2.536721804511278, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 47173 + }, + { + "epoch": 2.536781954887218, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 47174 + }, + { + "epoch": 2.536842105263158, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.8334, + "step": 47175 + }, + { + "epoch": 2.536902255639098, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 47176 + }, + { + "epoch": 2.5369624060150375, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 47177 + }, + { + "epoch": 2.537022556390977, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 47178 + }, + { + "epoch": 2.5370827067669173, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 47179 + }, + { + "epoch": 2.5371428571428574, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 47180 + }, + { + "epoch": 2.537203007518797, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.8909, + "step": 47181 + }, + { + "epoch": 2.5372631578947367, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 47182 + }, + { + "epoch": 2.5373233082706768, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 47183 + }, + { + "epoch": 2.5373834586466164, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 47184 + }, + { + "epoch": 2.5374436090225565, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9302, + "step": 47185 + }, + { + "epoch": 2.537503759398496, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 47186 + }, + { + "epoch": 2.5375639097744362, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 47187 + }, + { + "epoch": 2.537624060150376, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 47188 + }, + { + "epoch": 2.537684210526316, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 47189 + }, + { + "epoch": 2.5377443609022556, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 47190 + }, + { + "epoch": 2.5378045112781953, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 47191 + }, + { + "epoch": 2.5378646616541354, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 47192 + }, + { + "epoch": 2.537924812030075, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 47193 + }, + { + "epoch": 2.537984962406015, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 47194 + }, + { + "epoch": 2.5380451127819548, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 47195 + }, + { + "epoch": 2.538105263157895, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9255, + "step": 47196 + }, + { + "epoch": 2.5381654135338345, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 47197 + }, + { + "epoch": 2.5382255639097746, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 47198 + }, + { + "epoch": 2.5382857142857143, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 47199 + }, + { + "epoch": 2.538345864661654, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 47200 + }, + { + "epoch": 2.538406015037594, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 47201 + }, + { + "epoch": 2.538466165413534, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 47202 + }, + { + "epoch": 2.5385263157894737, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 47203 + }, + { + "epoch": 2.5385864661654134, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 47204 + }, + { + "epoch": 2.5386466165413535, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 47205 + }, + { + "epoch": 2.538706766917293, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 47206 + }, + { + "epoch": 2.5387669172932332, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 47207 + }, + { + "epoch": 2.538827067669173, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 47208 + }, + { + "epoch": 2.5388872180451125, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 47209 + }, + { + "epoch": 2.5389473684210526, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 47210 + }, + { + "epoch": 2.5390075187969927, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 47211 + }, + { + "epoch": 2.5390676691729324, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 47212 + }, + { + "epoch": 2.539127819548872, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 47213 + }, + { + "epoch": 2.539187969924812, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 47214 + }, + { + "epoch": 2.5392481203007518, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 47215 + }, + { + "epoch": 2.539308270676692, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 47216 + }, + { + "epoch": 2.5393684210526315, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 47217 + }, + { + "epoch": 2.5394285714285716, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 47218 + }, + { + "epoch": 2.5394887218045112, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 47219 + }, + { + "epoch": 2.5395488721804513, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 47220 + }, + { + "epoch": 2.539609022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 47221 + }, + { + "epoch": 2.5396691729323306, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 47222 + }, + { + "epoch": 2.5397293233082707, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 47223 + }, + { + "epoch": 2.5397894736842104, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 47224 + }, + { + "epoch": 2.5398496240601505, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 47225 + }, + { + "epoch": 2.53990977443609, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 47226 + }, + { + "epoch": 2.53996992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9024, + "step": 47227 + }, + { + "epoch": 2.54003007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 47228 + }, + { + "epoch": 2.54009022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 47229 + }, + { + "epoch": 2.5401503759398496, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9315, + "step": 47230 + }, + { + "epoch": 2.5402105263157893, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9317, + "step": 47231 + }, + { + "epoch": 2.5402706766917293, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 47232 + }, + { + "epoch": 2.5403308270676694, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 47233 + }, + { + "epoch": 2.540390977443609, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 47234 + }, + { + "epoch": 2.5404511278195487, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 47235 + }, + { + "epoch": 2.540511278195489, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 47236 + }, + { + "epoch": 2.5405714285714285, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 47237 + }, + { + "epoch": 2.5406315789473686, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 47238 + }, + { + "epoch": 2.5406917293233082, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 47239 + }, + { + "epoch": 2.540751879699248, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 47240 + }, + { + "epoch": 2.540812030075188, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 47241 + }, + { + "epoch": 2.540872180451128, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 47242 + }, + { + "epoch": 2.5409323308270677, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 47243 + }, + { + "epoch": 2.5409924812030074, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 47244 + }, + { + "epoch": 2.5410526315789475, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 47245 + }, + { + "epoch": 2.541112781954887, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 47246 + }, + { + "epoch": 2.541172932330827, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 47247 + }, + { + "epoch": 2.541233082706767, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 47248 + }, + { + "epoch": 2.5412932330827065, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9412, + "step": 47249 + }, + { + "epoch": 2.5413533834586466, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 47250 + }, + { + "epoch": 2.5414135338345867, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 47251 + }, + { + "epoch": 2.5414736842105263, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 47252 + }, + { + "epoch": 2.541533834586466, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 47253 + }, + { + "epoch": 2.541593984962406, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 47254 + }, + { + "epoch": 2.5416541353383457, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 47255 + }, + { + "epoch": 2.541714285714286, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 47256 + }, + { + "epoch": 2.5417744360902255, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 47257 + }, + { + "epoch": 2.5418345864661656, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 47258 + }, + { + "epoch": 2.541894736842105, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.2381, + "step": 47259 + }, + { + "epoch": 2.5419548872180453, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 47260 + }, + { + "epoch": 2.542015037593985, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 47261 + }, + { + "epoch": 2.5420751879699246, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 47262 + }, + { + "epoch": 2.5421353383458647, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 47263 + }, + { + "epoch": 2.542195488721805, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 47264 + }, + { + "epoch": 2.5422556390977444, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 47265 + }, + { + "epoch": 2.542315789473684, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 47266 + }, + { + "epoch": 2.542375939849624, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 47267 + }, + { + "epoch": 2.542436090225564, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 47268 + }, + { + "epoch": 2.542496240601504, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9479, + "step": 47269 + }, + { + "epoch": 2.5425563909774436, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 47270 + }, + { + "epoch": 2.542616541353383, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9082, + "step": 47271 + }, + { + "epoch": 2.5426766917293233, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 47272 + }, + { + "epoch": 2.5427368421052634, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 47273 + }, + { + "epoch": 2.542796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 47274 + }, + { + "epoch": 2.5428571428571427, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 47275 + }, + { + "epoch": 2.542917293233083, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 47276 + }, + { + "epoch": 2.5429774436090224, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 47277 + }, + { + "epoch": 2.5430375939849625, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 47278 + }, + { + "epoch": 2.543097744360902, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 47279 + }, + { + "epoch": 2.543157894736842, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 47280 + }, + { + "epoch": 2.543218045112782, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 47281 + }, + { + "epoch": 2.543278195488722, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 47282 + }, + { + "epoch": 2.5433383458646617, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9433, + "step": 47283 + }, + { + "epoch": 2.5433984962406013, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 47284 + }, + { + "epoch": 2.5434586466165414, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 47285 + }, + { + "epoch": 2.543518796992481, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 47286 + }, + { + "epoch": 2.543578947368421, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 47287 + }, + { + "epoch": 2.543639097744361, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 47288 + }, + { + "epoch": 2.543699248120301, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 47289 + }, + { + "epoch": 2.5437593984962406, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 47290 + }, + { + "epoch": 2.5438195488721806, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 47291 + }, + { + "epoch": 2.5438796992481203, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 47292 + }, + { + "epoch": 2.54393984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 47293 + }, + { + "epoch": 2.544, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 47294 + }, + { + "epoch": 2.5440601503759397, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9205, + "step": 47295 + }, + { + "epoch": 2.54412030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 47296 + }, + { + "epoch": 2.5441804511278194, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.3255, + "step": 47297 + }, + { + "epoch": 2.5442406015037595, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 47298 + }, + { + "epoch": 2.544300751879699, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 47299 + }, + { + "epoch": 2.5443609022556393, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 47300 + }, + { + "epoch": 2.544421052631579, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 47301 + }, + { + "epoch": 2.5444812030075186, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1337, + "step": 47302 + }, + { + "epoch": 2.5445413533834587, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 47303 + }, + { + "epoch": 2.5446015037593988, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 47304 + }, + { + "epoch": 2.5446616541353384, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.8897, + "step": 47305 + }, + { + "epoch": 2.544721804511278, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 47306 + }, + { + "epoch": 2.544781954887218, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 47307 + }, + { + "epoch": 2.544842105263158, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 47308 + }, + { + "epoch": 2.544902255639098, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 47309 + }, + { + "epoch": 2.5449624060150375, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 47310 + }, + { + "epoch": 2.545022556390977, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 47311 + }, + { + "epoch": 2.5450827067669173, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 47312 + }, + { + "epoch": 2.5451428571428574, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1774, + "step": 47313 + }, + { + "epoch": 2.545203007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 47314 + }, + { + "epoch": 2.5452631578947367, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 47315 + }, + { + "epoch": 2.5453233082706768, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 47316 + }, + { + "epoch": 2.5453834586466164, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 47317 + }, + { + "epoch": 2.5454436090225565, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 47318 + }, + { + "epoch": 2.545503759398496, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 47319 + }, + { + "epoch": 2.5455639097744363, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 47320 + }, + { + "epoch": 2.545624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 47321 + }, + { + "epoch": 2.545684210526316, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 47322 + }, + { + "epoch": 2.5457443609022556, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 47323 + }, + { + "epoch": 2.5458045112781953, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 47324 + }, + { + "epoch": 2.5458646616541354, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 47325 + }, + { + "epoch": 2.545924812030075, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 47326 + }, + { + "epoch": 2.545984962406015, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 47327 + }, + { + "epoch": 2.5460451127819548, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 47328 + }, + { + "epoch": 2.546105263157895, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 47329 + }, + { + "epoch": 2.5461654135338345, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 47330 + }, + { + "epoch": 2.5462255639097746, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 47331 + }, + { + "epoch": 2.5462857142857143, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 47332 + }, + { + "epoch": 2.546345864661654, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 47333 + }, + { + "epoch": 2.546406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 47334 + }, + { + "epoch": 2.546466165413534, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 47335 + }, + { + "epoch": 2.5465263157894737, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 47336 + }, + { + "epoch": 2.5465864661654134, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 47337 + }, + { + "epoch": 2.5466466165413535, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 47338 + }, + { + "epoch": 2.546706766917293, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 47339 + }, + { + "epoch": 2.5467669172932332, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 47340 + }, + { + "epoch": 2.546827067669173, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 47341 + }, + { + "epoch": 2.5468872180451125, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 47342 + }, + { + "epoch": 2.5469473684210526, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 47343 + }, + { + "epoch": 2.5470075187969927, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 47344 + }, + { + "epoch": 2.5470676691729324, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 47345 + }, + { + "epoch": 2.547127819548872, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 47346 + }, + { + "epoch": 2.547187969924812, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 47347 + }, + { + "epoch": 2.5472481203007518, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 47348 + }, + { + "epoch": 2.547308270676692, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 47349 + }, + { + "epoch": 2.5473684210526315, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 47350 + }, + { + "epoch": 2.547428571428571, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 47351 + }, + { + "epoch": 2.5474887218045112, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 47352 + }, + { + "epoch": 2.5475488721804513, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 47353 + }, + { + "epoch": 2.547609022556391, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 47354 + }, + { + "epoch": 2.5476691729323306, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 47355 + }, + { + "epoch": 2.5477293233082707, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 47356 + }, + { + "epoch": 2.5477894736842104, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 47357 + }, + { + "epoch": 2.5478496240601505, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 47358 + }, + { + "epoch": 2.54790977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 47359 + }, + { + "epoch": 2.54796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 47360 + }, + { + "epoch": 2.54803007518797, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 47361 + }, + { + "epoch": 2.54809022556391, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 47362 + }, + { + "epoch": 2.5481503759398496, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 47363 + }, + { + "epoch": 2.5482105263157893, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 47364 + }, + { + "epoch": 2.5482706766917294, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 47365 + }, + { + "epoch": 2.5483308270676694, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 47366 + }, + { + "epoch": 2.548390977443609, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9316, + "step": 47367 + }, + { + "epoch": 2.5484511278195487, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 47368 + }, + { + "epoch": 2.548511278195489, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9451, + "step": 47369 + }, + { + "epoch": 2.5485714285714285, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 47370 + }, + { + "epoch": 2.5486315789473686, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 47371 + }, + { + "epoch": 2.5486917293233082, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 47372 + }, + { + "epoch": 2.548751879699248, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 47373 + }, + { + "epoch": 2.548812030075188, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 47374 + }, + { + "epoch": 2.548872180451128, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 47375 + }, + { + "epoch": 2.5489323308270677, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 47376 + }, + { + "epoch": 2.5489924812030074, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 47377 + }, + { + "epoch": 2.5490526315789475, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 47378 + }, + { + "epoch": 2.549112781954887, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 47379 + }, + { + "epoch": 2.549172932330827, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 47380 + }, + { + "epoch": 2.549233082706767, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 47381 + }, + { + "epoch": 2.5492932330827065, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 47382 + }, + { + "epoch": 2.5493533834586466, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 47383 + }, + { + "epoch": 2.5494135338345867, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 47384 + }, + { + "epoch": 2.5494736842105263, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 47385 + }, + { + "epoch": 2.549533834586466, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 47386 + }, + { + "epoch": 2.549593984962406, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 47387 + }, + { + "epoch": 2.5496541353383457, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 47388 + }, + { + "epoch": 2.549714285714286, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 47389 + }, + { + "epoch": 2.5497744360902255, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 47390 + }, + { + "epoch": 2.5498345864661656, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.153, + "step": 47391 + }, + { + "epoch": 2.549894736842105, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 47392 + }, + { + "epoch": 2.5499548872180453, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 47393 + }, + { + "epoch": 2.550015037593985, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 47394 + }, + { + "epoch": 2.5500751879699246, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 47395 + }, + { + "epoch": 2.5501353383458647, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 47396 + }, + { + "epoch": 2.5501954887218043, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 47397 + }, + { + "epoch": 2.5502556390977444, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 47398 + }, + { + "epoch": 2.550315789473684, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 47399 + }, + { + "epoch": 2.550375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 47400 + }, + { + "epoch": 2.550436090225564, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 47401 + }, + { + "epoch": 2.550496240601504, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 47402 + }, + { + "epoch": 2.5505563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 47403 + }, + { + "epoch": 2.5506165413533832, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 47404 + }, + { + "epoch": 2.5506766917293233, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1031, + "step": 47405 + }, + { + "epoch": 2.5507368421052634, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 47406 + }, + { + "epoch": 2.550796992481203, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 47407 + }, + { + "epoch": 2.5508571428571427, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 47408 + }, + { + "epoch": 2.550917293233083, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 47409 + }, + { + "epoch": 2.5509774436090225, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 47410 + }, + { + "epoch": 2.5510375939849625, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 47411 + }, + { + "epoch": 2.551097744360902, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 47412 + }, + { + "epoch": 2.551157894736842, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 47413 + }, + { + "epoch": 2.551218045112782, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 47414 + }, + { + "epoch": 2.551278195488722, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 47415 + }, + { + "epoch": 2.5513383458646617, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 47416 + }, + { + "epoch": 2.5513984962406013, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 47417 + }, + { + "epoch": 2.5514586466165414, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 47418 + }, + { + "epoch": 2.551518796992481, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 47419 + }, + { + "epoch": 2.551578947368421, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 47420 + }, + { + "epoch": 2.551639097744361, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 47421 + }, + { + "epoch": 2.551699248120301, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 47422 + }, + { + "epoch": 2.5517593984962406, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 47423 + }, + { + "epoch": 2.5518195488721807, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 47424 + }, + { + "epoch": 2.5518796992481203, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 47425 + }, + { + "epoch": 2.55193984962406, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 47426 + }, + { + "epoch": 2.552, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 47427 + }, + { + "epoch": 2.5520601503759397, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 47428 + }, + { + "epoch": 2.55212030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 47429 + }, + { + "epoch": 2.5521804511278194, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 47430 + }, + { + "epoch": 2.5522406015037595, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 47431 + }, + { + "epoch": 2.552300751879699, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 47432 + }, + { + "epoch": 2.5523609022556393, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 47433 + }, + { + "epoch": 2.552421052631579, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 47434 + }, + { + "epoch": 2.5524812030075186, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 47435 + }, + { + "epoch": 2.5525413533834587, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 47436 + }, + { + "epoch": 2.5526015037593988, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 47437 + }, + { + "epoch": 2.5526616541353384, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.934, + "step": 47438 + }, + { + "epoch": 2.552721804511278, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 47439 + }, + { + "epoch": 2.552781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 47440 + }, + { + "epoch": 2.552842105263158, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 47441 + }, + { + "epoch": 2.552902255639098, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 47442 + }, + { + "epoch": 2.5529624060150375, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 47443 + }, + { + "epoch": 2.553022556390977, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 47444 + }, + { + "epoch": 2.5530827067669173, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 47445 + }, + { + "epoch": 2.5531428571428574, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 47446 + }, + { + "epoch": 2.553203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 47447 + }, + { + "epoch": 2.5532631578947367, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 47448 + }, + { + "epoch": 2.5533233082706768, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 47449 + }, + { + "epoch": 2.5533834586466164, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9207, + "step": 47450 + }, + { + "epoch": 2.5534436090225565, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 47451 + }, + { + "epoch": 2.553503759398496, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 47452 + }, + { + "epoch": 2.553563909774436, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 47453 + }, + { + "epoch": 2.553624060150376, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 47454 + }, + { + "epoch": 2.553684210526316, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 47455 + }, + { + "epoch": 2.5537443609022557, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 47456 + }, + { + "epoch": 2.5538045112781953, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 47457 + }, + { + "epoch": 2.5538646616541354, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 47458 + }, + { + "epoch": 2.553924812030075, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 47459 + }, + { + "epoch": 2.553984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 47460 + }, + { + "epoch": 2.554045112781955, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 47461 + }, + { + "epoch": 2.554105263157895, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 47462 + }, + { + "epoch": 2.5541654135338345, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 47463 + }, + { + "epoch": 2.5542255639097746, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 47464 + }, + { + "epoch": 2.5542857142857143, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 47465 + }, + { + "epoch": 2.554345864661654, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 47466 + }, + { + "epoch": 2.554406015037594, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 47467 + }, + { + "epoch": 2.554466165413534, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 47468 + }, + { + "epoch": 2.5545263157894738, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 47469 + }, + { + "epoch": 2.5545864661654134, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9282, + "step": 47470 + }, + { + "epoch": 2.5546466165413535, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 47471 + }, + { + "epoch": 2.554706766917293, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 47472 + }, + { + "epoch": 2.5547669172932332, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 47473 + }, + { + "epoch": 2.554827067669173, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 47474 + }, + { + "epoch": 2.5548872180451125, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 47475 + }, + { + "epoch": 2.5549473684210526, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 47476 + }, + { + "epoch": 2.5550075187969927, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 47477 + }, + { + "epoch": 2.5550676691729324, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 47478 + }, + { + "epoch": 2.555127819548872, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 47479 + }, + { + "epoch": 2.555187969924812, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 47480 + }, + { + "epoch": 2.5552481203007518, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 47481 + }, + { + "epoch": 2.555308270676692, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 47482 + }, + { + "epoch": 2.5553684210526315, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 47483 + }, + { + "epoch": 2.555428571428571, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 47484 + }, + { + "epoch": 2.5554887218045113, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 47485 + }, + { + "epoch": 2.5555488721804513, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1075, + "step": 47486 + }, + { + "epoch": 2.555609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 47487 + }, + { + "epoch": 2.5556691729323306, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 47488 + }, + { + "epoch": 2.5557293233082707, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 47489 + }, + { + "epoch": 2.5557894736842104, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 47490 + }, + { + "epoch": 2.5558496240601505, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 47491 + }, + { + "epoch": 2.55590977443609, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 47492 + }, + { + "epoch": 2.5559699248120302, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 47493 + }, + { + "epoch": 2.55603007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 47494 + }, + { + "epoch": 2.55609022556391, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 47495 + }, + { + "epoch": 2.5561503759398496, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 47496 + }, + { + "epoch": 2.5562105263157893, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 47497 + }, + { + "epoch": 2.5562706766917294, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 47498 + }, + { + "epoch": 2.556330827067669, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 47499 + }, + { + "epoch": 2.556390977443609, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 47500 + }, + { + "epoch": 2.5564511278195488, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 47501 + }, + { + "epoch": 2.556511278195489, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 47502 + }, + { + "epoch": 2.5565714285714285, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 47503 + }, + { + "epoch": 2.5566315789473686, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 47504 + }, + { + "epoch": 2.5566917293233082, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 47505 + }, + { + "epoch": 2.556751879699248, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 47506 + }, + { + "epoch": 2.556812030075188, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 47507 + }, + { + "epoch": 2.556872180451128, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 47508 + }, + { + "epoch": 2.5569323308270677, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9338, + "step": 47509 + }, + { + "epoch": 2.5569924812030074, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 47510 + }, + { + "epoch": 2.5570526315789475, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 47511 + }, + { + "epoch": 2.557112781954887, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 47512 + }, + { + "epoch": 2.557172932330827, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9456, + "step": 47513 + }, + { + "epoch": 2.557233082706767, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 47514 + }, + { + "epoch": 2.5572932330827065, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 47515 + }, + { + "epoch": 2.5573533834586466, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 47516 + }, + { + "epoch": 2.5574135338345867, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 47517 + }, + { + "epoch": 2.5574736842105263, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 47518 + }, + { + "epoch": 2.557533834586466, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 47519 + }, + { + "epoch": 2.557593984962406, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9514, + "step": 47520 + }, + { + "epoch": 2.5576541353383457, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 47521 + }, + { + "epoch": 2.557714285714286, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 47522 + }, + { + "epoch": 2.5577744360902255, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 0.8534, + "step": 47523 + }, + { + "epoch": 2.5578345864661656, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 47524 + }, + { + "epoch": 2.557894736842105, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 47525 + }, + { + "epoch": 2.5579548872180453, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.2448, + "step": 47526 + }, + { + "epoch": 2.558015037593985, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 47527 + }, + { + "epoch": 2.5580751879699246, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 47528 + }, + { + "epoch": 2.5581353383458647, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 47529 + }, + { + "epoch": 2.5581954887218044, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 47530 + }, + { + "epoch": 2.5582556390977444, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 47531 + }, + { + "epoch": 2.558315789473684, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 47532 + }, + { + "epoch": 2.558375939849624, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 47533 + }, + { + "epoch": 2.558436090225564, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 47534 + }, + { + "epoch": 2.558496240601504, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 47535 + }, + { + "epoch": 2.5585563909774436, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 47536 + }, + { + "epoch": 2.5586165413533832, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 47537 + }, + { + "epoch": 2.5586766917293233, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2598, + "step": 47538 + }, + { + "epoch": 2.5587368421052634, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 47539 + }, + { + "epoch": 2.558796992481203, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 47540 + }, + { + "epoch": 2.5588571428571427, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1332, + "step": 47541 + }, + { + "epoch": 2.558917293233083, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 47542 + }, + { + "epoch": 2.5589774436090225, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 47543 + }, + { + "epoch": 2.5590375939849626, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 47544 + }, + { + "epoch": 2.559097744360902, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 47545 + }, + { + "epoch": 2.559157894736842, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 47546 + }, + { + "epoch": 2.559218045112782, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 47547 + }, + { + "epoch": 2.559278195488722, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1592, + "step": 47548 + }, + { + "epoch": 2.5593383458646617, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 47549 + }, + { + "epoch": 2.5593984962406013, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 47550 + }, + { + "epoch": 2.5594586466165414, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 47551 + }, + { + "epoch": 2.559518796992481, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 47552 + }, + { + "epoch": 2.559578947368421, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 47553 + }, + { + "epoch": 2.559639097744361, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 47554 + }, + { + "epoch": 2.5596992481203005, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 47555 + }, + { + "epoch": 2.5597593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 47556 + }, + { + "epoch": 2.5598195488721807, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 47557 + }, + { + "epoch": 2.5598796992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 47558 + }, + { + "epoch": 2.55993984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 47559 + }, + { + "epoch": 2.56, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9386, + "step": 47560 + }, + { + "epoch": 2.5600601503759397, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 47561 + }, + { + "epoch": 2.56012030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 47562 + }, + { + "epoch": 2.5601804511278194, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 47563 + }, + { + "epoch": 2.5602406015037595, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 47564 + }, + { + "epoch": 2.560300751879699, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1347, + "step": 47565 + }, + { + "epoch": 2.5603609022556393, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 47566 + }, + { + "epoch": 2.560421052631579, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 47567 + }, + { + "epoch": 2.5604812030075186, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 47568 + }, + { + "epoch": 2.5605413533834587, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 47569 + }, + { + "epoch": 2.5606015037593988, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 47570 + }, + { + "epoch": 2.5606616541353384, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 47571 + }, + { + "epoch": 2.560721804511278, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 47572 + }, + { + "epoch": 2.560781954887218, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 47573 + }, + { + "epoch": 2.560842105263158, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 47574 + }, + { + "epoch": 2.560902255639098, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 47575 + }, + { + "epoch": 2.5609624060150376, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 47576 + }, + { + "epoch": 2.561022556390977, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 47577 + }, + { + "epoch": 2.5610827067669173, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 47578 + }, + { + "epoch": 2.5611428571428574, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 47579 + }, + { + "epoch": 2.561203007518797, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 47580 + }, + { + "epoch": 2.5612631578947367, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 47581 + }, + { + "epoch": 2.561323308270677, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 47582 + }, + { + "epoch": 2.5613834586466164, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 47583 + }, + { + "epoch": 2.5614436090225565, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 47584 + }, + { + "epoch": 2.561503759398496, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 47585 + }, + { + "epoch": 2.561563909774436, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1564, + "step": 47586 + }, + { + "epoch": 2.561624060150376, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 47587 + }, + { + "epoch": 2.561684210526316, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 47588 + }, + { + "epoch": 2.5617443609022557, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 47589 + }, + { + "epoch": 2.5618045112781953, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 47590 + }, + { + "epoch": 2.5618646616541354, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 47591 + }, + { + "epoch": 2.561924812030075, + "grad_norm": 1.8671875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 47592 + }, + { + "epoch": 2.561984962406015, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 47593 + }, + { + "epoch": 2.562045112781955, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 47594 + }, + { + "epoch": 2.562105263157895, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 47595 + }, + { + "epoch": 2.5621654135338345, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 47596 + }, + { + "epoch": 2.5622255639097746, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 47597 + }, + { + "epoch": 2.5622857142857143, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 47598 + }, + { + "epoch": 2.562345864661654, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 47599 + }, + { + "epoch": 2.562406015037594, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 47600 + }, + { + "epoch": 2.5624661654135337, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 47601 + }, + { + "epoch": 2.5625263157894738, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 47602 + }, + { + "epoch": 2.5625864661654134, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 47603 + }, + { + "epoch": 2.5626466165413535, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 47604 + }, + { + "epoch": 2.562706766917293, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 47605 + }, + { + "epoch": 2.5627669172932332, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 47606 + }, + { + "epoch": 2.562827067669173, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 47607 + }, + { + "epoch": 2.5628872180451125, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 47608 + }, + { + "epoch": 2.5629473684210526, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 47609 + }, + { + "epoch": 2.5630075187969927, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 47610 + }, + { + "epoch": 2.5630676691729324, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 47611 + }, + { + "epoch": 2.563127819548872, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 47612 + }, + { + "epoch": 2.563187969924812, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 47613 + }, + { + "epoch": 2.5632481203007518, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 47614 + }, + { + "epoch": 2.563308270676692, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 47615 + }, + { + "epoch": 2.5633684210526315, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9262, + "step": 47616 + }, + { + "epoch": 2.563428571428571, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 47617 + }, + { + "epoch": 2.5634887218045113, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 47618 + }, + { + "epoch": 2.5635488721804514, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9197, + "step": 47619 + }, + { + "epoch": 2.563609022556391, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 47620 + }, + { + "epoch": 2.5636691729323307, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 47621 + }, + { + "epoch": 2.5637293233082707, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9102, + "step": 47622 + }, + { + "epoch": 2.5637894736842104, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 47623 + }, + { + "epoch": 2.5638496240601505, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 47624 + }, + { + "epoch": 2.56390977443609, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 47625 + }, + { + "epoch": 2.5639699248120302, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 47626 + }, + { + "epoch": 2.56403007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 47627 + }, + { + "epoch": 2.56409022556391, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 47628 + }, + { + "epoch": 2.5641503759398496, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1488, + "step": 47629 + }, + { + "epoch": 2.5642105263157893, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 47630 + }, + { + "epoch": 2.5642706766917294, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 47631 + }, + { + "epoch": 2.564330827067669, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 47632 + }, + { + "epoch": 2.564390977443609, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 47633 + }, + { + "epoch": 2.5644511278195488, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 47634 + }, + { + "epoch": 2.564511278195489, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 47635 + }, + { + "epoch": 2.5645714285714285, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 47636 + }, + { + "epoch": 2.5646315789473686, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 47637 + }, + { + "epoch": 2.5646917293233082, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 47638 + }, + { + "epoch": 2.564751879699248, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 47639 + }, + { + "epoch": 2.564812030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 47640 + }, + { + "epoch": 2.564872180451128, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 47641 + }, + { + "epoch": 2.5649323308270677, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 47642 + }, + { + "epoch": 2.5649924812030074, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 47643 + }, + { + "epoch": 2.5650526315789475, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 47644 + }, + { + "epoch": 2.565112781954887, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 47645 + }, + { + "epoch": 2.565172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9174, + "step": 47646 + }, + { + "epoch": 2.565233082706767, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 47647 + }, + { + "epoch": 2.5652932330827065, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 47648 + }, + { + "epoch": 2.5653533834586466, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 47649 + }, + { + "epoch": 2.5654135338345867, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 47650 + }, + { + "epoch": 2.5654736842105264, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 47651 + }, + { + "epoch": 2.565533834586466, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 47652 + }, + { + "epoch": 2.565593984962406, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 47653 + }, + { + "epoch": 2.5656541353383457, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 47654 + }, + { + "epoch": 2.565714285714286, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 47655 + }, + { + "epoch": 2.5657744360902255, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9278, + "step": 47656 + }, + { + "epoch": 2.5658345864661656, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.944, + "step": 47657 + }, + { + "epoch": 2.5658947368421052, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 47658 + }, + { + "epoch": 2.5659548872180453, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 47659 + }, + { + "epoch": 2.566015037593985, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 47660 + }, + { + "epoch": 2.5660751879699246, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 47661 + }, + { + "epoch": 2.5661353383458647, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.15, + "step": 47662 + }, + { + "epoch": 2.5661954887218044, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 47663 + }, + { + "epoch": 2.5662556390977445, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 47664 + }, + { + "epoch": 2.566315789473684, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 47665 + }, + { + "epoch": 2.566375939849624, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 47666 + }, + { + "epoch": 2.566436090225564, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 47667 + }, + { + "epoch": 2.566496240601504, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 47668 + }, + { + "epoch": 2.5665563909774436, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 47669 + }, + { + "epoch": 2.5666165413533832, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 47670 + }, + { + "epoch": 2.5666766917293233, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 47671 + }, + { + "epoch": 2.5667368421052634, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 47672 + }, + { + "epoch": 2.566796992481203, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 47673 + }, + { + "epoch": 2.5668571428571427, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 47674 + }, + { + "epoch": 2.566917293233083, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 47675 + }, + { + "epoch": 2.5669774436090225, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 47676 + }, + { + "epoch": 2.5670375939849626, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 47677 + }, + { + "epoch": 2.567097744360902, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 47678 + }, + { + "epoch": 2.567157894736842, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 47679 + }, + { + "epoch": 2.567218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 47680 + }, + { + "epoch": 2.567278195488722, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 47681 + }, + { + "epoch": 2.5673383458646617, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.956, + "step": 47682 + }, + { + "epoch": 2.5673984962406013, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 47683 + }, + { + "epoch": 2.5674586466165414, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 47684 + }, + { + "epoch": 2.567518796992481, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 47685 + }, + { + "epoch": 2.567578947368421, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 47686 + }, + { + "epoch": 2.567639097744361, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 47687 + }, + { + "epoch": 2.5676992481203005, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 47688 + }, + { + "epoch": 2.5677593984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 47689 + }, + { + "epoch": 2.5678195488721807, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 47690 + }, + { + "epoch": 2.5678796992481203, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 47691 + }, + { + "epoch": 2.56793984962406, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 47692 + }, + { + "epoch": 2.568, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 47693 + }, + { + "epoch": 2.5680601503759397, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 47694 + }, + { + "epoch": 2.56812030075188, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 47695 + }, + { + "epoch": 2.5681804511278195, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 47696 + }, + { + "epoch": 2.5682406015037595, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 47697 + }, + { + "epoch": 2.568300751879699, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 47698 + }, + { + "epoch": 2.5683609022556393, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 47699 + }, + { + "epoch": 2.568421052631579, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 47700 + }, + { + "epoch": 2.5684812030075186, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 47701 + }, + { + "epoch": 2.5685413533834587, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 47702 + }, + { + "epoch": 2.5686015037593983, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 47703 + }, + { + "epoch": 2.5686616541353384, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 47704 + }, + { + "epoch": 2.568721804511278, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 47705 + }, + { + "epoch": 2.568781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 47706 + }, + { + "epoch": 2.568842105263158, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 47707 + }, + { + "epoch": 2.568902255639098, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 47708 + }, + { + "epoch": 2.5689624060150376, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 47709 + }, + { + "epoch": 2.569022556390977, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 47710 + }, + { + "epoch": 2.5690827067669173, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 47711 + }, + { + "epoch": 2.5691428571428574, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 47712 + }, + { + "epoch": 2.569203007518797, + "grad_norm": 1.5, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 47713 + }, + { + "epoch": 2.5692631578947367, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 47714 + }, + { + "epoch": 2.569323308270677, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 47715 + }, + { + "epoch": 2.5693834586466164, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 47716 + }, + { + "epoch": 2.5694436090225565, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 47717 + }, + { + "epoch": 2.569503759398496, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 47718 + }, + { + "epoch": 2.569563909774436, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 47719 + }, + { + "epoch": 2.569624060150376, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 47720 + }, + { + "epoch": 2.569684210526316, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 47721 + }, + { + "epoch": 2.5697443609022557, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 47722 + }, + { + "epoch": 2.5698045112781953, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 47723 + }, + { + "epoch": 2.5698646616541354, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 47724 + }, + { + "epoch": 2.569924812030075, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 47725 + }, + { + "epoch": 2.569984962406015, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 47726 + }, + { + "epoch": 2.570045112781955, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 47727 + }, + { + "epoch": 2.570105263157895, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 47728 + }, + { + "epoch": 2.5701654135338345, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 47729 + }, + { + "epoch": 2.5702255639097746, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 47730 + }, + { + "epoch": 2.5702857142857143, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 47731 + }, + { + "epoch": 2.570345864661654, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 47732 + }, + { + "epoch": 2.570406015037594, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 47733 + }, + { + "epoch": 2.5704661654135337, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 47734 + }, + { + "epoch": 2.5705263157894738, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 47735 + }, + { + "epoch": 2.5705864661654134, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 47736 + }, + { + "epoch": 2.5706466165413535, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 47737 + }, + { + "epoch": 2.570706766917293, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 47738 + }, + { + "epoch": 2.5707669172932333, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 47739 + }, + { + "epoch": 2.570827067669173, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 47740 + }, + { + "epoch": 2.5708872180451126, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 47741 + }, + { + "epoch": 2.5709473684210526, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 47742 + }, + { + "epoch": 2.5710075187969927, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 47743 + }, + { + "epoch": 2.5710676691729324, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 47744 + }, + { + "epoch": 2.571127819548872, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 47745 + }, + { + "epoch": 2.571187969924812, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 47746 + }, + { + "epoch": 2.571248120300752, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 47747 + }, + { + "epoch": 2.571308270676692, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.8952, + "step": 47748 + }, + { + "epoch": 2.5713684210526315, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 47749 + }, + { + "epoch": 2.571428571428571, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.939, + "step": 47750 + }, + { + "epoch": 2.5714887218045113, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 47751 + }, + { + "epoch": 2.5715488721804514, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 47752 + }, + { + "epoch": 2.571609022556391, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 47753 + }, + { + "epoch": 2.5716691729323307, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.2116, + "step": 47754 + }, + { + "epoch": 2.5717293233082708, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 47755 + }, + { + "epoch": 2.5717894736842104, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 47756 + }, + { + "epoch": 2.5718496240601505, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 47757 + }, + { + "epoch": 2.57190977443609, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 47758 + }, + { + "epoch": 2.5719699248120302, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 47759 + }, + { + "epoch": 2.57203007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 47760 + }, + { + "epoch": 2.57209022556391, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 47761 + }, + { + "epoch": 2.5721503759398496, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 47762 + }, + { + "epoch": 2.5722105263157893, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 47763 + }, + { + "epoch": 2.5722706766917294, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 47764 + }, + { + "epoch": 2.572330827067669, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 47765 + }, + { + "epoch": 2.572390977443609, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 47766 + }, + { + "epoch": 2.5724511278195488, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 47767 + }, + { + "epoch": 2.572511278195489, + "grad_norm": 1.3671875, + "learning_rate": 3e-05, + "loss": 0.9339, + "step": 47768 + }, + { + "epoch": 2.5725714285714285, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 47769 + }, + { + "epoch": 2.5726315789473686, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 47770 + }, + { + "epoch": 2.5726917293233083, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 47771 + }, + { + "epoch": 2.572751879699248, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 47772 + }, + { + "epoch": 2.572812030075188, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 47773 + }, + { + "epoch": 2.572872180451128, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9141, + "step": 47774 + }, + { + "epoch": 2.5729323308270677, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.932, + "step": 47775 + }, + { + "epoch": 2.5729924812030074, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 47776 + }, + { + "epoch": 2.5730526315789475, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 47777 + }, + { + "epoch": 2.573112781954887, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 47778 + }, + { + "epoch": 2.5731729323308272, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 47779 + }, + { + "epoch": 2.573233082706767, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9054, + "step": 47780 + }, + { + "epoch": 2.5732932330827065, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 47781 + }, + { + "epoch": 2.5733533834586466, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 47782 + }, + { + "epoch": 2.5734135338345867, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 47783 + }, + { + "epoch": 2.5734736842105264, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9313, + "step": 47784 + }, + { + "epoch": 2.573533834586466, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9433, + "step": 47785 + }, + { + "epoch": 2.573593984962406, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 47786 + }, + { + "epoch": 2.5736541353383457, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9269, + "step": 47787 + }, + { + "epoch": 2.573714285714286, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 47788 + }, + { + "epoch": 2.5737744360902255, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 47789 + }, + { + "epoch": 2.573834586466165, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 47790 + }, + { + "epoch": 2.5738947368421052, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 47791 + }, + { + "epoch": 2.5739548872180453, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 47792 + }, + { + "epoch": 2.574015037593985, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 47793 + }, + { + "epoch": 2.5740751879699246, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 47794 + }, + { + "epoch": 2.5741353383458647, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9302, + "step": 47795 + }, + { + "epoch": 2.5741954887218044, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 47796 + }, + { + "epoch": 2.5742556390977445, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 47797 + }, + { + "epoch": 2.574315789473684, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 47798 + }, + { + "epoch": 2.574375939849624, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 47799 + }, + { + "epoch": 2.574436090225564, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 47800 + }, + { + "epoch": 2.574496240601504, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 47801 + }, + { + "epoch": 2.5745563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 47802 + }, + { + "epoch": 2.5746165413533832, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 47803 + }, + { + "epoch": 2.5746766917293233, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 47804 + }, + { + "epoch": 2.5747368421052634, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 47805 + }, + { + "epoch": 2.574796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 47806 + }, + { + "epoch": 2.5748571428571427, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 47807 + }, + { + "epoch": 2.574917293233083, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.104, + "step": 47808 + }, + { + "epoch": 2.5749774436090225, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 47809 + }, + { + "epoch": 2.5750375939849626, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 47810 + }, + { + "epoch": 2.575097744360902, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 47811 + }, + { + "epoch": 2.575157894736842, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 47812 + }, + { + "epoch": 2.575218045112782, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.954, + "step": 47813 + }, + { + "epoch": 2.575278195488722, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 47814 + }, + { + "epoch": 2.5753383458646617, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 47815 + }, + { + "epoch": 2.5753984962406014, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 47816 + }, + { + "epoch": 2.5754586466165414, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 47817 + }, + { + "epoch": 2.575518796992481, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 47818 + }, + { + "epoch": 2.575578947368421, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 47819 + }, + { + "epoch": 2.575639097744361, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 47820 + }, + { + "epoch": 2.5756992481203005, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.157, + "step": 47821 + }, + { + "epoch": 2.5757593984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 47822 + }, + { + "epoch": 2.5758195488721807, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 47823 + }, + { + "epoch": 2.5758796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 47824 + }, + { + "epoch": 2.57593984962406, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 47825 + }, + { + "epoch": 2.576, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 47826 + }, + { + "epoch": 2.5760601503759397, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 47827 + }, + { + "epoch": 2.57612030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 47828 + }, + { + "epoch": 2.5761804511278195, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 47829 + }, + { + "epoch": 2.5762406015037596, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 47830 + }, + { + "epoch": 2.576300751879699, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 47831 + }, + { + "epoch": 2.5763609022556393, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 47832 + }, + { + "epoch": 2.576421052631579, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 47833 + }, + { + "epoch": 2.5764812030075186, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 47834 + }, + { + "epoch": 2.5765413533834587, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 47835 + }, + { + "epoch": 2.5766015037593983, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 47836 + }, + { + "epoch": 2.5766616541353384, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.259, + "step": 47837 + }, + { + "epoch": 2.576721804511278, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 47838 + }, + { + "epoch": 2.576781954887218, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 47839 + }, + { + "epoch": 2.576842105263158, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 47840 + }, + { + "epoch": 2.576902255639098, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 47841 + }, + { + "epoch": 2.5769624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 47842 + }, + { + "epoch": 2.577022556390977, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 47843 + }, + { + "epoch": 2.5770827067669173, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 47844 + }, + { + "epoch": 2.5771428571428574, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 47845 + }, + { + "epoch": 2.577203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 47846 + }, + { + "epoch": 2.5772631578947367, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 47847 + }, + { + "epoch": 2.577323308270677, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 47848 + }, + { + "epoch": 2.5773834586466164, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 47849 + }, + { + "epoch": 2.5774436090225565, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 47850 + }, + { + "epoch": 2.577503759398496, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 47851 + }, + { + "epoch": 2.577563909774436, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 47852 + }, + { + "epoch": 2.577624060150376, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9308, + "step": 47853 + }, + { + "epoch": 2.577684210526316, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 47854 + }, + { + "epoch": 2.5777443609022557, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 47855 + }, + { + "epoch": 2.5778045112781953, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 47856 + }, + { + "epoch": 2.5778646616541354, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 47857 + }, + { + "epoch": 2.577924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 47858 + }, + { + "epoch": 2.577984962406015, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1437, + "step": 47859 + }, + { + "epoch": 2.578045112781955, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 47860 + }, + { + "epoch": 2.578105263157895, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9505, + "step": 47861 + }, + { + "epoch": 2.5781654135338345, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9163, + "step": 47862 + }, + { + "epoch": 2.5782255639097746, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 47863 + }, + { + "epoch": 2.5782857142857143, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2314, + "step": 47864 + }, + { + "epoch": 2.578345864661654, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 47865 + }, + { + "epoch": 2.578406015037594, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 47866 + }, + { + "epoch": 2.5784661654135337, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 47867 + }, + { + "epoch": 2.5785263157894738, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 47868 + }, + { + "epoch": 2.5785864661654134, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 47869 + }, + { + "epoch": 2.5786466165413535, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 47870 + }, + { + "epoch": 2.578706766917293, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 47871 + }, + { + "epoch": 2.5787669172932333, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 47872 + }, + { + "epoch": 2.578827067669173, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 47873 + }, + { + "epoch": 2.5788872180451126, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 47874 + }, + { + "epoch": 2.5789473684210527, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 47875 + }, + { + "epoch": 2.5790075187969927, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 47876 + }, + { + "epoch": 2.5790676691729324, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1162, + "step": 47877 + }, + { + "epoch": 2.579127819548872, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 47878 + }, + { + "epoch": 2.579187969924812, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 47879 + }, + { + "epoch": 2.579248120300752, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 47880 + }, + { + "epoch": 2.579308270676692, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 47881 + }, + { + "epoch": 2.5793684210526315, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 47882 + }, + { + "epoch": 2.579428571428571, + "grad_norm": 1.4375, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 47883 + }, + { + "epoch": 2.5794887218045113, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 47884 + }, + { + "epoch": 2.5795488721804514, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 47885 + }, + { + "epoch": 2.579609022556391, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 47886 + }, + { + "epoch": 2.5796691729323307, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 47887 + }, + { + "epoch": 2.5797293233082708, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 47888 + }, + { + "epoch": 2.5797894736842104, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 47889 + }, + { + "epoch": 2.5798496240601505, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 47890 + }, + { + "epoch": 2.57990977443609, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 47891 + }, + { + "epoch": 2.57996992481203, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 47892 + }, + { + "epoch": 2.58003007518797, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2345, + "step": 47893 + }, + { + "epoch": 2.58009022556391, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 47894 + }, + { + "epoch": 2.5801503759398496, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 47895 + }, + { + "epoch": 2.5802105263157893, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 47896 + }, + { + "epoch": 2.5802706766917294, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 47897 + }, + { + "epoch": 2.580330827067669, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 47898 + }, + { + "epoch": 2.580390977443609, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 47899 + }, + { + "epoch": 2.5804511278195488, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 47900 + }, + { + "epoch": 2.580511278195489, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 47901 + }, + { + "epoch": 2.5805714285714285, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 47902 + }, + { + "epoch": 2.5806315789473686, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 47903 + }, + { + "epoch": 2.5806917293233083, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 47904 + }, + { + "epoch": 2.580751879699248, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 47905 + }, + { + "epoch": 2.580812030075188, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 47906 + }, + { + "epoch": 2.580872180451128, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 47907 + }, + { + "epoch": 2.5809323308270677, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 47908 + }, + { + "epoch": 2.5809924812030074, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.8845, + "step": 47909 + }, + { + "epoch": 2.5810526315789475, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 47910 + }, + { + "epoch": 2.581112781954887, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 47911 + }, + { + "epoch": 2.5811729323308272, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9213, + "step": 47912 + }, + { + "epoch": 2.581233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 47913 + }, + { + "epoch": 2.5812932330827065, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 47914 + }, + { + "epoch": 2.5813533834586466, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 47915 + }, + { + "epoch": 2.5814135338345867, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 47916 + }, + { + "epoch": 2.5814736842105264, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 47917 + }, + { + "epoch": 2.581533834586466, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 47918 + }, + { + "epoch": 2.581593984962406, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 47919 + }, + { + "epoch": 2.5816541353383458, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1877, + "step": 47920 + }, + { + "epoch": 2.581714285714286, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 47921 + }, + { + "epoch": 2.5817744360902255, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 47922 + }, + { + "epoch": 2.581834586466165, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 47923 + }, + { + "epoch": 2.5818947368421052, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 47924 + }, + { + "epoch": 2.5819548872180453, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 47925 + }, + { + "epoch": 2.582015037593985, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 47926 + }, + { + "epoch": 2.5820751879699246, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 47927 + }, + { + "epoch": 2.5821353383458647, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 47928 + }, + { + "epoch": 2.5821954887218044, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 47929 + }, + { + "epoch": 2.5822556390977445, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 47930 + }, + { + "epoch": 2.582315789473684, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 47931 + }, + { + "epoch": 2.582375939849624, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 47932 + }, + { + "epoch": 2.582436090225564, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 47933 + }, + { + "epoch": 2.582496240601504, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 47934 + }, + { + "epoch": 2.5825563909774436, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 47935 + }, + { + "epoch": 2.5826165413533833, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 47936 + }, + { + "epoch": 2.5826766917293233, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 47937 + }, + { + "epoch": 2.582736842105263, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 47938 + }, + { + "epoch": 2.582796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 47939 + }, + { + "epoch": 2.5828571428571427, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 47940 + }, + { + "epoch": 2.582917293233083, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9079, + "step": 47941 + }, + { + "epoch": 2.5829774436090225, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 47942 + }, + { + "epoch": 2.5830375939849626, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 47943 + }, + { + "epoch": 2.5830977443609022, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 47944 + }, + { + "epoch": 2.583157894736842, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 47945 + }, + { + "epoch": 2.583218045112782, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 47946 + }, + { + "epoch": 2.583278195488722, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 47947 + }, + { + "epoch": 2.5833383458646617, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 47948 + }, + { + "epoch": 2.5833984962406014, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 47949 + }, + { + "epoch": 2.5834586466165415, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 47950 + }, + { + "epoch": 2.583518796992481, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 47951 + }, + { + "epoch": 2.583578947368421, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 47952 + }, + { + "epoch": 2.583639097744361, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 47953 + }, + { + "epoch": 2.5836992481203005, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 47954 + }, + { + "epoch": 2.5837593984962406, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 47955 + }, + { + "epoch": 2.5838195488721807, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 47956 + }, + { + "epoch": 2.5838796992481203, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 47957 + }, + { + "epoch": 2.58393984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 47958 + }, + { + "epoch": 2.584, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 47959 + }, + { + "epoch": 2.5840601503759397, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 47960 + }, + { + "epoch": 2.58412030075188, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 47961 + }, + { + "epoch": 2.5841804511278195, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 47962 + }, + { + "epoch": 2.5842406015037596, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 47963 + }, + { + "epoch": 2.584300751879699, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 47964 + }, + { + "epoch": 2.5843609022556393, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 47965 + }, + { + "epoch": 2.584421052631579, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 47966 + }, + { + "epoch": 2.5844812030075186, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 47967 + }, + { + "epoch": 2.5845413533834587, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 47968 + }, + { + "epoch": 2.5846015037593983, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 47969 + }, + { + "epoch": 2.5846616541353384, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9433, + "step": 47970 + }, + { + "epoch": 2.584721804511278, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 47971 + }, + { + "epoch": 2.584781954887218, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 47972 + }, + { + "epoch": 2.584842105263158, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1351, + "step": 47973 + }, + { + "epoch": 2.584902255639098, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.3039, + "step": 47974 + }, + { + "epoch": 2.5849624060150376, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 47975 + }, + { + "epoch": 2.585022556390977, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 47976 + }, + { + "epoch": 2.5850827067669173, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 47977 + }, + { + "epoch": 2.5851428571428574, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 47978 + }, + { + "epoch": 2.585203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1447, + "step": 47979 + }, + { + "epoch": 2.5852631578947367, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9024, + "step": 47980 + }, + { + "epoch": 2.585323308270677, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 47981 + }, + { + "epoch": 2.5853834586466165, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9491, + "step": 47982 + }, + { + "epoch": 2.5854436090225565, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1785, + "step": 47983 + }, + { + "epoch": 2.585503759398496, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 47984 + }, + { + "epoch": 2.585563909774436, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 47985 + }, + { + "epoch": 2.585624060150376, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 47986 + }, + { + "epoch": 2.585684210526316, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 47987 + }, + { + "epoch": 2.5857443609022557, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 47988 + }, + { + "epoch": 2.5858045112781953, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 47989 + }, + { + "epoch": 2.5858646616541354, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 47990 + }, + { + "epoch": 2.585924812030075, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 47991 + }, + { + "epoch": 2.585984962406015, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 47992 + }, + { + "epoch": 2.586045112781955, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 47993 + }, + { + "epoch": 2.5861052631578945, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 47994 + }, + { + "epoch": 2.5861654135338346, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 47995 + }, + { + "epoch": 2.5862255639097746, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 47996 + }, + { + "epoch": 2.5862857142857143, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 47997 + }, + { + "epoch": 2.586345864661654, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 47998 + }, + { + "epoch": 2.586406015037594, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 47999 + }, + { + "epoch": 2.5864661654135337, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 48000 + }, + { + "epoch": 2.586526315789474, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 48001 + }, + { + "epoch": 2.5865864661654134, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 48002 + }, + { + "epoch": 2.5866466165413535, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1281, + "step": 48003 + }, + { + "epoch": 2.586706766917293, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 48004 + }, + { + "epoch": 2.5867669172932333, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 48005 + }, + { + "epoch": 2.586827067669173, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 48006 + }, + { + "epoch": 2.5868872180451126, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 48007 + }, + { + "epoch": 2.5869473684210527, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 48008 + }, + { + "epoch": 2.5870075187969928, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9316, + "step": 48009 + }, + { + "epoch": 2.5870676691729324, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 48010 + }, + { + "epoch": 2.587127819548872, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 48011 + }, + { + "epoch": 2.587187969924812, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 48012 + }, + { + "epoch": 2.587248120300752, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 48013 + }, + { + "epoch": 2.587308270676692, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 48014 + }, + { + "epoch": 2.5873684210526315, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 48015 + }, + { + "epoch": 2.587428571428571, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 48016 + }, + { + "epoch": 2.5874887218045113, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 48017 + }, + { + "epoch": 2.5875488721804514, + "grad_norm": 3.453125, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 48018 + }, + { + "epoch": 2.587609022556391, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 48019 + }, + { + "epoch": 2.5876691729323307, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 48020 + }, + { + "epoch": 2.5877293233082708, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 48021 + }, + { + "epoch": 2.5877894736842104, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 48022 + }, + { + "epoch": 2.5878496240601505, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 48023 + }, + { + "epoch": 2.58790977443609, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 48024 + }, + { + "epoch": 2.58796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 48025 + }, + { + "epoch": 2.58803007518797, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2021, + "step": 48026 + }, + { + "epoch": 2.58809022556391, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 48027 + }, + { + "epoch": 2.5881503759398496, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 48028 + }, + { + "epoch": 2.5882105263157893, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 48029 + }, + { + "epoch": 2.5882706766917294, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1198, + "step": 48030 + }, + { + "epoch": 2.588330827067669, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 48031 + }, + { + "epoch": 2.588390977443609, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 48032 + }, + { + "epoch": 2.588451127819549, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 48033 + }, + { + "epoch": 2.588511278195489, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1989, + "step": 48034 + }, + { + "epoch": 2.5885714285714285, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 48035 + }, + { + "epoch": 2.5886315789473686, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 48036 + }, + { + "epoch": 2.5886917293233083, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 48037 + }, + { + "epoch": 2.588751879699248, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 48038 + }, + { + "epoch": 2.588812030075188, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 48039 + }, + { + "epoch": 2.5888721804511277, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 48040 + }, + { + "epoch": 2.5889323308270678, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9272, + "step": 48041 + }, + { + "epoch": 2.5889924812030074, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1674, + "step": 48042 + }, + { + "epoch": 2.5890526315789475, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 48043 + }, + { + "epoch": 2.589112781954887, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.954, + "step": 48044 + }, + { + "epoch": 2.5891729323308272, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 48045 + }, + { + "epoch": 2.589233082706767, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 48046 + }, + { + "epoch": 2.5892932330827065, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9128, + "step": 48047 + }, + { + "epoch": 2.5893533834586466, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 48048 + }, + { + "epoch": 2.5894135338345867, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 48049 + }, + { + "epoch": 2.5894736842105264, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 48050 + }, + { + "epoch": 2.589533834586466, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 48051 + }, + { + "epoch": 2.589593984962406, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 48052 + }, + { + "epoch": 2.5896541353383458, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.2873, + "step": 48053 + }, + { + "epoch": 2.589714285714286, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 48054 + }, + { + "epoch": 2.5897744360902255, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1264, + "step": 48055 + }, + { + "epoch": 2.589834586466165, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 48056 + }, + { + "epoch": 2.5898947368421052, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 48057 + }, + { + "epoch": 2.5899548872180453, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 48058 + }, + { + "epoch": 2.590015037593985, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 48059 + }, + { + "epoch": 2.5900751879699246, + "grad_norm": 1.90625, + "learning_rate": 3e-05, + "loss": 1.0997, + "step": 48060 + }, + { + "epoch": 2.5901353383458647, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 48061 + }, + { + "epoch": 2.5901954887218044, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 48062 + }, + { + "epoch": 2.5902556390977445, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 48063 + }, + { + "epoch": 2.590315789473684, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 48064 + }, + { + "epoch": 2.590375939849624, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 48065 + }, + { + "epoch": 2.590436090225564, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 48066 + }, + { + "epoch": 2.590496240601504, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 48067 + }, + { + "epoch": 2.5905563909774436, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 48068 + }, + { + "epoch": 2.5906165413533833, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 48069 + }, + { + "epoch": 2.5906766917293234, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9325, + "step": 48070 + }, + { + "epoch": 2.590736842105263, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 48071 + }, + { + "epoch": 2.590796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 48072 + }, + { + "epoch": 2.5908571428571427, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.944, + "step": 48073 + }, + { + "epoch": 2.590917293233083, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 48074 + }, + { + "epoch": 2.5909774436090225, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 48075 + }, + { + "epoch": 2.5910375939849626, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 48076 + }, + { + "epoch": 2.5910977443609022, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 48077 + }, + { + "epoch": 2.591157894736842, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 48078 + }, + { + "epoch": 2.591218045112782, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 48079 + }, + { + "epoch": 2.591278195488722, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.2268, + "step": 48080 + }, + { + "epoch": 2.5913383458646617, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9019, + "step": 48081 + }, + { + "epoch": 2.5913984962406014, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 48082 + }, + { + "epoch": 2.5914586466165415, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 48083 + }, + { + "epoch": 2.591518796992481, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 48084 + }, + { + "epoch": 2.591578947368421, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 48085 + }, + { + "epoch": 2.591639097744361, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 48086 + }, + { + "epoch": 2.5916992481203005, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 48087 + }, + { + "epoch": 2.5917593984962406, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 48088 + }, + { + "epoch": 2.5918195488721807, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 48089 + }, + { + "epoch": 2.5918796992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 48090 + }, + { + "epoch": 2.59193984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 48091 + }, + { + "epoch": 2.592, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 48092 + }, + { + "epoch": 2.5920601503759397, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 48093 + }, + { + "epoch": 2.59212030075188, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 48094 + }, + { + "epoch": 2.5921804511278195, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 48095 + }, + { + "epoch": 2.592240601503759, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 48096 + }, + { + "epoch": 2.592300751879699, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 48097 + }, + { + "epoch": 2.5923609022556393, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 48098 + }, + { + "epoch": 2.592421052631579, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 48099 + }, + { + "epoch": 2.5924812030075186, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 48100 + }, + { + "epoch": 2.5925413533834587, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 48101 + }, + { + "epoch": 2.5926015037593984, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 48102 + }, + { + "epoch": 2.5926616541353384, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 48103 + }, + { + "epoch": 2.592721804511278, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 48104 + }, + { + "epoch": 2.592781954887218, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 48105 + }, + { + "epoch": 2.592842105263158, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 48106 + }, + { + "epoch": 2.592902255639098, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 48107 + }, + { + "epoch": 2.5929624060150376, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 48108 + }, + { + "epoch": 2.5930225563909772, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 48109 + }, + { + "epoch": 2.5930827067669173, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 48110 + }, + { + "epoch": 2.5931428571428574, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 48111 + }, + { + "epoch": 2.593203007518797, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 48112 + }, + { + "epoch": 2.5932631578947367, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 48113 + }, + { + "epoch": 2.593323308270677, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 48114 + }, + { + "epoch": 2.5933834586466165, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 48115 + }, + { + "epoch": 2.5934436090225566, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 48116 + }, + { + "epoch": 2.593503759398496, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9357, + "step": 48117 + }, + { + "epoch": 2.593563909774436, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 48118 + }, + { + "epoch": 2.593624060150376, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 48119 + }, + { + "epoch": 2.593684210526316, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 48120 + }, + { + "epoch": 2.5937443609022557, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 48121 + }, + { + "epoch": 2.5938045112781953, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 48122 + }, + { + "epoch": 2.5938646616541354, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 48123 + }, + { + "epoch": 2.593924812030075, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 48124 + }, + { + "epoch": 2.593984962406015, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 48125 + }, + { + "epoch": 2.594045112781955, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 48126 + }, + { + "epoch": 2.5941052631578945, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 48127 + }, + { + "epoch": 2.5941654135338346, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 48128 + }, + { + "epoch": 2.5942255639097747, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 48129 + }, + { + "epoch": 2.5942857142857143, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9298, + "step": 48130 + }, + { + "epoch": 2.594345864661654, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 48131 + }, + { + "epoch": 2.594406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 48132 + }, + { + "epoch": 2.5944661654135337, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 48133 + }, + { + "epoch": 2.594526315789474, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 48134 + }, + { + "epoch": 2.5945864661654134, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 48135 + }, + { + "epoch": 2.5946466165413535, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 48136 + }, + { + "epoch": 2.594706766917293, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 48137 + }, + { + "epoch": 2.5947669172932333, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 48138 + }, + { + "epoch": 2.594827067669173, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 48139 + }, + { + "epoch": 2.5948872180451126, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 48140 + }, + { + "epoch": 2.5949473684210527, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 48141 + }, + { + "epoch": 2.5950075187969923, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 48142 + }, + { + "epoch": 2.5950676691729324, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 48143 + }, + { + "epoch": 2.595127819548872, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 48144 + }, + { + "epoch": 2.595187969924812, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 48145 + }, + { + "epoch": 2.595248120300752, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 48146 + }, + { + "epoch": 2.595308270676692, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 48147 + }, + { + "epoch": 2.5953684210526315, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 48148 + }, + { + "epoch": 2.595428571428571, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 48149 + }, + { + "epoch": 2.5954887218045113, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 48150 + }, + { + "epoch": 2.5955488721804514, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 48151 + }, + { + "epoch": 2.595609022556391, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 48152 + }, + { + "epoch": 2.5956691729323307, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 48153 + }, + { + "epoch": 2.5957293233082708, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 48154 + }, + { + "epoch": 2.5957894736842104, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 48155 + }, + { + "epoch": 2.5958496240601505, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 48156 + }, + { + "epoch": 2.59590977443609, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 48157 + }, + { + "epoch": 2.59596992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 48158 + }, + { + "epoch": 2.59603007518797, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 48159 + }, + { + "epoch": 2.59609022556391, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 48160 + }, + { + "epoch": 2.5961503759398497, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 48161 + }, + { + "epoch": 2.5962105263157893, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 48162 + }, + { + "epoch": 2.5962706766917294, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 48163 + }, + { + "epoch": 2.596330827067669, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1272, + "step": 48164 + }, + { + "epoch": 2.596390977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 48165 + }, + { + "epoch": 2.596451127819549, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 48166 + }, + { + "epoch": 2.596511278195489, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1899, + "step": 48167 + }, + { + "epoch": 2.5965714285714285, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 48168 + }, + { + "epoch": 2.5966315789473686, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 48169 + }, + { + "epoch": 2.5966917293233083, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 48170 + }, + { + "epoch": 2.596751879699248, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 48171 + }, + { + "epoch": 2.596812030075188, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 48172 + }, + { + "epoch": 2.5968721804511277, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 48173 + }, + { + "epoch": 2.5969323308270678, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 48174 + }, + { + "epoch": 2.5969924812030074, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 48175 + }, + { + "epoch": 2.5970526315789475, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 48176 + }, + { + "epoch": 2.597112781954887, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 48177 + }, + { + "epoch": 2.5971729323308272, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9395, + "step": 48178 + }, + { + "epoch": 2.597233082706767, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 48179 + }, + { + "epoch": 2.5972932330827065, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 48180 + }, + { + "epoch": 2.5973533834586466, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 48181 + }, + { + "epoch": 2.5974135338345867, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9266, + "step": 48182 + }, + { + "epoch": 2.5974736842105264, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 48183 + }, + { + "epoch": 2.597533834586466, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 48184 + }, + { + "epoch": 2.597593984962406, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 48185 + }, + { + "epoch": 2.5976541353383458, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 48186 + }, + { + "epoch": 2.597714285714286, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 48187 + }, + { + "epoch": 2.5977744360902255, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 48188 + }, + { + "epoch": 2.597834586466165, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 48189 + }, + { + "epoch": 2.5978947368421053, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 48190 + }, + { + "epoch": 2.5979548872180454, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 48191 + }, + { + "epoch": 2.598015037593985, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 48192 + }, + { + "epoch": 2.5980751879699246, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 48193 + }, + { + "epoch": 2.5981353383458647, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 48194 + }, + { + "epoch": 2.5981954887218044, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 48195 + }, + { + "epoch": 2.5982556390977445, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 48196 + }, + { + "epoch": 2.598315789473684, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 48197 + }, + { + "epoch": 2.598375939849624, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 48198 + }, + { + "epoch": 2.598436090225564, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 48199 + }, + { + "epoch": 2.598496240601504, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 48200 + }, + { + "epoch": 2.5985563909774436, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 48201 + }, + { + "epoch": 2.5986165413533833, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 48202 + }, + { + "epoch": 2.5986766917293234, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 48203 + }, + { + "epoch": 2.598736842105263, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 48204 + }, + { + "epoch": 2.598796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 48205 + }, + { + "epoch": 2.5988571428571428, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 48206 + }, + { + "epoch": 2.598917293233083, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 48207 + }, + { + "epoch": 2.5989774436090225, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 48208 + }, + { + "epoch": 2.5990375939849626, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 48209 + }, + { + "epoch": 2.5990977443609022, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 48210 + }, + { + "epoch": 2.599157894736842, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 48211 + }, + { + "epoch": 2.599218045112782, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 48212 + }, + { + "epoch": 2.599278195488722, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 48213 + }, + { + "epoch": 2.5993383458646617, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 48214 + }, + { + "epoch": 2.5993984962406014, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 48215 + }, + { + "epoch": 2.5994586466165415, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 48216 + }, + { + "epoch": 2.599518796992481, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 48217 + }, + { + "epoch": 2.599578947368421, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 48218 + }, + { + "epoch": 2.599639097744361, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 48219 + }, + { + "epoch": 2.5996992481203005, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 48220 + }, + { + "epoch": 2.5997593984962406, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 48221 + }, + { + "epoch": 2.5998195488721807, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 48222 + }, + { + "epoch": 2.5998796992481203, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 48223 + }, + { + "epoch": 2.59993984962406, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 48224 + }, + { + "epoch": 2.6, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 48225 + }, + { + "epoch": 2.6000601503759397, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 48226 + }, + { + "epoch": 2.60012030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 48227 + }, + { + "epoch": 2.6001804511278195, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 48228 + }, + { + "epoch": 2.600240601503759, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 48229 + }, + { + "epoch": 2.6003007518796992, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 48230 + }, + { + "epoch": 2.6003609022556393, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 48231 + }, + { + "epoch": 2.600421052631579, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 48232 + }, + { + "epoch": 2.6004812030075186, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 48233 + }, + { + "epoch": 2.6005413533834587, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 48234 + }, + { + "epoch": 2.6006015037593984, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 48235 + }, + { + "epoch": 2.6006616541353385, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 48236 + }, + { + "epoch": 2.600721804511278, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 48237 + }, + { + "epoch": 2.600781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 48238 + }, + { + "epoch": 2.600842105263158, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 48239 + }, + { + "epoch": 2.600902255639098, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 48240 + }, + { + "epoch": 2.6009624060150376, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9177, + "step": 48241 + }, + { + "epoch": 2.6010225563909772, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 48242 + }, + { + "epoch": 2.6010827067669173, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 48243 + }, + { + "epoch": 2.601142857142857, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 48244 + }, + { + "epoch": 2.601203007518797, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 48245 + }, + { + "epoch": 2.6012631578947367, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 48246 + }, + { + "epoch": 2.601323308270677, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 48247 + }, + { + "epoch": 2.6013834586466165, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 48248 + }, + { + "epoch": 2.6014436090225566, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 48249 + }, + { + "epoch": 2.601503759398496, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 48250 + }, + { + "epoch": 2.601563909774436, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 48251 + }, + { + "epoch": 2.601624060150376, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 48252 + }, + { + "epoch": 2.601684210526316, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 48253 + }, + { + "epoch": 2.6017443609022557, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 48254 + }, + { + "epoch": 2.6018045112781953, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 48255 + }, + { + "epoch": 2.6018646616541354, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 48256 + }, + { + "epoch": 2.601924812030075, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 48257 + }, + { + "epoch": 2.601984962406015, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 48258 + }, + { + "epoch": 2.602045112781955, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 48259 + }, + { + "epoch": 2.6021052631578945, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 48260 + }, + { + "epoch": 2.6021654135338346, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 48261 + }, + { + "epoch": 2.6022255639097747, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 48262 + }, + { + "epoch": 2.6022857142857143, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 48263 + }, + { + "epoch": 2.602345864661654, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 48264 + }, + { + "epoch": 2.602406015037594, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 48265 + }, + { + "epoch": 2.6024661654135337, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 48266 + }, + { + "epoch": 2.602526315789474, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 48267 + }, + { + "epoch": 2.6025864661654134, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 48268 + }, + { + "epoch": 2.6026466165413535, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 48269 + }, + { + "epoch": 2.602706766917293, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 48270 + }, + { + "epoch": 2.6027669172932333, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 48271 + }, + { + "epoch": 2.602827067669173, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 48272 + }, + { + "epoch": 2.6028872180451126, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 48273 + }, + { + "epoch": 2.6029473684210527, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 48274 + }, + { + "epoch": 2.6030075187969923, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 48275 + }, + { + "epoch": 2.6030676691729324, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 48276 + }, + { + "epoch": 2.603127819548872, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 48277 + }, + { + "epoch": 2.603187969924812, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 48278 + }, + { + "epoch": 2.603248120300752, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 48279 + }, + { + "epoch": 2.603308270676692, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 48280 + }, + { + "epoch": 2.6033684210526316, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 48281 + }, + { + "epoch": 2.603428571428571, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 48282 + }, + { + "epoch": 2.6034887218045113, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 48283 + }, + { + "epoch": 2.6035488721804514, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 48284 + }, + { + "epoch": 2.603609022556391, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 48285 + }, + { + "epoch": 2.6036691729323307, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.148, + "step": 48286 + }, + { + "epoch": 2.603729323308271, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 48287 + }, + { + "epoch": 2.6037894736842104, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9361, + "step": 48288 + }, + { + "epoch": 2.6038496240601505, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 48289 + }, + { + "epoch": 2.60390977443609, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 48290 + }, + { + "epoch": 2.60396992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 48291 + }, + { + "epoch": 2.60403007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 48292 + }, + { + "epoch": 2.60409022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 48293 + }, + { + "epoch": 2.6041503759398497, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 48294 + }, + { + "epoch": 2.6042105263157893, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 48295 + }, + { + "epoch": 2.6042706766917294, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 48296 + }, + { + "epoch": 2.604330827067669, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 48297 + }, + { + "epoch": 2.604390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 48298 + }, + { + "epoch": 2.604451127819549, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 48299 + }, + { + "epoch": 2.604511278195489, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 48300 + }, + { + "epoch": 2.6045714285714285, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 48301 + }, + { + "epoch": 2.6046315789473686, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1309, + "step": 48302 + }, + { + "epoch": 2.6046917293233083, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 48303 + }, + { + "epoch": 2.604751879699248, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 48304 + }, + { + "epoch": 2.604812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 48305 + }, + { + "epoch": 2.6048721804511277, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 48306 + }, + { + "epoch": 2.6049323308270678, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 48307 + }, + { + "epoch": 2.6049924812030074, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 48308 + }, + { + "epoch": 2.6050526315789475, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 48309 + }, + { + "epoch": 2.605112781954887, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 48310 + }, + { + "epoch": 2.6051729323308273, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 48311 + }, + { + "epoch": 2.605233082706767, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 48312 + }, + { + "epoch": 2.6052932330827065, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 48313 + }, + { + "epoch": 2.6053533834586466, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 48314 + }, + { + "epoch": 2.6054135338345867, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9428, + "step": 48315 + }, + { + "epoch": 2.6054736842105264, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 48316 + }, + { + "epoch": 2.605533834586466, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 48317 + }, + { + "epoch": 2.605593984962406, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 48318 + }, + { + "epoch": 2.605654135338346, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 48319 + }, + { + "epoch": 2.605714285714286, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 48320 + }, + { + "epoch": 2.6057744360902255, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 48321 + }, + { + "epoch": 2.605834586466165, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 48322 + }, + { + "epoch": 2.6058947368421053, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 48323 + }, + { + "epoch": 2.6059548872180454, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 48324 + }, + { + "epoch": 2.606015037593985, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 48325 + }, + { + "epoch": 2.6060751879699247, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 48326 + }, + { + "epoch": 2.6061353383458647, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 48327 + }, + { + "epoch": 2.6061954887218044, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 48328 + }, + { + "epoch": 2.6062556390977445, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 48329 + }, + { + "epoch": 2.606315789473684, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 48330 + }, + { + "epoch": 2.606375939849624, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 48331 + }, + { + "epoch": 2.606436090225564, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 48332 + }, + { + "epoch": 2.606496240601504, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9443, + "step": 48333 + }, + { + "epoch": 2.6065563909774436, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 48334 + }, + { + "epoch": 2.6066165413533833, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 48335 + }, + { + "epoch": 2.6066766917293234, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 48336 + }, + { + "epoch": 2.606736842105263, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 48337 + }, + { + "epoch": 2.606796992481203, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 48338 + }, + { + "epoch": 2.6068571428571428, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 48339 + }, + { + "epoch": 2.606917293233083, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 48340 + }, + { + "epoch": 2.6069774436090225, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 48341 + }, + { + "epoch": 2.6070375939849626, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 48342 + }, + { + "epoch": 2.6070977443609022, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 48343 + }, + { + "epoch": 2.607157894736842, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 48344 + }, + { + "epoch": 2.607218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 48345 + }, + { + "epoch": 2.6072781954887216, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 48346 + }, + { + "epoch": 2.6073383458646617, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 48347 + }, + { + "epoch": 2.6073984962406014, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 48348 + }, + { + "epoch": 2.6074586466165415, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 48349 + }, + { + "epoch": 2.607518796992481, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 48350 + }, + { + "epoch": 2.607578947368421, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 48351 + }, + { + "epoch": 2.607639097744361, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 48352 + }, + { + "epoch": 2.6076992481203005, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 48353 + }, + { + "epoch": 2.6077593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 48354 + }, + { + "epoch": 2.6078195488721807, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 48355 + }, + { + "epoch": 2.6078796992481204, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 48356 + }, + { + "epoch": 2.60793984962406, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 48357 + }, + { + "epoch": 2.608, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 48358 + }, + { + "epoch": 2.6080601503759397, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9167, + "step": 48359 + }, + { + "epoch": 2.60812030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 48360 + }, + { + "epoch": 2.6081804511278195, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 48361 + }, + { + "epoch": 2.608240601503759, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.911, + "step": 48362 + }, + { + "epoch": 2.6083007518796992, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 48363 + }, + { + "epoch": 2.6083609022556393, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 48364 + }, + { + "epoch": 2.608421052631579, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 48365 + }, + { + "epoch": 2.6084812030075186, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 48366 + }, + { + "epoch": 2.6085413533834587, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 48367 + }, + { + "epoch": 2.6086015037593984, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 48368 + }, + { + "epoch": 2.6086616541353385, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 48369 + }, + { + "epoch": 2.608721804511278, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 48370 + }, + { + "epoch": 2.608781954887218, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 48371 + }, + { + "epoch": 2.608842105263158, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 48372 + }, + { + "epoch": 2.608902255639098, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 48373 + }, + { + "epoch": 2.6089624060150376, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 48374 + }, + { + "epoch": 2.6090225563909772, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 48375 + }, + { + "epoch": 2.6090827067669173, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 48376 + }, + { + "epoch": 2.609142857142857, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 48377 + }, + { + "epoch": 2.609203007518797, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 48378 + }, + { + "epoch": 2.6092631578947367, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 48379 + }, + { + "epoch": 2.609323308270677, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 48380 + }, + { + "epoch": 2.6093834586466165, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 48381 + }, + { + "epoch": 2.6094436090225566, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 48382 + }, + { + "epoch": 2.609503759398496, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 48383 + }, + { + "epoch": 2.609563909774436, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 48384 + }, + { + "epoch": 2.609624060150376, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 48385 + }, + { + "epoch": 2.609684210526316, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 48386 + }, + { + "epoch": 2.6097443609022557, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 48387 + }, + { + "epoch": 2.6098045112781953, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 48388 + }, + { + "epoch": 2.6098646616541354, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 48389 + }, + { + "epoch": 2.609924812030075, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9352, + "step": 48390 + }, + { + "epoch": 2.609984962406015, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 48391 + }, + { + "epoch": 2.610045112781955, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9228, + "step": 48392 + }, + { + "epoch": 2.6101052631578945, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 48393 + }, + { + "epoch": 2.6101654135338346, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9209, + "step": 48394 + }, + { + "epoch": 2.6102255639097747, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 48395 + }, + { + "epoch": 2.6102857142857143, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 48396 + }, + { + "epoch": 2.610345864661654, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 48397 + }, + { + "epoch": 2.610406015037594, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 48398 + }, + { + "epoch": 2.6104661654135337, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 48399 + }, + { + "epoch": 2.610526315789474, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 48400 + }, + { + "epoch": 2.6105864661654135, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 48401 + }, + { + "epoch": 2.6106466165413535, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 48402 + }, + { + "epoch": 2.610706766917293, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 48403 + }, + { + "epoch": 2.6107669172932333, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 48404 + }, + { + "epoch": 2.610827067669173, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9233, + "step": 48405 + }, + { + "epoch": 2.6108872180451126, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 48406 + }, + { + "epoch": 2.6109473684210527, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 48407 + }, + { + "epoch": 2.6110075187969923, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 48408 + }, + { + "epoch": 2.6110676691729324, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 48409 + }, + { + "epoch": 2.611127819548872, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 48410 + }, + { + "epoch": 2.611187969924812, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 48411 + }, + { + "epoch": 2.611248120300752, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9314, + "step": 48412 + }, + { + "epoch": 2.611308270676692, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 48413 + }, + { + "epoch": 2.6113684210526316, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 48414 + }, + { + "epoch": 2.611428571428571, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 48415 + }, + { + "epoch": 2.6114887218045113, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1286, + "step": 48416 + }, + { + "epoch": 2.6115488721804514, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 48417 + }, + { + "epoch": 2.611609022556391, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 48418 + }, + { + "epoch": 2.6116691729323307, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 48419 + }, + { + "epoch": 2.611729323308271, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 48420 + }, + { + "epoch": 2.6117894736842104, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 48421 + }, + { + "epoch": 2.6118496240601505, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 48422 + }, + { + "epoch": 2.61190977443609, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 48423 + }, + { + "epoch": 2.61196992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 48424 + }, + { + "epoch": 2.61203007518797, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 48425 + }, + { + "epoch": 2.61209022556391, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 48426 + }, + { + "epoch": 2.6121503759398497, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 48427 + }, + { + "epoch": 2.6122105263157893, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 48428 + }, + { + "epoch": 2.6122706766917294, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 48429 + }, + { + "epoch": 2.612330827067669, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 48430 + }, + { + "epoch": 2.612390977443609, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 48431 + }, + { + "epoch": 2.612451127819549, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 48432 + }, + { + "epoch": 2.6125112781954885, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 48433 + }, + { + "epoch": 2.6125714285714285, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 48434 + }, + { + "epoch": 2.6126315789473686, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 48435 + }, + { + "epoch": 2.6126917293233083, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 48436 + }, + { + "epoch": 2.612751879699248, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 48437 + }, + { + "epoch": 2.612812030075188, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 48438 + }, + { + "epoch": 2.6128721804511277, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 48439 + }, + { + "epoch": 2.6129323308270678, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 48440 + }, + { + "epoch": 2.6129924812030074, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 48441 + }, + { + "epoch": 2.6130526315789475, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9479, + "step": 48442 + }, + { + "epoch": 2.613112781954887, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 48443 + }, + { + "epoch": 2.6131729323308273, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 48444 + }, + { + "epoch": 2.613233082706767, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.941, + "step": 48445 + }, + { + "epoch": 2.6132932330827066, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 48446 + }, + { + "epoch": 2.6133533834586467, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 48447 + }, + { + "epoch": 2.6134135338345863, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.936, + "step": 48448 + }, + { + "epoch": 2.6134736842105264, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 48449 + }, + { + "epoch": 2.613533834586466, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9001, + "step": 48450 + }, + { + "epoch": 2.613593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 48451 + }, + { + "epoch": 2.613654135338346, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 48452 + }, + { + "epoch": 2.613714285714286, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 48453 + }, + { + "epoch": 2.6137744360902255, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 48454 + }, + { + "epoch": 2.613834586466165, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 48455 + }, + { + "epoch": 2.6138947368421053, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1397, + "step": 48456 + }, + { + "epoch": 2.6139548872180454, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 48457 + }, + { + "epoch": 2.614015037593985, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 48458 + }, + { + "epoch": 2.6140751879699247, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 48459 + }, + { + "epoch": 2.6141353383458648, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 48460 + }, + { + "epoch": 2.6141954887218044, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 48461 + }, + { + "epoch": 2.6142556390977445, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 48462 + }, + { + "epoch": 2.614315789473684, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 48463 + }, + { + "epoch": 2.614375939849624, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 48464 + }, + { + "epoch": 2.614436090225564, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 48465 + }, + { + "epoch": 2.614496240601504, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 48466 + }, + { + "epoch": 2.6145563909774436, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 48467 + }, + { + "epoch": 2.6146165413533833, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.8919, + "step": 48468 + }, + { + "epoch": 2.6146766917293234, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 48469 + }, + { + "epoch": 2.614736842105263, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 48470 + }, + { + "epoch": 2.614796992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 48471 + }, + { + "epoch": 2.6148571428571428, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 48472 + }, + { + "epoch": 2.614917293233083, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 48473 + }, + { + "epoch": 2.6149774436090225, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 48474 + }, + { + "epoch": 2.6150375939849626, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9133, + "step": 48475 + }, + { + "epoch": 2.6150977443609023, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 48476 + }, + { + "epoch": 2.615157894736842, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 48477 + }, + { + "epoch": 2.615218045112782, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 48478 + }, + { + "epoch": 2.6152781954887216, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 48479 + }, + { + "epoch": 2.6153383458646617, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 48480 + }, + { + "epoch": 2.6153984962406014, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 48481 + }, + { + "epoch": 2.6154586466165415, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 48482 + }, + { + "epoch": 2.615518796992481, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 48483 + }, + { + "epoch": 2.6155789473684212, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 48484 + }, + { + "epoch": 2.615639097744361, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 48485 + }, + { + "epoch": 2.6156992481203005, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1368, + "step": 48486 + }, + { + "epoch": 2.6157593984962406, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 48487 + }, + { + "epoch": 2.6158195488721807, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 48488 + }, + { + "epoch": 2.6158796992481204, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 48489 + }, + { + "epoch": 2.61593984962406, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 48490 + }, + { + "epoch": 2.616, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 48491 + }, + { + "epoch": 2.6160601503759398, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 48492 + }, + { + "epoch": 2.61612030075188, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 48493 + }, + { + "epoch": 2.6161804511278195, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9406, + "step": 48494 + }, + { + "epoch": 2.616240601503759, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 48495 + }, + { + "epoch": 2.6163007518796992, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 48496 + }, + { + "epoch": 2.6163609022556393, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 48497 + }, + { + "epoch": 2.616421052631579, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 48498 + }, + { + "epoch": 2.6164812030075186, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 48499 + }, + { + "epoch": 2.6165413533834587, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 48500 + }, + { + "epoch": 2.6166015037593984, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 48501 + }, + { + "epoch": 2.6166616541353385, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.227, + "step": 48502 + }, + { + "epoch": 2.616721804511278, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 48503 + }, + { + "epoch": 2.616781954887218, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 48504 + }, + { + "epoch": 2.616842105263158, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 48505 + }, + { + "epoch": 2.616902255639098, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 48506 + }, + { + "epoch": 2.6169624060150376, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 48507 + }, + { + "epoch": 2.6170225563909773, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 48508 + }, + { + "epoch": 2.6170827067669173, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 48509 + }, + { + "epoch": 2.617142857142857, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 48510 + }, + { + "epoch": 2.617203007518797, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 48511 + }, + { + "epoch": 2.6172631578947367, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 48512 + }, + { + "epoch": 2.617323308270677, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 48513 + }, + { + "epoch": 2.6173834586466165, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 48514 + }, + { + "epoch": 2.6174436090225566, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 48515 + }, + { + "epoch": 2.617503759398496, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 48516 + }, + { + "epoch": 2.617563909774436, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 48517 + }, + { + "epoch": 2.617624060150376, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 48518 + }, + { + "epoch": 2.617684210526316, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 48519 + }, + { + "epoch": 2.6177443609022557, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 48520 + }, + { + "epoch": 2.6178045112781954, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 48521 + }, + { + "epoch": 2.6178646616541354, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 48522 + }, + { + "epoch": 2.617924812030075, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 48523 + }, + { + "epoch": 2.617984962406015, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 48524 + }, + { + "epoch": 2.618045112781955, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9217, + "step": 48525 + }, + { + "epoch": 2.6181052631578945, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 48526 + }, + { + "epoch": 2.6181654135338346, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 48527 + }, + { + "epoch": 2.6182255639097747, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 48528 + }, + { + "epoch": 2.6182857142857143, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 48529 + }, + { + "epoch": 2.618345864661654, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 48530 + }, + { + "epoch": 2.618406015037594, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 48531 + }, + { + "epoch": 2.6184661654135337, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1833, + "step": 48532 + }, + { + "epoch": 2.618526315789474, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 48533 + }, + { + "epoch": 2.6185864661654135, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 48534 + }, + { + "epoch": 2.618646616541353, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 48535 + }, + { + "epoch": 2.618706766917293, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 48536 + }, + { + "epoch": 2.6187669172932333, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 48537 + }, + { + "epoch": 2.618827067669173, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 48538 + }, + { + "epoch": 2.6188872180451126, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 48539 + }, + { + "epoch": 2.6189473684210527, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 48540 + }, + { + "epoch": 2.6190075187969923, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 48541 + }, + { + "epoch": 2.6190676691729324, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 48542 + }, + { + "epoch": 2.619127819548872, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 48543 + }, + { + "epoch": 2.619187969924812, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 48544 + }, + { + "epoch": 2.619248120300752, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 48545 + }, + { + "epoch": 2.619308270676692, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 48546 + }, + { + "epoch": 2.6193684210526316, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 48547 + }, + { + "epoch": 2.619428571428571, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 48548 + }, + { + "epoch": 2.6194887218045113, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 48549 + }, + { + "epoch": 2.6195488721804514, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 48550 + }, + { + "epoch": 2.619609022556391, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 48551 + }, + { + "epoch": 2.6196691729323307, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 48552 + }, + { + "epoch": 2.619729323308271, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 48553 + }, + { + "epoch": 2.6197894736842104, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 48554 + }, + { + "epoch": 2.6198496240601505, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 48555 + }, + { + "epoch": 2.61990977443609, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 48556 + }, + { + "epoch": 2.61996992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 48557 + }, + { + "epoch": 2.62003007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.8936, + "step": 48558 + }, + { + "epoch": 2.62009022556391, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 48559 + }, + { + "epoch": 2.6201503759398497, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 48560 + }, + { + "epoch": 2.6202105263157893, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 48561 + }, + { + "epoch": 2.6202706766917294, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 48562 + }, + { + "epoch": 2.620330827067669, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 48563 + }, + { + "epoch": 2.620390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 48564 + }, + { + "epoch": 2.620451127819549, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 48565 + }, + { + "epoch": 2.6205112781954885, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 48566 + }, + { + "epoch": 2.6205714285714286, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 48567 + }, + { + "epoch": 2.6206315789473686, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 48568 + }, + { + "epoch": 2.6206917293233083, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 48569 + }, + { + "epoch": 2.620751879699248, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 48570 + }, + { + "epoch": 2.620812030075188, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 48571 + }, + { + "epoch": 2.6208721804511277, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 48572 + }, + { + "epoch": 2.620932330827068, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 48573 + }, + { + "epoch": 2.6209924812030074, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 48574 + }, + { + "epoch": 2.6210526315789475, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 48575 + }, + { + "epoch": 2.621112781954887, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 48576 + }, + { + "epoch": 2.6211729323308273, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 48577 + }, + { + "epoch": 2.621233082706767, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 48578 + }, + { + "epoch": 2.6212932330827066, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 48579 + }, + { + "epoch": 2.6213533834586467, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 48580 + }, + { + "epoch": 2.6214135338345863, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1903, + "step": 48581 + }, + { + "epoch": 2.6214736842105264, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 48582 + }, + { + "epoch": 2.621533834586466, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 48583 + }, + { + "epoch": 2.621593984962406, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 48584 + }, + { + "epoch": 2.621654135338346, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 48585 + }, + { + "epoch": 2.621714285714286, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9338, + "step": 48586 + }, + { + "epoch": 2.6217744360902255, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 48587 + }, + { + "epoch": 2.621834586466165, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 48588 + }, + { + "epoch": 2.6218947368421053, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 48589 + }, + { + "epoch": 2.6219548872180454, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 48590 + }, + { + "epoch": 2.622015037593985, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 48591 + }, + { + "epoch": 2.6220751879699247, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 48592 + }, + { + "epoch": 2.6221353383458648, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 48593 + }, + { + "epoch": 2.6221954887218044, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9423, + "step": 48594 + }, + { + "epoch": 2.6222556390977445, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 48595 + }, + { + "epoch": 2.622315789473684, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 48596 + }, + { + "epoch": 2.622375939849624, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 48597 + }, + { + "epoch": 2.622436090225564, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9383, + "step": 48598 + }, + { + "epoch": 2.622496240601504, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 48599 + }, + { + "epoch": 2.6225563909774436, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 48600 + }, + { + "epoch": 2.6226165413533833, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 48601 + }, + { + "epoch": 2.6226766917293234, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 48602 + }, + { + "epoch": 2.622736842105263, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 48603 + }, + { + "epoch": 2.622796992481203, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 48604 + }, + { + "epoch": 2.6228571428571428, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 48605 + }, + { + "epoch": 2.622917293233083, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 48606 + }, + { + "epoch": 2.6229774436090225, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 48607 + }, + { + "epoch": 2.6230375939849626, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 48608 + }, + { + "epoch": 2.6230977443609023, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 48609 + }, + { + "epoch": 2.623157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 48610 + }, + { + "epoch": 2.623218045112782, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 48611 + }, + { + "epoch": 2.6232781954887217, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 48612 + }, + { + "epoch": 2.6233383458646617, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 48613 + }, + { + "epoch": 2.6233984962406014, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1686, + "step": 48614 + }, + { + "epoch": 2.6234586466165415, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 48615 + }, + { + "epoch": 2.623518796992481, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 48616 + }, + { + "epoch": 2.6235789473684212, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 48617 + }, + { + "epoch": 2.623639097744361, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 48618 + }, + { + "epoch": 2.6236992481203005, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 48619 + }, + { + "epoch": 2.6237593984962406, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 48620 + }, + { + "epoch": 2.6238195488721807, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 48621 + }, + { + "epoch": 2.6238796992481204, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 48622 + }, + { + "epoch": 2.62393984962406, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 48623 + }, + { + "epoch": 2.624, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 48624 + }, + { + "epoch": 2.6240601503759398, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 48625 + }, + { + "epoch": 2.62412030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 48626 + }, + { + "epoch": 2.6241804511278195, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 48627 + }, + { + "epoch": 2.624240601503759, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 48628 + }, + { + "epoch": 2.6243007518796992, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.122, + "step": 48629 + }, + { + "epoch": 2.6243609022556393, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 48630 + }, + { + "epoch": 2.624421052631579, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 48631 + }, + { + "epoch": 2.6244812030075186, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 48632 + }, + { + "epoch": 2.6245413533834587, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.924, + "step": 48633 + }, + { + "epoch": 2.6246015037593984, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 48634 + }, + { + "epoch": 2.6246616541353385, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.2425, + "step": 48635 + }, + { + "epoch": 2.624721804511278, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 48636 + }, + { + "epoch": 2.6247819548872178, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.8921, + "step": 48637 + }, + { + "epoch": 2.624842105263158, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 48638 + }, + { + "epoch": 2.624902255639098, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 48639 + }, + { + "epoch": 2.6249624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 48640 + }, + { + "epoch": 2.6250225563909773, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 48641 + }, + { + "epoch": 2.6250827067669174, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 48642 + }, + { + "epoch": 2.625142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 48643 + }, + { + "epoch": 2.625203007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 48644 + }, + { + "epoch": 2.6252631578947367, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1965, + "step": 48645 + }, + { + "epoch": 2.625323308270677, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 48646 + }, + { + "epoch": 2.6253834586466165, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9301, + "step": 48647 + }, + { + "epoch": 2.6254436090225566, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 48648 + }, + { + "epoch": 2.6255037593984962, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 48649 + }, + { + "epoch": 2.625563909774436, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 48650 + }, + { + "epoch": 2.625624060150376, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 48651 + }, + { + "epoch": 2.625684210526316, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 48652 + }, + { + "epoch": 2.6257443609022557, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.957, + "step": 48653 + }, + { + "epoch": 2.6258045112781954, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 48654 + }, + { + "epoch": 2.6258646616541355, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 48655 + }, + { + "epoch": 2.625924812030075, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9288, + "step": 48656 + }, + { + "epoch": 2.625984962406015, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 48657 + }, + { + "epoch": 2.626045112781955, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 48658 + }, + { + "epoch": 2.6261052631578945, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 48659 + }, + { + "epoch": 2.6261654135338346, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 48660 + }, + { + "epoch": 2.6262255639097747, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.2257, + "step": 48661 + }, + { + "epoch": 2.6262857142857143, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 48662 + }, + { + "epoch": 2.626345864661654, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9394, + "step": 48663 + }, + { + "epoch": 2.626406015037594, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 48664 + }, + { + "epoch": 2.6264661654135337, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 48665 + }, + { + "epoch": 2.626526315789474, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9489, + "step": 48666 + }, + { + "epoch": 2.6265864661654135, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 48667 + }, + { + "epoch": 2.626646616541353, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1278, + "step": 48668 + }, + { + "epoch": 2.626706766917293, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 48669 + }, + { + "epoch": 2.6267669172932333, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 48670 + }, + { + "epoch": 2.626827067669173, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 48671 + }, + { + "epoch": 2.6268872180451126, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.926, + "step": 48672 + }, + { + "epoch": 2.6269473684210527, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 48673 + }, + { + "epoch": 2.6270075187969923, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 48674 + }, + { + "epoch": 2.6270676691729324, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 48675 + }, + { + "epoch": 2.627127819548872, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 48676 + }, + { + "epoch": 2.627187969924812, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 48677 + }, + { + "epoch": 2.627248120300752, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 48678 + }, + { + "epoch": 2.627308270676692, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 48679 + }, + { + "epoch": 2.6273684210526316, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 48680 + }, + { + "epoch": 2.6274285714285712, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 48681 + }, + { + "epoch": 2.6274887218045113, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 48682 + }, + { + "epoch": 2.627548872180451, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 48683 + }, + { + "epoch": 2.627609022556391, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 48684 + }, + { + "epoch": 2.6276691729323307, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.173, + "step": 48685 + }, + { + "epoch": 2.627729323308271, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 48686 + }, + { + "epoch": 2.6277894736842105, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 48687 + }, + { + "epoch": 2.6278496240601505, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 48688 + }, + { + "epoch": 2.62790977443609, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 48689 + }, + { + "epoch": 2.62796992481203, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 48690 + }, + { + "epoch": 2.62803007518797, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 48691 + }, + { + "epoch": 2.62809022556391, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 48692 + }, + { + "epoch": 2.6281503759398497, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 48693 + }, + { + "epoch": 2.6282105263157893, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 48694 + }, + { + "epoch": 2.6282706766917294, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 48695 + }, + { + "epoch": 2.628330827067669, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 48696 + }, + { + "epoch": 2.628390977443609, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 48697 + }, + { + "epoch": 2.628451127819549, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 48698 + }, + { + "epoch": 2.6285112781954885, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 48699 + }, + { + "epoch": 2.6285714285714286, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 48700 + }, + { + "epoch": 2.6286315789473687, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 48701 + }, + { + "epoch": 2.6286917293233083, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 48702 + }, + { + "epoch": 2.628751879699248, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9084, + "step": 48703 + }, + { + "epoch": 2.628812030075188, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 48704 + }, + { + "epoch": 2.6288721804511277, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 48705 + }, + { + "epoch": 2.628932330827068, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9079, + "step": 48706 + }, + { + "epoch": 2.6289924812030074, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 48707 + }, + { + "epoch": 2.6290526315789475, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 48708 + }, + { + "epoch": 2.629112781954887, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 48709 + }, + { + "epoch": 2.6291729323308273, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 48710 + }, + { + "epoch": 2.629233082706767, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 48711 + }, + { + "epoch": 2.6292932330827066, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 48712 + }, + { + "epoch": 2.6293533834586467, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 48713 + }, + { + "epoch": 2.6294135338345863, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 48714 + }, + { + "epoch": 2.6294736842105264, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 48715 + }, + { + "epoch": 2.629533834586466, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 48716 + }, + { + "epoch": 2.629593984962406, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 48717 + }, + { + "epoch": 2.629654135338346, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 48718 + }, + { + "epoch": 2.629714285714286, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 48719 + }, + { + "epoch": 2.6297744360902255, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 48720 + }, + { + "epoch": 2.629834586466165, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 48721 + }, + { + "epoch": 2.6298947368421053, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9138, + "step": 48722 + }, + { + "epoch": 2.6299548872180454, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 48723 + }, + { + "epoch": 2.630015037593985, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 48724 + }, + { + "epoch": 2.6300751879699247, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 48725 + }, + { + "epoch": 2.6301353383458648, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 48726 + }, + { + "epoch": 2.6301954887218044, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 48727 + }, + { + "epoch": 2.6302556390977445, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 48728 + }, + { + "epoch": 2.630315789473684, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 48729 + }, + { + "epoch": 2.630375939849624, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 48730 + }, + { + "epoch": 2.630436090225564, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 48731 + }, + { + "epoch": 2.630496240601504, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 48732 + }, + { + "epoch": 2.6305563909774436, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 48733 + }, + { + "epoch": 2.6306165413533833, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 48734 + }, + { + "epoch": 2.6306766917293234, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 48735 + }, + { + "epoch": 2.630736842105263, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 48736 + }, + { + "epoch": 2.630796992481203, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 48737 + }, + { + "epoch": 2.630857142857143, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 48738 + }, + { + "epoch": 2.6309172932330824, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 48739 + }, + { + "epoch": 2.6309774436090225, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 48740 + }, + { + "epoch": 2.6310375939849626, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 48741 + }, + { + "epoch": 2.6310977443609023, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 48742 + }, + { + "epoch": 2.631157894736842, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 48743 + }, + { + "epoch": 2.631218045112782, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 48744 + }, + { + "epoch": 2.6312781954887217, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9373, + "step": 48745 + }, + { + "epoch": 2.6313383458646618, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 48746 + }, + { + "epoch": 2.6313984962406014, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 48747 + }, + { + "epoch": 2.6314586466165415, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 48748 + }, + { + "epoch": 2.631518796992481, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 48749 + }, + { + "epoch": 2.6315789473684212, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 48750 + }, + { + "epoch": 2.631639097744361, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 48751 + }, + { + "epoch": 2.6316992481203005, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 48752 + }, + { + "epoch": 2.6317593984962406, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 48753 + }, + { + "epoch": 2.6318195488721807, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 48754 + }, + { + "epoch": 2.6318796992481204, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 48755 + }, + { + "epoch": 2.63193984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 48756 + }, + { + "epoch": 2.632, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 48757 + }, + { + "epoch": 2.6320601503759398, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 48758 + }, + { + "epoch": 2.63212030075188, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 48759 + }, + { + "epoch": 2.6321804511278195, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 48760 + }, + { + "epoch": 2.632240601503759, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 48761 + }, + { + "epoch": 2.6323007518796993, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 48762 + }, + { + "epoch": 2.6323609022556393, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 48763 + }, + { + "epoch": 2.632421052631579, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 48764 + }, + { + "epoch": 2.6324812030075186, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 48765 + }, + { + "epoch": 2.6325413533834587, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 48766 + }, + { + "epoch": 2.6326015037593984, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 48767 + }, + { + "epoch": 2.6326616541353385, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 48768 + }, + { + "epoch": 2.632721804511278, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 48769 + }, + { + "epoch": 2.632781954887218, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 48770 + }, + { + "epoch": 2.632842105263158, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 48771 + }, + { + "epoch": 2.632902255639098, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 48772 + }, + { + "epoch": 2.6329624060150376, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 48773 + }, + { + "epoch": 2.6330225563909773, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 48774 + }, + { + "epoch": 2.6330827067669174, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1913, + "step": 48775 + }, + { + "epoch": 2.633142857142857, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9332, + "step": 48776 + }, + { + "epoch": 2.633203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 48777 + }, + { + "epoch": 2.6332631578947368, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 48778 + }, + { + "epoch": 2.633323308270677, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 48779 + }, + { + "epoch": 2.6333834586466165, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 48780 + }, + { + "epoch": 2.6334436090225566, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 48781 + }, + { + "epoch": 2.6335037593984962, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 48782 + }, + { + "epoch": 2.633563909774436, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 48783 + }, + { + "epoch": 2.633624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 48784 + }, + { + "epoch": 2.6336842105263156, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 48785 + }, + { + "epoch": 2.6337443609022557, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 48786 + }, + { + "epoch": 2.6338045112781954, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 48787 + }, + { + "epoch": 2.6338646616541355, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 48788 + }, + { + "epoch": 2.633924812030075, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9333, + "step": 48789 + }, + { + "epoch": 2.633984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 48790 + }, + { + "epoch": 2.634045112781955, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 48791 + }, + { + "epoch": 2.6341052631578945, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 48792 + }, + { + "epoch": 2.6341654135338346, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 48793 + }, + { + "epoch": 2.6342255639097747, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 48794 + }, + { + "epoch": 2.6342857142857143, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 48795 + }, + { + "epoch": 2.634345864661654, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 48796 + }, + { + "epoch": 2.634406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 48797 + }, + { + "epoch": 2.6344661654135337, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 48798 + }, + { + "epoch": 2.634526315789474, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 48799 + }, + { + "epoch": 2.6345864661654135, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 48800 + }, + { + "epoch": 2.634646616541353, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 48801 + }, + { + "epoch": 2.634706766917293, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 48802 + }, + { + "epoch": 2.6347669172932333, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 48803 + }, + { + "epoch": 2.634827067669173, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 48804 + }, + { + "epoch": 2.6348872180451126, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 48805 + }, + { + "epoch": 2.6349473684210527, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 48806 + }, + { + "epoch": 2.6350075187969924, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 48807 + }, + { + "epoch": 2.6350676691729324, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 48808 + }, + { + "epoch": 2.635127819548872, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 48809 + }, + { + "epoch": 2.635187969924812, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 48810 + }, + { + "epoch": 2.635248120300752, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 48811 + }, + { + "epoch": 2.635308270676692, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.137, + "step": 48812 + }, + { + "epoch": 2.6353684210526316, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 48813 + }, + { + "epoch": 2.6354285714285712, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 48814 + }, + { + "epoch": 2.6354887218045113, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 48815 + }, + { + "epoch": 2.635548872180451, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1524, + "step": 48816 + }, + { + "epoch": 2.635609022556391, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 48817 + }, + { + "epoch": 2.6356691729323307, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 48818 + }, + { + "epoch": 2.635729323308271, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 48819 + }, + { + "epoch": 2.6357894736842105, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 48820 + }, + { + "epoch": 2.6358496240601506, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 48821 + }, + { + "epoch": 2.63590977443609, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 48822 + }, + { + "epoch": 2.63596992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 48823 + }, + { + "epoch": 2.63603007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 48824 + }, + { + "epoch": 2.63609022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 48825 + }, + { + "epoch": 2.6361503759398497, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 48826 + }, + { + "epoch": 2.6362105263157893, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 48827 + }, + { + "epoch": 2.6362706766917294, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 48828 + }, + { + "epoch": 2.636330827067669, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 48829 + }, + { + "epoch": 2.636390977443609, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 48830 + }, + { + "epoch": 2.636451127819549, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 48831 + }, + { + "epoch": 2.6365112781954885, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9497, + "step": 48832 + }, + { + "epoch": 2.6365714285714286, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 48833 + }, + { + "epoch": 2.6366315789473687, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 48834 + }, + { + "epoch": 2.6366917293233083, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 48835 + }, + { + "epoch": 2.636751879699248, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 48836 + }, + { + "epoch": 2.636812030075188, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 48837 + }, + { + "epoch": 2.6368721804511277, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 48838 + }, + { + "epoch": 2.636932330827068, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 48839 + }, + { + "epoch": 2.6369924812030074, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.124, + "step": 48840 + }, + { + "epoch": 2.637052631578947, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 48841 + }, + { + "epoch": 2.637112781954887, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 48842 + }, + { + "epoch": 2.6371729323308273, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 48843 + }, + { + "epoch": 2.637233082706767, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 48844 + }, + { + "epoch": 2.6372932330827066, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 48845 + }, + { + "epoch": 2.6373533834586467, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 48846 + }, + { + "epoch": 2.6374135338345863, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 48847 + }, + { + "epoch": 2.6374736842105264, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.8894, + "step": 48848 + }, + { + "epoch": 2.637533834586466, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 48849 + }, + { + "epoch": 2.637593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 48850 + }, + { + "epoch": 2.637654135338346, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 48851 + }, + { + "epoch": 2.637714285714286, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 48852 + }, + { + "epoch": 2.6377744360902255, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 48853 + }, + { + "epoch": 2.637834586466165, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 48854 + }, + { + "epoch": 2.6378947368421053, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 48855 + }, + { + "epoch": 2.6379548872180454, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 48856 + }, + { + "epoch": 2.638015037593985, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 48857 + }, + { + "epoch": 2.6380751879699247, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 48858 + }, + { + "epoch": 2.6381353383458648, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 48859 + }, + { + "epoch": 2.6381954887218044, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 48860 + }, + { + "epoch": 2.6382556390977445, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 48861 + }, + { + "epoch": 2.638315789473684, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 48862 + }, + { + "epoch": 2.638375939849624, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 48863 + }, + { + "epoch": 2.638436090225564, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 48864 + }, + { + "epoch": 2.638496240601504, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 48865 + }, + { + "epoch": 2.6385563909774437, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 48866 + }, + { + "epoch": 2.6386165413533833, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 48867 + }, + { + "epoch": 2.6386766917293234, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 48868 + }, + { + "epoch": 2.638736842105263, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 48869 + }, + { + "epoch": 2.638796992481203, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 48870 + }, + { + "epoch": 2.638857142857143, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 48871 + }, + { + "epoch": 2.6389172932330824, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 48872 + }, + { + "epoch": 2.6389774436090225, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9014, + "step": 48873 + }, + { + "epoch": 2.6390375939849626, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 48874 + }, + { + "epoch": 2.6390977443609023, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 48875 + }, + { + "epoch": 2.639157894736842, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 48876 + }, + { + "epoch": 2.639218045112782, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 48877 + }, + { + "epoch": 2.6392781954887217, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 48878 + }, + { + "epoch": 2.6393383458646618, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 48879 + }, + { + "epoch": 2.6393984962406014, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 48880 + }, + { + "epoch": 2.6394586466165415, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 48881 + }, + { + "epoch": 2.639518796992481, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 48882 + }, + { + "epoch": 2.6395789473684212, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 48883 + }, + { + "epoch": 2.639639097744361, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 48884 + }, + { + "epoch": 2.6396992481203005, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 48885 + }, + { + "epoch": 2.6397593984962406, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 48886 + }, + { + "epoch": 2.6398195488721803, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 48887 + }, + { + "epoch": 2.6398796992481204, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 48888 + }, + { + "epoch": 2.63993984962406, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 48889 + }, + { + "epoch": 2.64, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 48890 + }, + { + "epoch": 2.6400601503759398, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 48891 + }, + { + "epoch": 2.64012030075188, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 48892 + }, + { + "epoch": 2.6401804511278195, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 48893 + }, + { + "epoch": 2.640240601503759, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 48894 + }, + { + "epoch": 2.6403007518796993, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 48895 + }, + { + "epoch": 2.6403609022556394, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 48896 + }, + { + "epoch": 2.640421052631579, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 48897 + }, + { + "epoch": 2.6404812030075187, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.1422, + "step": 48898 + }, + { + "epoch": 2.6405413533834587, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 48899 + }, + { + "epoch": 2.6406015037593984, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 48900 + }, + { + "epoch": 2.6406616541353385, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 48901 + }, + { + "epoch": 2.640721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 48902 + }, + { + "epoch": 2.640781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 48903 + }, + { + "epoch": 2.640842105263158, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 48904 + }, + { + "epoch": 2.640902255639098, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 48905 + }, + { + "epoch": 2.6409624060150376, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 48906 + }, + { + "epoch": 2.6410225563909773, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 48907 + }, + { + "epoch": 2.6410827067669174, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 48908 + }, + { + "epoch": 2.641142857142857, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 48909 + }, + { + "epoch": 2.641203007518797, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 48910 + }, + { + "epoch": 2.6412631578947368, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 48911 + }, + { + "epoch": 2.641323308270677, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 48912 + }, + { + "epoch": 2.6413834586466165, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 48913 + }, + { + "epoch": 2.6414436090225566, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 48914 + }, + { + "epoch": 2.6415037593984962, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 48915 + }, + { + "epoch": 2.641563909774436, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.8751, + "step": 48916 + }, + { + "epoch": 2.641624060150376, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9504, + "step": 48917 + }, + { + "epoch": 2.6416842105263156, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 48918 + }, + { + "epoch": 2.6417443609022557, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 48919 + }, + { + "epoch": 2.6418045112781954, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 48920 + }, + { + "epoch": 2.6418646616541355, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 48921 + }, + { + "epoch": 2.641924812030075, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 48922 + }, + { + "epoch": 2.641984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 48923 + }, + { + "epoch": 2.642045112781955, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 48924 + }, + { + "epoch": 2.6421052631578945, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 48925 + }, + { + "epoch": 2.6421654135338346, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 48926 + }, + { + "epoch": 2.6422255639097747, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 48927 + }, + { + "epoch": 2.6422857142857143, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 48928 + }, + { + "epoch": 2.642345864661654, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9127, + "step": 48929 + }, + { + "epoch": 2.642406015037594, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9287, + "step": 48930 + }, + { + "epoch": 2.6424661654135337, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 48931 + }, + { + "epoch": 2.642526315789474, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 48932 + }, + { + "epoch": 2.6425864661654135, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 48933 + }, + { + "epoch": 2.642646616541353, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 48934 + }, + { + "epoch": 2.6427067669172932, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 48935 + }, + { + "epoch": 2.6427669172932333, + "grad_norm": 2.15625, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 48936 + }, + { + "epoch": 2.642827067669173, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 48937 + }, + { + "epoch": 2.6428872180451126, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 48938 + }, + { + "epoch": 2.6429473684210527, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 48939 + }, + { + "epoch": 2.6430075187969924, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 48940 + }, + { + "epoch": 2.6430676691729325, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 48941 + }, + { + "epoch": 2.643127819548872, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1707, + "step": 48942 + }, + { + "epoch": 2.6431879699248118, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9351, + "step": 48943 + }, + { + "epoch": 2.643248120300752, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 48944 + }, + { + "epoch": 2.643308270676692, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 48945 + }, + { + "epoch": 2.6433684210526316, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 48946 + }, + { + "epoch": 2.6434285714285712, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 48947 + }, + { + "epoch": 2.6434887218045113, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 48948 + }, + { + "epoch": 2.643548872180451, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 48949 + }, + { + "epoch": 2.643609022556391, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 48950 + }, + { + "epoch": 2.6436691729323307, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 48951 + }, + { + "epoch": 2.643729323308271, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9447, + "step": 48952 + }, + { + "epoch": 2.6437894736842105, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 48953 + }, + { + "epoch": 2.6438496240601506, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 48954 + }, + { + "epoch": 2.64390977443609, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 48955 + }, + { + "epoch": 2.64396992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.903, + "step": 48956 + }, + { + "epoch": 2.64403007518797, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 48957 + }, + { + "epoch": 2.64409022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9026, + "step": 48958 + }, + { + "epoch": 2.6441503759398497, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 48959 + }, + { + "epoch": 2.6442105263157893, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 48960 + }, + { + "epoch": 2.6442706766917294, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 48961 + }, + { + "epoch": 2.644330827067669, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 48962 + }, + { + "epoch": 2.644390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 48963 + }, + { + "epoch": 2.644451127819549, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 48964 + }, + { + "epoch": 2.6445112781954885, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 48965 + }, + { + "epoch": 2.6445714285714286, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 48966 + }, + { + "epoch": 2.6446315789473687, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 48967 + }, + { + "epoch": 2.6446917293233083, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 48968 + }, + { + "epoch": 2.644751879699248, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 48969 + }, + { + "epoch": 2.644812030075188, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0852, + "step": 48970 + }, + { + "epoch": 2.6448721804511277, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 48971 + }, + { + "epoch": 2.644932330827068, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 48972 + }, + { + "epoch": 2.6449924812030075, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 48973 + }, + { + "epoch": 2.645052631578947, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 48974 + }, + { + "epoch": 2.645112781954887, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 48975 + }, + { + "epoch": 2.6451729323308273, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 48976 + }, + { + "epoch": 2.645233082706767, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 48977 + }, + { + "epoch": 2.6452932330827066, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 48978 + }, + { + "epoch": 2.6453533834586467, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 48979 + }, + { + "epoch": 2.6454135338345863, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 48980 + }, + { + "epoch": 2.6454736842105264, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1211, + "step": 48981 + }, + { + "epoch": 2.645533834586466, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 48982 + }, + { + "epoch": 2.645593984962406, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 48983 + }, + { + "epoch": 2.645654135338346, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 48984 + }, + { + "epoch": 2.645714285714286, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 48985 + }, + { + "epoch": 2.6457744360902256, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 48986 + }, + { + "epoch": 2.645834586466165, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 48987 + }, + { + "epoch": 2.6458947368421053, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 48988 + }, + { + "epoch": 2.645954887218045, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 48989 + }, + { + "epoch": 2.646015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 48990 + }, + { + "epoch": 2.6460751879699247, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 48991 + }, + { + "epoch": 2.646135338345865, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9281, + "step": 48992 + }, + { + "epoch": 2.6461954887218044, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 48993 + }, + { + "epoch": 2.6462556390977445, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 48994 + }, + { + "epoch": 2.646315789473684, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 48995 + }, + { + "epoch": 2.646375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 48996 + }, + { + "epoch": 2.646436090225564, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 48997 + }, + { + "epoch": 2.646496240601504, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 48998 + }, + { + "epoch": 2.6465563909774437, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 48999 + }, + { + "epoch": 2.6466165413533833, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 49000 + }, + { + "epoch": 2.6466766917293234, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 49001 + }, + { + "epoch": 2.646736842105263, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 49002 + }, + { + "epoch": 2.646796992481203, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 49003 + }, + { + "epoch": 2.646857142857143, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 49004 + }, + { + "epoch": 2.6469172932330824, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 49005 + }, + { + "epoch": 2.6469774436090225, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 49006 + }, + { + "epoch": 2.6470375939849626, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 49007 + }, + { + "epoch": 2.6470977443609023, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 49008 + }, + { + "epoch": 2.647157894736842, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 49009 + }, + { + "epoch": 2.647218045112782, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 49010 + }, + { + "epoch": 2.6472781954887217, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 49011 + }, + { + "epoch": 2.6473383458646618, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 49012 + }, + { + "epoch": 2.6473984962406014, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 49013 + }, + { + "epoch": 2.6474586466165415, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 49014 + }, + { + "epoch": 2.647518796992481, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 49015 + }, + { + "epoch": 2.6475789473684213, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 49016 + }, + { + "epoch": 2.647639097744361, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2282, + "step": 49017 + }, + { + "epoch": 2.6476992481203006, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 49018 + }, + { + "epoch": 2.6477593984962406, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 49019 + }, + { + "epoch": 2.6478195488721803, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 49020 + }, + { + "epoch": 2.6478796992481204, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 49021 + }, + { + "epoch": 2.64793984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 49022 + }, + { + "epoch": 2.648, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 49023 + }, + { + "epoch": 2.64806015037594, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 49024 + }, + { + "epoch": 2.64812030075188, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1265, + "step": 49025 + }, + { + "epoch": 2.6481804511278195, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 49026 + }, + { + "epoch": 2.648240601503759, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 49027 + }, + { + "epoch": 2.6483007518796993, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 49028 + }, + { + "epoch": 2.6483609022556394, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9129, + "step": 49029 + }, + { + "epoch": 2.648421052631579, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 49030 + }, + { + "epoch": 2.6484812030075187, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 49031 + }, + { + "epoch": 2.6485413533834588, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 49032 + }, + { + "epoch": 2.6486015037593984, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 49033 + }, + { + "epoch": 2.6486616541353385, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 49034 + }, + { + "epoch": 2.648721804511278, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 49035 + }, + { + "epoch": 2.648781954887218, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 49036 + }, + { + "epoch": 2.648842105263158, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 49037 + }, + { + "epoch": 2.648902255639098, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 49038 + }, + { + "epoch": 2.6489624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9126, + "step": 49039 + }, + { + "epoch": 2.6490225563909773, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 49040 + }, + { + "epoch": 2.6490827067669174, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 49041 + }, + { + "epoch": 2.649142857142857, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9388, + "step": 49042 + }, + { + "epoch": 2.649203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 49043 + }, + { + "epoch": 2.6492631578947368, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 49044 + }, + { + "epoch": 2.649323308270677, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 49045 + }, + { + "epoch": 2.6493834586466165, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 49046 + }, + { + "epoch": 2.6494436090225566, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 49047 + }, + { + "epoch": 2.6495037593984962, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 49048 + }, + { + "epoch": 2.649563909774436, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 49049 + }, + { + "epoch": 2.649624060150376, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 49050 + }, + { + "epoch": 2.6496842105263156, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 49051 + }, + { + "epoch": 2.6497443609022557, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 49052 + }, + { + "epoch": 2.6498045112781954, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 49053 + }, + { + "epoch": 2.6498646616541355, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 49054 + }, + { + "epoch": 2.649924812030075, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 49055 + }, + { + "epoch": 2.649984962406015, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 49056 + }, + { + "epoch": 2.650045112781955, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9096, + "step": 49057 + }, + { + "epoch": 2.6501052631578945, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 49058 + }, + { + "epoch": 2.6501654135338346, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 49059 + }, + { + "epoch": 2.6502255639097747, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 49060 + }, + { + "epoch": 2.6502857142857144, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 49061 + }, + { + "epoch": 2.650345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 49062 + }, + { + "epoch": 2.650406015037594, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 49063 + }, + { + "epoch": 2.6504661654135337, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 49064 + }, + { + "epoch": 2.650526315789474, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 49065 + }, + { + "epoch": 2.6505864661654135, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 49066 + }, + { + "epoch": 2.650646616541353, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 49067 + }, + { + "epoch": 2.6507067669172932, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 49068 + }, + { + "epoch": 2.6507669172932333, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 49069 + }, + { + "epoch": 2.650827067669173, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 49070 + }, + { + "epoch": 2.6508872180451126, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 49071 + }, + { + "epoch": 2.6509473684210527, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 49072 + }, + { + "epoch": 2.6510075187969924, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 49073 + }, + { + "epoch": 2.6510676691729325, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 49074 + }, + { + "epoch": 2.651127819548872, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 49075 + }, + { + "epoch": 2.6511879699248118, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 49076 + }, + { + "epoch": 2.651248120300752, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 49077 + }, + { + "epoch": 2.651308270676692, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 49078 + }, + { + "epoch": 2.6513684210526316, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 49079 + }, + { + "epoch": 2.6514285714285712, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 49080 + }, + { + "epoch": 2.6514887218045113, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 49081 + }, + { + "epoch": 2.651548872180451, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 49082 + }, + { + "epoch": 2.651609022556391, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 49083 + }, + { + "epoch": 2.6516691729323307, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 49084 + }, + { + "epoch": 2.651729323308271, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 49085 + }, + { + "epoch": 2.6517894736842105, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 49086 + }, + { + "epoch": 2.6518496240601506, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 49087 + }, + { + "epoch": 2.65190977443609, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 49088 + }, + { + "epoch": 2.65196992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 49089 + }, + { + "epoch": 2.65203007518797, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 49090 + }, + { + "epoch": 2.6520902255639096, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 49091 + }, + { + "epoch": 2.6521503759398497, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 49092 + }, + { + "epoch": 2.6522105263157894, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9276, + "step": 49093 + }, + { + "epoch": 2.6522706766917294, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 49094 + }, + { + "epoch": 2.652330827067669, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 49095 + }, + { + "epoch": 2.652390977443609, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 49096 + }, + { + "epoch": 2.652451127819549, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 49097 + }, + { + "epoch": 2.6525112781954885, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 49098 + }, + { + "epoch": 2.6525714285714286, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 49099 + }, + { + "epoch": 2.6526315789473687, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 49100 + }, + { + "epoch": 2.6526917293233083, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 49101 + }, + { + "epoch": 2.652751879699248, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 49102 + }, + { + "epoch": 2.652812030075188, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 49103 + }, + { + "epoch": 2.6528721804511277, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 49104 + }, + { + "epoch": 2.652932330827068, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 49105 + }, + { + "epoch": 2.6529924812030075, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 49106 + }, + { + "epoch": 2.653052631578947, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 49107 + }, + { + "epoch": 2.653112781954887, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 49108 + }, + { + "epoch": 2.6531729323308273, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 49109 + }, + { + "epoch": 2.653233082706767, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 49110 + }, + { + "epoch": 2.6532932330827066, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 49111 + }, + { + "epoch": 2.6533533834586467, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 49112 + }, + { + "epoch": 2.6534135338345863, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 49113 + }, + { + "epoch": 2.6534736842105264, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 49114 + }, + { + "epoch": 2.653533834586466, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 49115 + }, + { + "epoch": 2.653593984962406, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 49116 + }, + { + "epoch": 2.653654135338346, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 49117 + }, + { + "epoch": 2.653714285714286, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 49118 + }, + { + "epoch": 2.6537744360902256, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 49119 + }, + { + "epoch": 2.653834586466165, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 49120 + }, + { + "epoch": 2.6538947368421053, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.3345, + "step": 49121 + }, + { + "epoch": 2.653954887218045, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 49122 + }, + { + "epoch": 2.654015037593985, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 49123 + }, + { + "epoch": 2.6540751879699247, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 49124 + }, + { + "epoch": 2.654135338345865, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 49125 + }, + { + "epoch": 2.6541954887218044, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 49126 + }, + { + "epoch": 2.6542556390977445, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 49127 + }, + { + "epoch": 2.654315789473684, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 49128 + }, + { + "epoch": 2.654375939849624, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 49129 + }, + { + "epoch": 2.654436090225564, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 49130 + }, + { + "epoch": 2.654496240601504, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 49131 + }, + { + "epoch": 2.6545563909774437, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 49132 + }, + { + "epoch": 2.6546165413533833, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 49133 + }, + { + "epoch": 2.6546766917293234, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 49134 + }, + { + "epoch": 2.654736842105263, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 49135 + }, + { + "epoch": 2.654796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 49136 + }, + { + "epoch": 2.654857142857143, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 49137 + }, + { + "epoch": 2.6549172932330825, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 49138 + }, + { + "epoch": 2.6549774436090225, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 49139 + }, + { + "epoch": 2.6550375939849626, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 49140 + }, + { + "epoch": 2.6550977443609023, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 49141 + }, + { + "epoch": 2.655157894736842, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 49142 + }, + { + "epoch": 2.655218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 49143 + }, + { + "epoch": 2.6552781954887217, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 49144 + }, + { + "epoch": 2.6553383458646618, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 49145 + }, + { + "epoch": 2.6553984962406014, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 49146 + }, + { + "epoch": 2.6554586466165415, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 49147 + }, + { + "epoch": 2.655518796992481, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 49148 + }, + { + "epoch": 2.6555789473684213, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 49149 + }, + { + "epoch": 2.655639097744361, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 49150 + }, + { + "epoch": 2.6556992481203006, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 49151 + }, + { + "epoch": 2.6557593984962407, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 49152 + }, + { + "epoch": 2.6558195488721803, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 49153 + }, + { + "epoch": 2.6558796992481204, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 49154 + }, + { + "epoch": 2.65593984962406, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 49155 + }, + { + "epoch": 2.656, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 49156 + }, + { + "epoch": 2.65606015037594, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 49157 + }, + { + "epoch": 2.65612030075188, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 49158 + }, + { + "epoch": 2.6561804511278195, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 49159 + }, + { + "epoch": 2.656240601503759, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 49160 + }, + { + "epoch": 2.6563007518796993, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 49161 + }, + { + "epoch": 2.6563609022556394, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 49162 + }, + { + "epoch": 2.656421052631579, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.885, + "step": 49163 + }, + { + "epoch": 2.6564812030075187, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.946, + "step": 49164 + }, + { + "epoch": 2.6565413533834588, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 49165 + }, + { + "epoch": 2.6566015037593984, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 49166 + }, + { + "epoch": 2.6566616541353385, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 49167 + }, + { + "epoch": 2.656721804511278, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9351, + "step": 49168 + }, + { + "epoch": 2.656781954887218, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 49169 + }, + { + "epoch": 2.656842105263158, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 49170 + }, + { + "epoch": 2.656902255639098, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 49171 + }, + { + "epoch": 2.6569624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 49172 + }, + { + "epoch": 2.6570225563909773, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 49173 + }, + { + "epoch": 2.6570827067669174, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 49174 + }, + { + "epoch": 2.657142857142857, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 49175 + }, + { + "epoch": 2.657203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 49176 + }, + { + "epoch": 2.6572631578947368, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 49177 + }, + { + "epoch": 2.6573233082706764, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 49178 + }, + { + "epoch": 2.6573834586466165, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 49179 + }, + { + "epoch": 2.6574436090225566, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 49180 + }, + { + "epoch": 2.6575037593984963, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 49181 + }, + { + "epoch": 2.657563909774436, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 49182 + }, + { + "epoch": 2.657624060150376, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 49183 + }, + { + "epoch": 2.6576842105263156, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 49184 + }, + { + "epoch": 2.6577443609022557, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 49185 + }, + { + "epoch": 2.6578045112781954, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 49186 + }, + { + "epoch": 2.6578646616541355, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 49187 + }, + { + "epoch": 2.657924812030075, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 49188 + }, + { + "epoch": 2.6579849624060152, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.8955, + "step": 49189 + }, + { + "epoch": 2.658045112781955, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 49190 + }, + { + "epoch": 2.6581052631578945, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 49191 + }, + { + "epoch": 2.6581654135338346, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 49192 + }, + { + "epoch": 2.6582255639097747, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 49193 + }, + { + "epoch": 2.6582857142857144, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 49194 + }, + { + "epoch": 2.658345864661654, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 49195 + }, + { + "epoch": 2.658406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 49196 + }, + { + "epoch": 2.6584661654135338, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 49197 + }, + { + "epoch": 2.658526315789474, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 49198 + }, + { + "epoch": 2.6585864661654135, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 49199 + }, + { + "epoch": 2.658646616541353, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 49200 + }, + { + "epoch": 2.6587067669172932, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 49201 + }, + { + "epoch": 2.6587669172932333, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 49202 + }, + { + "epoch": 2.658827067669173, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 49203 + }, + { + "epoch": 2.6588872180451126, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 49204 + }, + { + "epoch": 2.6589473684210527, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 49205 + }, + { + "epoch": 2.6590075187969924, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 49206 + }, + { + "epoch": 2.6590676691729325, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 49207 + }, + { + "epoch": 2.659127819548872, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 49208 + }, + { + "epoch": 2.6591879699248118, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 49209 + }, + { + "epoch": 2.659248120300752, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 49210 + }, + { + "epoch": 2.659308270676692, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 49211 + }, + { + "epoch": 2.6593684210526316, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 49212 + }, + { + "epoch": 2.6594285714285713, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 49213 + }, + { + "epoch": 2.6594887218045113, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 49214 + }, + { + "epoch": 2.659548872180451, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 49215 + }, + { + "epoch": 2.659609022556391, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 49216 + }, + { + "epoch": 2.6596691729323307, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 49217 + }, + { + "epoch": 2.659729323308271, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 49218 + }, + { + "epoch": 2.6597894736842105, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 49219 + }, + { + "epoch": 2.6598496240601506, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 49220 + }, + { + "epoch": 2.6599097744360902, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 49221 + }, + { + "epoch": 2.65996992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 49222 + }, + { + "epoch": 2.66003007518797, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 49223 + }, + { + "epoch": 2.6600902255639096, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 49224 + }, + { + "epoch": 2.6601503759398497, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 49225 + }, + { + "epoch": 2.6602105263157894, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 49226 + }, + { + "epoch": 2.6602706766917295, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9388, + "step": 49227 + }, + { + "epoch": 2.660330827067669, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0733, + "step": 49228 + }, + { + "epoch": 2.660390977443609, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 49229 + }, + { + "epoch": 2.660451127819549, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 49230 + }, + { + "epoch": 2.6605112781954885, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 49231 + }, + { + "epoch": 2.6605714285714286, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 49232 + }, + { + "epoch": 2.6606315789473687, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 49233 + }, + { + "epoch": 2.6606917293233083, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1247, + "step": 49234 + }, + { + "epoch": 2.660751879699248, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9323, + "step": 49235 + }, + { + "epoch": 2.660812030075188, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 49236 + }, + { + "epoch": 2.6608721804511277, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 49237 + }, + { + "epoch": 2.660932330827068, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 49238 + }, + { + "epoch": 2.6609924812030075, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 49239 + }, + { + "epoch": 2.661052631578947, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 49240 + }, + { + "epoch": 2.661112781954887, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 49241 + }, + { + "epoch": 2.6611729323308273, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 49242 + }, + { + "epoch": 2.661233082706767, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 49243 + }, + { + "epoch": 2.6612932330827066, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 49244 + }, + { + "epoch": 2.6613533834586467, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 49245 + }, + { + "epoch": 2.6614135338345863, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 49246 + }, + { + "epoch": 2.6614736842105264, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 49247 + }, + { + "epoch": 2.661533834586466, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 49248 + }, + { + "epoch": 2.661593984962406, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 49249 + }, + { + "epoch": 2.661654135338346, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 49250 + }, + { + "epoch": 2.661714285714286, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1487, + "step": 49251 + }, + { + "epoch": 2.6617744360902256, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 49252 + }, + { + "epoch": 2.661834586466165, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.8676, + "step": 49253 + }, + { + "epoch": 2.6618947368421053, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 49254 + }, + { + "epoch": 2.661954887218045, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 49255 + }, + { + "epoch": 2.662015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 49256 + }, + { + "epoch": 2.6620751879699247, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9337, + "step": 49257 + }, + { + "epoch": 2.662135338345865, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9313, + "step": 49258 + }, + { + "epoch": 2.6621954887218044, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 49259 + }, + { + "epoch": 2.6622556390977445, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 49260 + }, + { + "epoch": 2.662315789473684, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 49261 + }, + { + "epoch": 2.662375939849624, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 49262 + }, + { + "epoch": 2.662436090225564, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 49263 + }, + { + "epoch": 2.662496240601504, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 49264 + }, + { + "epoch": 2.6625563909774437, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 49265 + }, + { + "epoch": 2.6626165413533833, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 49266 + }, + { + "epoch": 2.6626766917293234, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 49267 + }, + { + "epoch": 2.662736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 49268 + }, + { + "epoch": 2.662796992481203, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0989, + "step": 49269 + }, + { + "epoch": 2.662857142857143, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 49270 + }, + { + "epoch": 2.6629172932330825, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 49271 + }, + { + "epoch": 2.6629774436090226, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 49272 + }, + { + "epoch": 2.6630375939849626, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9396, + "step": 49273 + }, + { + "epoch": 2.6630977443609023, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 49274 + }, + { + "epoch": 2.663157894736842, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 49275 + }, + { + "epoch": 2.663218045112782, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 49276 + }, + { + "epoch": 2.6632781954887217, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 49277 + }, + { + "epoch": 2.663338345864662, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9312, + "step": 49278 + }, + { + "epoch": 2.6633984962406014, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 49279 + }, + { + "epoch": 2.663458646616541, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 49280 + }, + { + "epoch": 2.663518796992481, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 49281 + }, + { + "epoch": 2.6635789473684213, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 49282 + }, + { + "epoch": 2.663639097744361, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 49283 + }, + { + "epoch": 2.6636992481203006, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 49284 + }, + { + "epoch": 2.6637593984962407, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 49285 + }, + { + "epoch": 2.6638195488721803, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 49286 + }, + { + "epoch": 2.6638796992481204, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 49287 + }, + { + "epoch": 2.66393984962406, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 49288 + }, + { + "epoch": 2.664, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9238, + "step": 49289 + }, + { + "epoch": 2.66406015037594, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 49290 + }, + { + "epoch": 2.66412030075188, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 49291 + }, + { + "epoch": 2.6641804511278195, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 49292 + }, + { + "epoch": 2.664240601503759, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 49293 + }, + { + "epoch": 2.6643007518796993, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9282, + "step": 49294 + }, + { + "epoch": 2.6643609022556394, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 49295 + }, + { + "epoch": 2.664421052631579, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.938, + "step": 49296 + }, + { + "epoch": 2.6644812030075187, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 49297 + }, + { + "epoch": 2.6645413533834588, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 49298 + }, + { + "epoch": 2.6646015037593984, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 49299 + }, + { + "epoch": 2.6646616541353385, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 49300 + }, + { + "epoch": 2.664721804511278, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 49301 + }, + { + "epoch": 2.664781954887218, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 49302 + }, + { + "epoch": 2.664842105263158, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 49303 + }, + { + "epoch": 2.664902255639098, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9338, + "step": 49304 + }, + { + "epoch": 2.6649624060150376, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 49305 + }, + { + "epoch": 2.6650225563909773, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 49306 + }, + { + "epoch": 2.6650827067669174, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 49307 + }, + { + "epoch": 2.665142857142857, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 49308 + }, + { + "epoch": 2.665203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 49309 + }, + { + "epoch": 2.665263157894737, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 49310 + }, + { + "epoch": 2.6653233082706764, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 49311 + }, + { + "epoch": 2.6653834586466165, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9262, + "step": 49312 + }, + { + "epoch": 2.6654436090225566, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9667, + "step": 49313 + }, + { + "epoch": 2.6655037593984963, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 49314 + }, + { + "epoch": 2.665563909774436, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 49315 + }, + { + "epoch": 2.665624060150376, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 49316 + }, + { + "epoch": 2.6656842105263157, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 49317 + }, + { + "epoch": 2.6657443609022557, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 49318 + }, + { + "epoch": 2.6658045112781954, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 49319 + }, + { + "epoch": 2.6658646616541355, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.8979, + "step": 49320 + }, + { + "epoch": 2.665924812030075, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 49321 + }, + { + "epoch": 2.6659849624060152, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 49322 + }, + { + "epoch": 2.666045112781955, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9586, + "step": 49323 + }, + { + "epoch": 2.6661052631578945, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 49324 + }, + { + "epoch": 2.6661654135338346, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 49325 + }, + { + "epoch": 2.6662255639097743, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 49326 + }, + { + "epoch": 2.6662857142857144, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1245, + "step": 49327 + }, + { + "epoch": 2.666345864661654, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 49328 + }, + { + "epoch": 2.666406015037594, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 49329 + }, + { + "epoch": 2.6664661654135338, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 49330 + }, + { + "epoch": 2.666526315789474, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 49331 + }, + { + "epoch": 2.6665864661654135, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1758, + "step": 49332 + }, + { + "epoch": 2.666646616541353, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1489, + "step": 49333 + }, + { + "epoch": 2.6667067669172932, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 49334 + }, + { + "epoch": 2.6667669172932333, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 49335 + }, + { + "epoch": 2.666827067669173, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.143, + "step": 49336 + }, + { + "epoch": 2.6668872180451126, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 49337 + }, + { + "epoch": 2.6669473684210527, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 49338 + }, + { + "epoch": 2.6670075187969924, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 49339 + }, + { + "epoch": 2.6670676691729325, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 49340 + }, + { + "epoch": 2.667127819548872, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 49341 + }, + { + "epoch": 2.6671879699248118, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 49342 + }, + { + "epoch": 2.667248120300752, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1154, + "step": 49343 + }, + { + "epoch": 2.667308270676692, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 49344 + }, + { + "epoch": 2.6673684210526316, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 49345 + }, + { + "epoch": 2.6674285714285713, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 49346 + }, + { + "epoch": 2.6674887218045114, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 49347 + }, + { + "epoch": 2.667548872180451, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 49348 + }, + { + "epoch": 2.667609022556391, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 49349 + }, + { + "epoch": 2.6676691729323307, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9442, + "step": 49350 + }, + { + "epoch": 2.667729323308271, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.2275, + "step": 49351 + }, + { + "epoch": 2.6677894736842105, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 49352 + }, + { + "epoch": 2.6678496240601506, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 49353 + }, + { + "epoch": 2.6679097744360902, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 49354 + }, + { + "epoch": 2.66796992481203, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 49355 + }, + { + "epoch": 2.66803007518797, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 49356 + }, + { + "epoch": 2.6680902255639096, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 49357 + }, + { + "epoch": 2.6681503759398497, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1939, + "step": 49358 + }, + { + "epoch": 2.6682105263157894, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0984, + "step": 49359 + }, + { + "epoch": 2.6682706766917295, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 49360 + }, + { + "epoch": 2.668330827067669, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.3415, + "step": 49361 + }, + { + "epoch": 2.668390977443609, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 49362 + }, + { + "epoch": 2.668451127819549, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 49363 + }, + { + "epoch": 2.6685112781954885, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 49364 + }, + { + "epoch": 2.6685714285714286, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 49365 + }, + { + "epoch": 2.6686315789473687, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 49366 + }, + { + "epoch": 2.6686917293233083, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 49367 + }, + { + "epoch": 2.668751879699248, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 49368 + }, + { + "epoch": 2.668812030075188, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 49369 + }, + { + "epoch": 2.6688721804511277, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 49370 + }, + { + "epoch": 2.668932330827068, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 49371 + }, + { + "epoch": 2.6689924812030075, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 49372 + }, + { + "epoch": 2.669052631578947, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 49373 + }, + { + "epoch": 2.669112781954887, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 49374 + }, + { + "epoch": 2.6691729323308273, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 49375 + }, + { + "epoch": 2.669233082706767, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 49376 + }, + { + "epoch": 2.6692932330827066, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9324, + "step": 49377 + }, + { + "epoch": 2.6693533834586467, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 49378 + }, + { + "epoch": 2.6694135338345863, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 49379 + }, + { + "epoch": 2.6694736842105264, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 49380 + }, + { + "epoch": 2.669533834586466, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 49381 + }, + { + "epoch": 2.6695939849624057, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 49382 + }, + { + "epoch": 2.669654135338346, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9352, + "step": 49383 + }, + { + "epoch": 2.669714285714286, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1404, + "step": 49384 + }, + { + "epoch": 2.6697744360902256, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 49385 + }, + { + "epoch": 2.6698345864661652, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 49386 + }, + { + "epoch": 2.6698947368421053, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 49387 + }, + { + "epoch": 2.669954887218045, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 49388 + }, + { + "epoch": 2.670015037593985, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9347, + "step": 49389 + }, + { + "epoch": 2.6700751879699247, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 49390 + }, + { + "epoch": 2.670135338345865, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 49391 + }, + { + "epoch": 2.6701954887218045, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 49392 + }, + { + "epoch": 2.6702556390977445, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 49393 + }, + { + "epoch": 2.670315789473684, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1913, + "step": 49394 + }, + { + "epoch": 2.670375939849624, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 49395 + }, + { + "epoch": 2.670436090225564, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 49396 + }, + { + "epoch": 2.670496240601504, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 49397 + }, + { + "epoch": 2.6705563909774437, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 49398 + }, + { + "epoch": 2.6706165413533833, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 49399 + }, + { + "epoch": 2.6706766917293234, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 49400 + }, + { + "epoch": 2.670736842105263, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 49401 + }, + { + "epoch": 2.670796992481203, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 49402 + }, + { + "epoch": 2.670857142857143, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 49403 + }, + { + "epoch": 2.6709172932330825, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 49404 + }, + { + "epoch": 2.6709774436090226, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 49405 + }, + { + "epoch": 2.6710375939849627, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 49406 + }, + { + "epoch": 2.6710977443609023, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 49407 + }, + { + "epoch": 2.671157894736842, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 49408 + }, + { + "epoch": 2.671218045112782, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 49409 + }, + { + "epoch": 2.6712781954887217, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 49410 + }, + { + "epoch": 2.671338345864662, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 49411 + }, + { + "epoch": 2.6713984962406014, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.8684, + "step": 49412 + }, + { + "epoch": 2.671458646616541, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 49413 + }, + { + "epoch": 2.671518796992481, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 49414 + }, + { + "epoch": 2.6715789473684213, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 49415 + }, + { + "epoch": 2.671639097744361, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 49416 + }, + { + "epoch": 2.6716992481203006, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 49417 + }, + { + "epoch": 2.6717593984962407, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 49418 + }, + { + "epoch": 2.6718195488721803, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 49419 + }, + { + "epoch": 2.6718796992481204, + "grad_norm": 1.9765625, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 49420 + }, + { + "epoch": 2.67193984962406, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 49421 + }, + { + "epoch": 2.672, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 49422 + }, + { + "epoch": 2.67206015037594, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 49423 + }, + { + "epoch": 2.67212030075188, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 49424 + }, + { + "epoch": 2.6721804511278195, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 49425 + }, + { + "epoch": 2.672240601503759, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 49426 + }, + { + "epoch": 2.6723007518796993, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 49427 + }, + { + "epoch": 2.672360902255639, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 49428 + }, + { + "epoch": 2.672421052631579, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 49429 + }, + { + "epoch": 2.6724812030075187, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 49430 + }, + { + "epoch": 2.6725413533834588, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 49431 + }, + { + "epoch": 2.6726015037593984, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 49432 + }, + { + "epoch": 2.6726616541353385, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 49433 + }, + { + "epoch": 2.672721804511278, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 49434 + }, + { + "epoch": 2.672781954887218, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 49435 + }, + { + "epoch": 2.672842105263158, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9136, + "step": 49436 + }, + { + "epoch": 2.672902255639098, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 49437 + }, + { + "epoch": 2.6729624060150377, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 49438 + }, + { + "epoch": 2.6730225563909773, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 49439 + }, + { + "epoch": 2.6730827067669174, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 49440 + }, + { + "epoch": 2.673142857142857, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 49441 + }, + { + "epoch": 2.673203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 49442 + }, + { + "epoch": 2.673263157894737, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 49443 + }, + { + "epoch": 2.6733233082706764, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 49444 + }, + { + "epoch": 2.6733834586466165, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 49445 + }, + { + "epoch": 2.6734436090225566, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 49446 + }, + { + "epoch": 2.6735037593984963, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 49447 + }, + { + "epoch": 2.673563909774436, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 49448 + }, + { + "epoch": 2.673624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 49449 + }, + { + "epoch": 2.6736842105263157, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 49450 + }, + { + "epoch": 2.6737443609022558, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 49451 + }, + { + "epoch": 2.6738045112781954, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9388, + "step": 49452 + }, + { + "epoch": 2.6738646616541355, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 49453 + }, + { + "epoch": 2.673924812030075, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9388, + "step": 49454 + }, + { + "epoch": 2.6739849624060152, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 49455 + }, + { + "epoch": 2.674045112781955, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 49456 + }, + { + "epoch": 2.6741052631578945, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 49457 + }, + { + "epoch": 2.6741654135338346, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 49458 + }, + { + "epoch": 2.6742255639097743, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 49459 + }, + { + "epoch": 2.6742857142857144, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 49460 + }, + { + "epoch": 2.674345864661654, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 49461 + }, + { + "epoch": 2.674406015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 49462 + }, + { + "epoch": 2.6744661654135338, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 49463 + }, + { + "epoch": 2.674526315789474, + "grad_norm": 1.4140625, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 49464 + }, + { + "epoch": 2.6745864661654135, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 49465 + }, + { + "epoch": 2.674646616541353, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 49466 + }, + { + "epoch": 2.6747067669172933, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 49467 + }, + { + "epoch": 2.6747669172932333, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 49468 + }, + { + "epoch": 2.674827067669173, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 49469 + }, + { + "epoch": 2.6748872180451126, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 49470 + }, + { + "epoch": 2.6749473684210527, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 49471 + }, + { + "epoch": 2.6750075187969924, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 49472 + }, + { + "epoch": 2.6750676691729325, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 49473 + }, + { + "epoch": 2.675127819548872, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 49474 + }, + { + "epoch": 2.675187969924812, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 49475 + }, + { + "epoch": 2.675248120300752, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 49476 + }, + { + "epoch": 2.675308270676692, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9489, + "step": 49477 + }, + { + "epoch": 2.6753684210526316, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 49478 + }, + { + "epoch": 2.6754285714285713, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 49479 + }, + { + "epoch": 2.6754887218045114, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 49480 + }, + { + "epoch": 2.675548872180451, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 49481 + }, + { + "epoch": 2.675609022556391, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 49482 + }, + { + "epoch": 2.6756691729323308, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 49483 + }, + { + "epoch": 2.6757293233082704, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 49484 + }, + { + "epoch": 2.6757894736842105, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 49485 + }, + { + "epoch": 2.6758496240601506, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1358, + "step": 49486 + }, + { + "epoch": 2.6759097744360902, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 49487 + }, + { + "epoch": 2.67596992481203, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 49488 + }, + { + "epoch": 2.67603007518797, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9016, + "step": 49489 + }, + { + "epoch": 2.6760902255639096, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 49490 + }, + { + "epoch": 2.6761503759398497, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 49491 + }, + { + "epoch": 2.6762105263157894, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.118, + "step": 49492 + }, + { + "epoch": 2.6762706766917295, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 49493 + }, + { + "epoch": 2.676330827067669, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 49494 + }, + { + "epoch": 2.676390977443609, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 49495 + }, + { + "epoch": 2.676451127819549, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 49496 + }, + { + "epoch": 2.6765112781954885, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 49497 + }, + { + "epoch": 2.6765714285714286, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 49498 + }, + { + "epoch": 2.6766315789473687, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 49499 + }, + { + "epoch": 2.6766917293233083, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 49500 + }, + { + "epoch": 2.676751879699248, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 49501 + }, + { + "epoch": 2.676812030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 49502 + }, + { + "epoch": 2.6768721804511277, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 49503 + }, + { + "epoch": 2.676932330827068, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 49504 + }, + { + "epoch": 2.6769924812030075, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 49505 + }, + { + "epoch": 2.677052631578947, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 49506 + }, + { + "epoch": 2.677112781954887, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 49507 + }, + { + "epoch": 2.6771729323308273, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 49508 + }, + { + "epoch": 2.677233082706767, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 49509 + }, + { + "epoch": 2.6772932330827066, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 49510 + }, + { + "epoch": 2.6773533834586467, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 49511 + }, + { + "epoch": 2.6774135338345864, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 49512 + }, + { + "epoch": 2.6774736842105265, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 49513 + }, + { + "epoch": 2.677533834586466, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 49514 + }, + { + "epoch": 2.6775939849624057, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 49515 + }, + { + "epoch": 2.677654135338346, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 49516 + }, + { + "epoch": 2.677714285714286, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 49517 + }, + { + "epoch": 2.6777744360902256, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 49518 + }, + { + "epoch": 2.6778345864661652, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 49519 + }, + { + "epoch": 2.6778947368421053, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1232, + "step": 49520 + }, + { + "epoch": 2.677954887218045, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 49521 + }, + { + "epoch": 2.678015037593985, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 49522 + }, + { + "epoch": 2.6780751879699247, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 49523 + }, + { + "epoch": 2.678135338345865, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 49524 + }, + { + "epoch": 2.6781954887218045, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 49525 + }, + { + "epoch": 2.6782556390977446, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 49526 + }, + { + "epoch": 2.678315789473684, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 49527 + }, + { + "epoch": 2.678375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 49528 + }, + { + "epoch": 2.678436090225564, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 49529 + }, + { + "epoch": 2.6784962406015036, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.2507, + "step": 49530 + }, + { + "epoch": 2.6785563909774437, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 49531 + }, + { + "epoch": 2.6786165413533833, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 49532 + }, + { + "epoch": 2.6786766917293234, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 49533 + }, + { + "epoch": 2.678736842105263, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 49534 + }, + { + "epoch": 2.678796992481203, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 49535 + }, + { + "epoch": 2.678857142857143, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 49536 + }, + { + "epoch": 2.6789172932330825, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9411, + "step": 49537 + }, + { + "epoch": 2.6789774436090226, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 49538 + }, + { + "epoch": 2.6790375939849627, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 49539 + }, + { + "epoch": 2.6790977443609023, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 49540 + }, + { + "epoch": 2.679157894736842, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 49541 + }, + { + "epoch": 2.679218045112782, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 49542 + }, + { + "epoch": 2.6792781954887217, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 49543 + }, + { + "epoch": 2.679338345864662, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 49544 + }, + { + "epoch": 2.6793984962406014, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 49545 + }, + { + "epoch": 2.679458646616541, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 49546 + }, + { + "epoch": 2.679518796992481, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 49547 + }, + { + "epoch": 2.6795789473684213, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 49548 + }, + { + "epoch": 2.679639097744361, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.2293, + "step": 49549 + }, + { + "epoch": 2.6796992481203006, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 49550 + }, + { + "epoch": 2.6797593984962407, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 49551 + }, + { + "epoch": 2.6798195488721803, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 49552 + }, + { + "epoch": 2.6798796992481204, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 49553 + }, + { + "epoch": 2.67993984962406, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 49554 + }, + { + "epoch": 2.68, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 49555 + }, + { + "epoch": 2.68006015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1919, + "step": 49556 + }, + { + "epoch": 2.68012030075188, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 49557 + }, + { + "epoch": 2.6801804511278196, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 49558 + }, + { + "epoch": 2.680240601503759, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 49559 + }, + { + "epoch": 2.6803007518796993, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.929, + "step": 49560 + }, + { + "epoch": 2.680360902255639, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 49561 + }, + { + "epoch": 2.680421052631579, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 49562 + }, + { + "epoch": 2.6804812030075187, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 49563 + }, + { + "epoch": 2.680541353383459, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 49564 + }, + { + "epoch": 2.6806015037593984, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9469, + "step": 49565 + }, + { + "epoch": 2.6806616541353385, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 49566 + }, + { + "epoch": 2.680721804511278, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 49567 + }, + { + "epoch": 2.680781954887218, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 49568 + }, + { + "epoch": 2.680842105263158, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 49569 + }, + { + "epoch": 2.680902255639098, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 49570 + }, + { + "epoch": 2.6809624060150377, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 49571 + }, + { + "epoch": 2.6810225563909773, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 49572 + }, + { + "epoch": 2.6810827067669174, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 49573 + }, + { + "epoch": 2.681142857142857, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 49574 + }, + { + "epoch": 2.681203007518797, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 49575 + }, + { + "epoch": 2.681263157894737, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 49576 + }, + { + "epoch": 2.6813233082706764, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 49577 + }, + { + "epoch": 2.6813834586466165, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 49578 + }, + { + "epoch": 2.6814436090225566, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 49579 + }, + { + "epoch": 2.6815037593984963, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 49580 + }, + { + "epoch": 2.681563909774436, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 49581 + }, + { + "epoch": 2.681624060150376, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 49582 + }, + { + "epoch": 2.6816842105263157, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 49583 + }, + { + "epoch": 2.6817443609022558, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 49584 + }, + { + "epoch": 2.6818045112781954, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9306, + "step": 49585 + }, + { + "epoch": 2.681864661654135, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 49586 + }, + { + "epoch": 2.681924812030075, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 49587 + }, + { + "epoch": 2.6819849624060152, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 49588 + }, + { + "epoch": 2.682045112781955, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 49589 + }, + { + "epoch": 2.6821052631578945, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 49590 + }, + { + "epoch": 2.6821654135338346, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 49591 + }, + { + "epoch": 2.6822255639097743, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 49592 + }, + { + "epoch": 2.6822857142857144, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 49593 + }, + { + "epoch": 2.682345864661654, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 49594 + }, + { + "epoch": 2.682406015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 49595 + }, + { + "epoch": 2.6824661654135338, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 49596 + }, + { + "epoch": 2.682526315789474, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 49597 + }, + { + "epoch": 2.6825864661654135, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 49598 + }, + { + "epoch": 2.682646616541353, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 49599 + }, + { + "epoch": 2.6827067669172933, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 49600 + }, + { + "epoch": 2.6827669172932334, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 49601 + }, + { + "epoch": 2.682827067669173, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 49602 + }, + { + "epoch": 2.6828872180451127, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 49603 + }, + { + "epoch": 2.6829473684210527, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1174, + "step": 49604 + }, + { + "epoch": 2.6830075187969924, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 49605 + }, + { + "epoch": 2.6830676691729325, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 49606 + }, + { + "epoch": 2.683127819548872, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 49607 + }, + { + "epoch": 2.683187969924812, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 49608 + }, + { + "epoch": 2.683248120300752, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 49609 + }, + { + "epoch": 2.683308270676692, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 49610 + }, + { + "epoch": 2.6833684210526316, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 49611 + }, + { + "epoch": 2.6834285714285713, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 49612 + }, + { + "epoch": 2.6834887218045114, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 49613 + }, + { + "epoch": 2.683548872180451, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1356, + "step": 49614 + }, + { + "epoch": 2.683609022556391, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 49615 + }, + { + "epoch": 2.6836691729323308, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 49616 + }, + { + "epoch": 2.6837293233082704, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 49617 + }, + { + "epoch": 2.6837894736842105, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 49618 + }, + { + "epoch": 2.6838496240601506, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 49619 + }, + { + "epoch": 2.6839097744360902, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 49620 + }, + { + "epoch": 2.68396992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 49621 + }, + { + "epoch": 2.68403007518797, + "grad_norm": 2.3125, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 49622 + }, + { + "epoch": 2.6840902255639096, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 49623 + }, + { + "epoch": 2.6841503759398497, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 49624 + }, + { + "epoch": 2.6842105263157894, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 49625 + }, + { + "epoch": 2.6842706766917295, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 49626 + }, + { + "epoch": 2.684330827067669, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 49627 + }, + { + "epoch": 2.684390977443609, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 49628 + }, + { + "epoch": 2.684451127819549, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 49629 + }, + { + "epoch": 2.6845112781954885, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1177, + "step": 49630 + }, + { + "epoch": 2.6845714285714286, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 49631 + }, + { + "epoch": 2.6846315789473683, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 49632 + }, + { + "epoch": 2.6846917293233084, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.922, + "step": 49633 + }, + { + "epoch": 2.684751879699248, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 49634 + }, + { + "epoch": 2.684812030075188, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 49635 + }, + { + "epoch": 2.6848721804511277, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 49636 + }, + { + "epoch": 2.684932330827068, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 49637 + }, + { + "epoch": 2.6849924812030075, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 49638 + }, + { + "epoch": 2.685052631578947, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 49639 + }, + { + "epoch": 2.6851127819548872, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 49640 + }, + { + "epoch": 2.6851729323308273, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 49641 + }, + { + "epoch": 2.685233082706767, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 49642 + }, + { + "epoch": 2.6852932330827066, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1931, + "step": 49643 + }, + { + "epoch": 2.6853533834586467, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 49644 + }, + { + "epoch": 2.6854135338345864, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 49645 + }, + { + "epoch": 2.6854736842105265, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 49646 + }, + { + "epoch": 2.685533834586466, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 49647 + }, + { + "epoch": 2.6855939849624058, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 49648 + }, + { + "epoch": 2.685654135338346, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 49649 + }, + { + "epoch": 2.685714285714286, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 49650 + }, + { + "epoch": 2.6857744360902256, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.931, + "step": 49651 + }, + { + "epoch": 2.6858345864661652, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 49652 + }, + { + "epoch": 2.6858947368421053, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 49653 + }, + { + "epoch": 2.685954887218045, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 49654 + }, + { + "epoch": 2.686015037593985, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 49655 + }, + { + "epoch": 2.6860751879699247, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 49656 + }, + { + "epoch": 2.686135338345865, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 49657 + }, + { + "epoch": 2.6861954887218045, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 49658 + }, + { + "epoch": 2.6862556390977446, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 49659 + }, + { + "epoch": 2.686315789473684, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 49660 + }, + { + "epoch": 2.686375939849624, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 49661 + }, + { + "epoch": 2.686436090225564, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 49662 + }, + { + "epoch": 2.6864962406015036, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 49663 + }, + { + "epoch": 2.6865563909774437, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 49664 + }, + { + "epoch": 2.6866165413533833, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 49665 + }, + { + "epoch": 2.6866766917293234, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 49666 + }, + { + "epoch": 2.686736842105263, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 49667 + }, + { + "epoch": 2.686796992481203, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 49668 + }, + { + "epoch": 2.686857142857143, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 49669 + }, + { + "epoch": 2.6869172932330825, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 49670 + }, + { + "epoch": 2.6869774436090226, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 49671 + }, + { + "epoch": 2.6870375939849627, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 49672 + }, + { + "epoch": 2.6870977443609023, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 49673 + }, + { + "epoch": 2.687157894736842, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 49674 + }, + { + "epoch": 2.687218045112782, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 49675 + }, + { + "epoch": 2.6872781954887217, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 49676 + }, + { + "epoch": 2.687338345864662, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 49677 + }, + { + "epoch": 2.6873984962406015, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 49678 + }, + { + "epoch": 2.687458646616541, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 49679 + }, + { + "epoch": 2.687518796992481, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 49680 + }, + { + "epoch": 2.6875789473684213, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 49681 + }, + { + "epoch": 2.687639097744361, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 49682 + }, + { + "epoch": 2.6876992481203006, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 49683 + }, + { + "epoch": 2.6877593984962407, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 49684 + }, + { + "epoch": 2.6878195488721803, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 49685 + }, + { + "epoch": 2.6878796992481204, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 49686 + }, + { + "epoch": 2.68793984962406, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 49687 + }, + { + "epoch": 2.6879999999999997, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 49688 + }, + { + "epoch": 2.68806015037594, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 49689 + }, + { + "epoch": 2.68812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 49690 + }, + { + "epoch": 2.6881804511278196, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 49691 + }, + { + "epoch": 2.688240601503759, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 49692 + }, + { + "epoch": 2.6883007518796993, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 49693 + }, + { + "epoch": 2.688360902255639, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 49694 + }, + { + "epoch": 2.688421052631579, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 49695 + }, + { + "epoch": 2.6884812030075187, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 49696 + }, + { + "epoch": 2.688541353383459, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 49697 + }, + { + "epoch": 2.6886015037593984, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 49698 + }, + { + "epoch": 2.6886616541353385, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 49699 + }, + { + "epoch": 2.688721804511278, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 49700 + }, + { + "epoch": 2.688781954887218, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9154, + "step": 49701 + }, + { + "epoch": 2.688842105263158, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 49702 + }, + { + "epoch": 2.688902255639098, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 49703 + }, + { + "epoch": 2.6889624060150377, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 49704 + }, + { + "epoch": 2.6890225563909773, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 49705 + }, + { + "epoch": 2.6890827067669174, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 49706 + }, + { + "epoch": 2.689142857142857, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 49707 + }, + { + "epoch": 2.689203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 49708 + }, + { + "epoch": 2.689263157894737, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 49709 + }, + { + "epoch": 2.6893233082706764, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 49710 + }, + { + "epoch": 2.6893834586466165, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.84, + "step": 49711 + }, + { + "epoch": 2.6894436090225566, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 49712 + }, + { + "epoch": 2.6895037593984963, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 49713 + }, + { + "epoch": 2.689563909774436, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 49714 + }, + { + "epoch": 2.689624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 49715 + }, + { + "epoch": 2.6896842105263157, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 49716 + }, + { + "epoch": 2.6897443609022558, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9357, + "step": 49717 + }, + { + "epoch": 2.6898045112781954, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.193, + "step": 49718 + }, + { + "epoch": 2.689864661654135, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.3712, + "step": 49719 + }, + { + "epoch": 2.689924812030075, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 49720 + }, + { + "epoch": 2.6899849624060153, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 49721 + }, + { + "epoch": 2.690045112781955, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 49722 + }, + { + "epoch": 2.6901052631578946, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 49723 + }, + { + "epoch": 2.6901654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 49724 + }, + { + "epoch": 2.6902255639097743, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 49725 + }, + { + "epoch": 2.6902857142857144, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9533, + "step": 49726 + }, + { + "epoch": 2.690345864661654, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 49727 + }, + { + "epoch": 2.690406015037594, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9036, + "step": 49728 + }, + { + "epoch": 2.690466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 49729 + }, + { + "epoch": 2.690526315789474, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 49730 + }, + { + "epoch": 2.6905864661654135, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 49731 + }, + { + "epoch": 2.690646616541353, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 49732 + }, + { + "epoch": 2.6907067669172933, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 49733 + }, + { + "epoch": 2.690766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 49734 + }, + { + "epoch": 2.690827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 49735 + }, + { + "epoch": 2.6908872180451127, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 49736 + }, + { + "epoch": 2.6909473684210528, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 49737 + }, + { + "epoch": 2.6910075187969924, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 49738 + }, + { + "epoch": 2.6910676691729325, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 49739 + }, + { + "epoch": 2.691127819548872, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 49740 + }, + { + "epoch": 2.691187969924812, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 49741 + }, + { + "epoch": 2.691248120300752, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 49742 + }, + { + "epoch": 2.691308270676692, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 49743 + }, + { + "epoch": 2.6913684210526316, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 49744 + }, + { + "epoch": 2.6914285714285713, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 49745 + }, + { + "epoch": 2.6914887218045114, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 49746 + }, + { + "epoch": 2.691548872180451, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 49747 + }, + { + "epoch": 2.691609022556391, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 49748 + }, + { + "epoch": 2.6916691729323308, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 49749 + }, + { + "epoch": 2.6917293233082704, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 49750 + }, + { + "epoch": 2.6917894736842105, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 49751 + }, + { + "epoch": 2.6918496240601506, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 49752 + }, + { + "epoch": 2.6919097744360903, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 49753 + }, + { + "epoch": 2.69196992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 49754 + }, + { + "epoch": 2.69203007518797, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 49755 + }, + { + "epoch": 2.6920902255639096, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 49756 + }, + { + "epoch": 2.6921503759398497, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 49757 + }, + { + "epoch": 2.6922105263157894, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 49758 + }, + { + "epoch": 2.6922706766917295, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 49759 + }, + { + "epoch": 2.692330827067669, + "grad_norm": 1.5546875, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 49760 + }, + { + "epoch": 2.6923909774436092, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 49761 + }, + { + "epoch": 2.692451127819549, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 49762 + }, + { + "epoch": 2.6925112781954885, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 49763 + }, + { + "epoch": 2.6925714285714286, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 49764 + }, + { + "epoch": 2.6926315789473683, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 49765 + }, + { + "epoch": 2.6926917293233084, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9509, + "step": 49766 + }, + { + "epoch": 2.692751879699248, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 49767 + }, + { + "epoch": 2.692812030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.8808, + "step": 49768 + }, + { + "epoch": 2.6928721804511278, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1824, + "step": 49769 + }, + { + "epoch": 2.692932330827068, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 49770 + }, + { + "epoch": 2.6929924812030075, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 49771 + }, + { + "epoch": 2.693052631578947, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 49772 + }, + { + "epoch": 2.6931127819548872, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 49773 + }, + { + "epoch": 2.6931729323308273, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 49774 + }, + { + "epoch": 2.693233082706767, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 49775 + }, + { + "epoch": 2.6932932330827066, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 49776 + }, + { + "epoch": 2.6933533834586467, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9414, + "step": 49777 + }, + { + "epoch": 2.6934135338345864, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 49778 + }, + { + "epoch": 2.6934736842105265, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 49779 + }, + { + "epoch": 2.693533834586466, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 49780 + }, + { + "epoch": 2.6935939849624058, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 49781 + }, + { + "epoch": 2.693654135338346, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 49782 + }, + { + "epoch": 2.693714285714286, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 49783 + }, + { + "epoch": 2.6937744360902256, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 49784 + }, + { + "epoch": 2.6938345864661652, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 49785 + }, + { + "epoch": 2.6938947368421053, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 49786 + }, + { + "epoch": 2.693954887218045, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 49787 + }, + { + "epoch": 2.694015037593985, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 49788 + }, + { + "epoch": 2.6940751879699247, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 49789 + }, + { + "epoch": 2.694135338345865, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 49790 + }, + { + "epoch": 2.6941954887218045, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 49791 + }, + { + "epoch": 2.6942556390977446, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.094, + "step": 49792 + }, + { + "epoch": 2.694315789473684, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 49793 + }, + { + "epoch": 2.694375939849624, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 49794 + }, + { + "epoch": 2.694436090225564, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9061, + "step": 49795 + }, + { + "epoch": 2.6944962406015036, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 49796 + }, + { + "epoch": 2.6945563909774437, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 49797 + }, + { + "epoch": 2.6946165413533834, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 49798 + }, + { + "epoch": 2.6946766917293234, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 49799 + }, + { + "epoch": 2.694736842105263, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 49800 + }, + { + "epoch": 2.694796992481203, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 49801 + }, + { + "epoch": 2.694857142857143, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 49802 + }, + { + "epoch": 2.6949172932330825, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 49803 + }, + { + "epoch": 2.6949774436090226, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 49804 + }, + { + "epoch": 2.6950375939849627, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 49805 + }, + { + "epoch": 2.6950977443609023, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 49806 + }, + { + "epoch": 2.695157894736842, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 49807 + }, + { + "epoch": 2.695218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 49808 + }, + { + "epoch": 2.6952781954887217, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 49809 + }, + { + "epoch": 2.695338345864662, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 49810 + }, + { + "epoch": 2.6953984962406015, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 49811 + }, + { + "epoch": 2.695458646616541, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 49812 + }, + { + "epoch": 2.695518796992481, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 49813 + }, + { + "epoch": 2.6955789473684213, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 49814 + }, + { + "epoch": 2.695639097744361, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 49815 + }, + { + "epoch": 2.6956992481203006, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0961, + "step": 49816 + }, + { + "epoch": 2.6957593984962407, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 49817 + }, + { + "epoch": 2.6958195488721803, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 49818 + }, + { + "epoch": 2.6958796992481204, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 49819 + }, + { + "epoch": 2.69593984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 49820 + }, + { + "epoch": 2.6959999999999997, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 49821 + }, + { + "epoch": 2.69606015037594, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1495, + "step": 49822 + }, + { + "epoch": 2.69612030075188, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 49823 + }, + { + "epoch": 2.6961804511278196, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.2255, + "step": 49824 + }, + { + "epoch": 2.696240601503759, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 49825 + }, + { + "epoch": 2.6963007518796993, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9312, + "step": 49826 + }, + { + "epoch": 2.696360902255639, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 49827 + }, + { + "epoch": 2.696421052631579, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 49828 + }, + { + "epoch": 2.6964812030075187, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 49829 + }, + { + "epoch": 2.696541353383459, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 49830 + }, + { + "epoch": 2.6966015037593984, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9213, + "step": 49831 + }, + { + "epoch": 2.6966616541353385, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 49832 + }, + { + "epoch": 2.696721804511278, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 49833 + }, + { + "epoch": 2.696781954887218, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 49834 + }, + { + "epoch": 2.696842105263158, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 49835 + }, + { + "epoch": 2.6969022556390976, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 49836 + }, + { + "epoch": 2.6969624060150377, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 49837 + }, + { + "epoch": 2.6970225563909773, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 49838 + }, + { + "epoch": 2.6970827067669174, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 49839 + }, + { + "epoch": 2.697142857142857, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 49840 + }, + { + "epoch": 2.697203007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 49841 + }, + { + "epoch": 2.697263157894737, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 49842 + }, + { + "epoch": 2.6973233082706765, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 49843 + }, + { + "epoch": 2.6973834586466165, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 49844 + }, + { + "epoch": 2.6974436090225566, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 49845 + }, + { + "epoch": 2.6975037593984963, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 49846 + }, + { + "epoch": 2.697563909774436, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 49847 + }, + { + "epoch": 2.697624060150376, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 49848 + }, + { + "epoch": 2.6976842105263157, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 49849 + }, + { + "epoch": 2.697744360902256, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 49850 + }, + { + "epoch": 2.6978045112781954, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.944, + "step": 49851 + }, + { + "epoch": 2.697864661654135, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 49852 + }, + { + "epoch": 2.697924812030075, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 49853 + }, + { + "epoch": 2.6979849624060153, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 49854 + }, + { + "epoch": 2.698045112781955, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 49855 + }, + { + "epoch": 2.6981052631578946, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 49856 + }, + { + "epoch": 2.6981654135338347, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 49857 + }, + { + "epoch": 2.6982255639097743, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 49858 + }, + { + "epoch": 2.6982857142857144, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 49859 + }, + { + "epoch": 2.698345864661654, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 49860 + }, + { + "epoch": 2.698406015037594, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 49861 + }, + { + "epoch": 2.698466165413534, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 49862 + }, + { + "epoch": 2.698526315789474, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.3114, + "step": 49863 + }, + { + "epoch": 2.6985864661654135, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 49864 + }, + { + "epoch": 2.698646616541353, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 49865 + }, + { + "epoch": 2.6987067669172933, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 49866 + }, + { + "epoch": 2.698766917293233, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 49867 + }, + { + "epoch": 2.698827067669173, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 49868 + }, + { + "epoch": 2.6988872180451127, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 49869 + }, + { + "epoch": 2.6989473684210528, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 49870 + }, + { + "epoch": 2.6990075187969924, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 49871 + }, + { + "epoch": 2.6990676691729325, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0941, + "step": 49872 + }, + { + "epoch": 2.699127819548872, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 49873 + }, + { + "epoch": 2.699187969924812, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 49874 + }, + { + "epoch": 2.699248120300752, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 49875 + }, + { + "epoch": 2.699308270676692, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9325, + "step": 49876 + }, + { + "epoch": 2.6993684210526316, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 49877 + }, + { + "epoch": 2.6994285714285713, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.1011, + "step": 49878 + }, + { + "epoch": 2.6994887218045114, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 49879 + }, + { + "epoch": 2.699548872180451, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9323, + "step": 49880 + }, + { + "epoch": 2.699609022556391, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 49881 + }, + { + "epoch": 2.6996691729323308, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1291, + "step": 49882 + }, + { + "epoch": 2.6997293233082704, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 49883 + }, + { + "epoch": 2.6997894736842105, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1457, + "step": 49884 + }, + { + "epoch": 2.6998496240601506, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 49885 + }, + { + "epoch": 2.6999097744360903, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 49886 + }, + { + "epoch": 2.69996992481203, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 49887 + }, + { + "epoch": 2.70003007518797, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9445, + "step": 49888 + }, + { + "epoch": 2.7000902255639097, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 49889 + }, + { + "epoch": 2.7001503759398497, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 49890 + }, + { + "epoch": 2.7002105263157894, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 49891 + }, + { + "epoch": 2.7002706766917295, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 49892 + }, + { + "epoch": 2.700330827067669, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 49893 + }, + { + "epoch": 2.7003909774436092, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 49894 + }, + { + "epoch": 2.700451127819549, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 49895 + }, + { + "epoch": 2.7005112781954885, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 49896 + }, + { + "epoch": 2.7005714285714286, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 49897 + }, + { + "epoch": 2.7006315789473683, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 49898 + }, + { + "epoch": 2.7006917293233084, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 49899 + }, + { + "epoch": 2.700751879699248, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 0.9151, + "step": 49900 + }, + { + "epoch": 2.700812030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.8981, + "step": 49901 + }, + { + "epoch": 2.7008721804511278, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 49902 + }, + { + "epoch": 2.700932330827068, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 49903 + }, + { + "epoch": 2.7009924812030075, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 49904 + }, + { + "epoch": 2.701052631578947, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 49905 + }, + { + "epoch": 2.7011127819548872, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 49906 + }, + { + "epoch": 2.7011729323308273, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 49907 + }, + { + "epoch": 2.701233082706767, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 49908 + }, + { + "epoch": 2.7012932330827066, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 49909 + }, + { + "epoch": 2.7013533834586467, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 49910 + }, + { + "epoch": 2.7014135338345864, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 49911 + }, + { + "epoch": 2.7014736842105265, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 49912 + }, + { + "epoch": 2.701533834586466, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.191, + "step": 49913 + }, + { + "epoch": 2.7015939849624058, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 49914 + }, + { + "epoch": 2.701654135338346, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 49915 + }, + { + "epoch": 2.701714285714286, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 49916 + }, + { + "epoch": 2.7017744360902256, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 49917 + }, + { + "epoch": 2.7018345864661653, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 49918 + }, + { + "epoch": 2.7018947368421053, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 49919 + }, + { + "epoch": 2.701954887218045, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 49920 + }, + { + "epoch": 2.702015037593985, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 49921 + }, + { + "epoch": 2.7020751879699247, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 49922 + }, + { + "epoch": 2.7021353383458644, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9153, + "step": 49923 + }, + { + "epoch": 2.7021954887218045, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 49924 + }, + { + "epoch": 2.7022556390977446, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 49925 + }, + { + "epoch": 2.7023157894736842, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 49926 + }, + { + "epoch": 2.702375939849624, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 49927 + }, + { + "epoch": 2.702436090225564, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 49928 + }, + { + "epoch": 2.7024962406015036, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 49929 + }, + { + "epoch": 2.7025563909774437, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 49930 + }, + { + "epoch": 2.7026165413533834, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 49931 + }, + { + "epoch": 2.7026766917293235, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 49932 + }, + { + "epoch": 2.702736842105263, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 49933 + }, + { + "epoch": 2.702796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 49934 + }, + { + "epoch": 2.702857142857143, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 49935 + }, + { + "epoch": 2.7029172932330825, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 49936 + }, + { + "epoch": 2.7029774436090226, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 49937 + }, + { + "epoch": 2.7030375939849627, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 49938 + }, + { + "epoch": 2.7030977443609023, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 49939 + }, + { + "epoch": 2.703157894736842, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 49940 + }, + { + "epoch": 2.703218045112782, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 49941 + }, + { + "epoch": 2.7032781954887217, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 49942 + }, + { + "epoch": 2.703338345864662, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 49943 + }, + { + "epoch": 2.7033984962406015, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 49944 + }, + { + "epoch": 2.703458646616541, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 49945 + }, + { + "epoch": 2.703518796992481, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 49946 + }, + { + "epoch": 2.7035789473684213, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 49947 + }, + { + "epoch": 2.703639097744361, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 49948 + }, + { + "epoch": 2.7036992481203006, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 49949 + }, + { + "epoch": 2.7037593984962407, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 49950 + }, + { + "epoch": 2.7038195488721803, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 49951 + }, + { + "epoch": 2.7038796992481204, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 49952 + }, + { + "epoch": 2.70393984962406, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 49953 + }, + { + "epoch": 2.7039999999999997, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 49954 + }, + { + "epoch": 2.70406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 49955 + }, + { + "epoch": 2.70412030075188, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2424, + "step": 49956 + }, + { + "epoch": 2.7041804511278196, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 49957 + }, + { + "epoch": 2.7042406015037592, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 49958 + }, + { + "epoch": 2.7043007518796993, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 49959 + }, + { + "epoch": 2.704360902255639, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 49960 + }, + { + "epoch": 2.704421052631579, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 49961 + }, + { + "epoch": 2.7044812030075187, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.921, + "step": 49962 + }, + { + "epoch": 2.704541353383459, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 49963 + }, + { + "epoch": 2.7046015037593985, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 49964 + }, + { + "epoch": 2.7046616541353385, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 49965 + }, + { + "epoch": 2.704721804511278, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 49966 + }, + { + "epoch": 2.704781954887218, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 49967 + }, + { + "epoch": 2.704842105263158, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 49968 + }, + { + "epoch": 2.7049022556390976, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 49969 + }, + { + "epoch": 2.7049624060150377, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 49970 + }, + { + "epoch": 2.7050225563909773, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 49971 + }, + { + "epoch": 2.7050827067669174, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 49972 + }, + { + "epoch": 2.705142857142857, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 49973 + }, + { + "epoch": 2.705203007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 49974 + }, + { + "epoch": 2.705263157894737, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 49975 + }, + { + "epoch": 2.7053233082706765, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 49976 + }, + { + "epoch": 2.7053834586466166, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 49977 + }, + { + "epoch": 2.7054436090225567, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 49978 + }, + { + "epoch": 2.7055037593984963, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 49979 + }, + { + "epoch": 2.705563909774436, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 49980 + }, + { + "epoch": 2.705624060150376, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 49981 + }, + { + "epoch": 2.7056842105263157, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 49982 + }, + { + "epoch": 2.705744360902256, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 49983 + }, + { + "epoch": 2.7058045112781954, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 49984 + }, + { + "epoch": 2.705864661654135, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 49985 + }, + { + "epoch": 2.705924812030075, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 49986 + }, + { + "epoch": 2.7059849624060153, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 49987 + }, + { + "epoch": 2.706045112781955, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 49988 + }, + { + "epoch": 2.7061052631578946, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9414, + "step": 49989 + }, + { + "epoch": 2.7061654135338347, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 49990 + }, + { + "epoch": 2.7062255639097743, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1701, + "step": 49991 + }, + { + "epoch": 2.7062857142857144, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 49992 + }, + { + "epoch": 2.706345864661654, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 49993 + }, + { + "epoch": 2.706406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 49994 + }, + { + "epoch": 2.706466165413534, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 49995 + }, + { + "epoch": 2.706526315789474, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 49996 + }, + { + "epoch": 2.7065864661654135, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 49997 + }, + { + "epoch": 2.706646616541353, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 49998 + }, + { + "epoch": 2.7067067669172933, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 49999 + }, + { + "epoch": 2.706766917293233, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 50000 + }, + { + "epoch": 2.706827067669173, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 50001 + }, + { + "epoch": 2.7068872180451127, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 50002 + }, + { + "epoch": 2.7069473684210528, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9258, + "step": 50003 + }, + { + "epoch": 2.7070075187969924, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 50004 + }, + { + "epoch": 2.7070676691729325, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 50005 + }, + { + "epoch": 2.707127819548872, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 50006 + }, + { + "epoch": 2.707187969924812, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 50007 + }, + { + "epoch": 2.707248120300752, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 50008 + }, + { + "epoch": 2.707308270676692, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 50009 + }, + { + "epoch": 2.7073684210526316, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1077, + "step": 50010 + }, + { + "epoch": 2.7074285714285713, + "grad_norm": 2.578125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 50011 + }, + { + "epoch": 2.7074887218045114, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 50012 + }, + { + "epoch": 2.707548872180451, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 50013 + }, + { + "epoch": 2.707609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 50014 + }, + { + "epoch": 2.707669172932331, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 50015 + }, + { + "epoch": 2.7077293233082704, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 50016 + }, + { + "epoch": 2.7077894736842105, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 50017 + }, + { + "epoch": 2.7078496240601506, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 50018 + }, + { + "epoch": 2.7079097744360903, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 50019 + }, + { + "epoch": 2.70796992481203, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 50020 + }, + { + "epoch": 2.70803007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9187, + "step": 50021 + }, + { + "epoch": 2.7080902255639097, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 50022 + }, + { + "epoch": 2.7081503759398498, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 50023 + }, + { + "epoch": 2.7082105263157894, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 50024 + }, + { + "epoch": 2.708270676691729, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 50025 + }, + { + "epoch": 2.708330827067669, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 50026 + }, + { + "epoch": 2.7083909774436092, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 50027 + }, + { + "epoch": 2.708451127819549, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 50028 + }, + { + "epoch": 2.7085112781954885, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 50029 + }, + { + "epoch": 2.7085714285714286, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 0.9428, + "step": 50030 + }, + { + "epoch": 2.7086315789473683, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 50031 + }, + { + "epoch": 2.7086917293233084, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 50032 + }, + { + "epoch": 2.708751879699248, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 50033 + }, + { + "epoch": 2.708812030075188, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 50034 + }, + { + "epoch": 2.7088721804511278, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 50035 + }, + { + "epoch": 2.708932330827068, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 50036 + }, + { + "epoch": 2.7089924812030075, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 50037 + }, + { + "epoch": 2.709052631578947, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 50038 + }, + { + "epoch": 2.7091127819548873, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 50039 + }, + { + "epoch": 2.7091729323308273, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 50040 + }, + { + "epoch": 2.709233082706767, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 50041 + }, + { + "epoch": 2.7092932330827066, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0913, + "step": 50042 + }, + { + "epoch": 2.7093533834586467, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 50043 + }, + { + "epoch": 2.7094135338345864, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 50044 + }, + { + "epoch": 2.7094736842105265, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 50045 + }, + { + "epoch": 2.709533834586466, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 50046 + }, + { + "epoch": 2.7095939849624058, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 50047 + }, + { + "epoch": 2.709654135338346, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 50048 + }, + { + "epoch": 2.709714285714286, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 50049 + }, + { + "epoch": 2.7097744360902256, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1227, + "step": 50050 + }, + { + "epoch": 2.7098345864661653, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 50051 + }, + { + "epoch": 2.7098947368421054, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 50052 + }, + { + "epoch": 2.709954887218045, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 50053 + }, + { + "epoch": 2.710015037593985, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 50054 + }, + { + "epoch": 2.7100751879699247, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 50055 + }, + { + "epoch": 2.7101353383458644, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1147, + "step": 50056 + }, + { + "epoch": 2.7101954887218045, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9083, + "step": 50057 + }, + { + "epoch": 2.7102556390977446, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 50058 + }, + { + "epoch": 2.7103157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 50059 + }, + { + "epoch": 2.710375939849624, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 50060 + }, + { + "epoch": 2.710436090225564, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 50061 + }, + { + "epoch": 2.7104962406015036, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 50062 + }, + { + "epoch": 2.7105563909774437, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 50063 + }, + { + "epoch": 2.7106165413533834, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 50064 + }, + { + "epoch": 2.7106766917293235, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 50065 + }, + { + "epoch": 2.710736842105263, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 50066 + }, + { + "epoch": 2.710796992481203, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 50067 + }, + { + "epoch": 2.710857142857143, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 50068 + }, + { + "epoch": 2.7109172932330825, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 50069 + }, + { + "epoch": 2.7109774436090226, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 50070 + }, + { + "epoch": 2.7110375939849622, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 50071 + }, + { + "epoch": 2.7110977443609023, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 50072 + }, + { + "epoch": 2.711157894736842, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9595, + "step": 50073 + }, + { + "epoch": 2.711218045112782, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 50074 + }, + { + "epoch": 2.7112781954887217, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1256, + "step": 50075 + }, + { + "epoch": 2.711338345864662, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 50076 + }, + { + "epoch": 2.7113984962406015, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 50077 + }, + { + "epoch": 2.711458646616541, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 50078 + }, + { + "epoch": 2.711518796992481, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 50079 + }, + { + "epoch": 2.7115789473684213, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 50080 + }, + { + "epoch": 2.711639097744361, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 50081 + }, + { + "epoch": 2.7116992481203006, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 50082 + }, + { + "epoch": 2.7117593984962407, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 50083 + }, + { + "epoch": 2.7118195488721804, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 50084 + }, + { + "epoch": 2.7118796992481204, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9114, + "step": 50085 + }, + { + "epoch": 2.71193984962406, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 50086 + }, + { + "epoch": 2.7119999999999997, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 50087 + }, + { + "epoch": 2.71206015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 50088 + }, + { + "epoch": 2.71212030075188, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 50089 + }, + { + "epoch": 2.7121804511278196, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 50090 + }, + { + "epoch": 2.7122406015037592, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 50091 + }, + { + "epoch": 2.7123007518796993, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 50092 + }, + { + "epoch": 2.712360902255639, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 50093 + }, + { + "epoch": 2.712421052631579, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 50094 + }, + { + "epoch": 2.7124812030075187, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 50095 + }, + { + "epoch": 2.712541353383459, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 50096 + }, + { + "epoch": 2.7126015037593985, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 50097 + }, + { + "epoch": 2.7126616541353386, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 50098 + }, + { + "epoch": 2.712721804511278, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 50099 + }, + { + "epoch": 2.712781954887218, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 50100 + }, + { + "epoch": 2.712842105263158, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 50101 + }, + { + "epoch": 2.7129022556390976, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 50102 + }, + { + "epoch": 2.7129624060150377, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 50103 + }, + { + "epoch": 2.7130225563909773, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 50104 + }, + { + "epoch": 2.7130827067669174, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 50105 + }, + { + "epoch": 2.713142857142857, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 50106 + }, + { + "epoch": 2.713203007518797, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 50107 + }, + { + "epoch": 2.713263157894737, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 50108 + }, + { + "epoch": 2.7133233082706765, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 50109 + }, + { + "epoch": 2.7133834586466166, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9533, + "step": 50110 + }, + { + "epoch": 2.7134436090225567, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 50111 + }, + { + "epoch": 2.7135037593984963, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 50112 + }, + { + "epoch": 2.713563909774436, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 50113 + }, + { + "epoch": 2.713624060150376, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 50114 + }, + { + "epoch": 2.7136842105263157, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 50115 + }, + { + "epoch": 2.713744360902256, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 50116 + }, + { + "epoch": 2.7138045112781954, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 50117 + }, + { + "epoch": 2.713864661654135, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 50118 + }, + { + "epoch": 2.713924812030075, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 50119 + }, + { + "epoch": 2.7139849624060153, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 50120 + }, + { + "epoch": 2.714045112781955, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 50121 + }, + { + "epoch": 2.7141052631578946, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 50122 + }, + { + "epoch": 2.7141654135338347, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 50123 + }, + { + "epoch": 2.7142255639097743, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.8456, + "step": 50124 + }, + { + "epoch": 2.7142857142857144, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 50125 + }, + { + "epoch": 2.714345864661654, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 50126 + }, + { + "epoch": 2.7144060150375937, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 50127 + }, + { + "epoch": 2.714466165413534, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 50128 + }, + { + "epoch": 2.714526315789474, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 50129 + }, + { + "epoch": 2.7145864661654135, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 50130 + }, + { + "epoch": 2.714646616541353, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 50131 + }, + { + "epoch": 2.7147067669172933, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 50132 + }, + { + "epoch": 2.714766917293233, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 50133 + }, + { + "epoch": 2.714827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 50134 + }, + { + "epoch": 2.7148872180451127, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 50135 + }, + { + "epoch": 2.7149473684210528, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 50136 + }, + { + "epoch": 2.7150075187969924, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 50137 + }, + { + "epoch": 2.7150676691729325, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 50138 + }, + { + "epoch": 2.715127819548872, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9097, + "step": 50139 + }, + { + "epoch": 2.715187969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 50140 + }, + { + "epoch": 2.715248120300752, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 50141 + }, + { + "epoch": 2.715308270676692, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 50142 + }, + { + "epoch": 2.7153684210526317, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 50143 + }, + { + "epoch": 2.7154285714285713, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 50144 + }, + { + "epoch": 2.7154887218045114, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1049, + "step": 50145 + }, + { + "epoch": 2.715548872180451, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 50146 + }, + { + "epoch": 2.715609022556391, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 50147 + }, + { + "epoch": 2.715669172932331, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 50148 + }, + { + "epoch": 2.7157293233082704, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 50149 + }, + { + "epoch": 2.7157894736842105, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9037, + "step": 50150 + }, + { + "epoch": 2.7158496240601506, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9266, + "step": 50151 + }, + { + "epoch": 2.7159097744360903, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 50152 + }, + { + "epoch": 2.71596992481203, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 50153 + }, + { + "epoch": 2.71603007518797, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 50154 + }, + { + "epoch": 2.7160902255639097, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 50155 + }, + { + "epoch": 2.7161503759398498, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1285, + "step": 50156 + }, + { + "epoch": 2.7162105263157894, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 50157 + }, + { + "epoch": 2.716270676691729, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 50158 + }, + { + "epoch": 2.716330827067669, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 50159 + }, + { + "epoch": 2.7163909774436092, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 50160 + }, + { + "epoch": 2.716451127819549, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 50161 + }, + { + "epoch": 2.7165112781954885, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 50162 + }, + { + "epoch": 2.7165714285714286, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 50163 + }, + { + "epoch": 2.7166315789473683, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 50164 + }, + { + "epoch": 2.7166917293233084, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 50165 + }, + { + "epoch": 2.716751879699248, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 50166 + }, + { + "epoch": 2.716812030075188, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 50167 + }, + { + "epoch": 2.7168721804511278, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 50168 + }, + { + "epoch": 2.716932330827068, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.928, + "step": 50169 + }, + { + "epoch": 2.7169924812030075, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 50170 + }, + { + "epoch": 2.717052631578947, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 50171 + }, + { + "epoch": 2.7171127819548873, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 50172 + }, + { + "epoch": 2.717172932330827, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 50173 + }, + { + "epoch": 2.717233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 50174 + }, + { + "epoch": 2.7172932330827066, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 50175 + }, + { + "epoch": 2.7173533834586467, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 50176 + }, + { + "epoch": 2.7174135338345864, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 50177 + }, + { + "epoch": 2.7174736842105265, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 50178 + }, + { + "epoch": 2.717533834586466, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 50179 + }, + { + "epoch": 2.717593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 50180 + }, + { + "epoch": 2.717654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 50181 + }, + { + "epoch": 2.717714285714286, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 50182 + }, + { + "epoch": 2.7177744360902256, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 50183 + }, + { + "epoch": 2.7178345864661653, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 50184 + }, + { + "epoch": 2.7178947368421054, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 50185 + }, + { + "epoch": 2.717954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 50186 + }, + { + "epoch": 2.718015037593985, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 50187 + }, + { + "epoch": 2.7180751879699248, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 50188 + }, + { + "epoch": 2.7181353383458644, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 50189 + }, + { + "epoch": 2.7181954887218045, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 50190 + }, + { + "epoch": 2.7182556390977446, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 50191 + }, + { + "epoch": 2.7183157894736842, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 50192 + }, + { + "epoch": 2.718375939849624, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 50193 + }, + { + "epoch": 2.718436090225564, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9429, + "step": 50194 + }, + { + "epoch": 2.7184962406015036, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 50195 + }, + { + "epoch": 2.7185563909774437, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.8966, + "step": 50196 + }, + { + "epoch": 2.7186165413533834, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 50197 + }, + { + "epoch": 2.7186766917293235, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 50198 + }, + { + "epoch": 2.718736842105263, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 50199 + }, + { + "epoch": 2.718796992481203, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 50200 + }, + { + "epoch": 2.718857142857143, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 50201 + }, + { + "epoch": 2.7189172932330825, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 50202 + }, + { + "epoch": 2.7189774436090226, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 50203 + }, + { + "epoch": 2.7190375939849623, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 50204 + }, + { + "epoch": 2.7190977443609023, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 50205 + }, + { + "epoch": 2.719157894736842, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 50206 + }, + { + "epoch": 2.719218045112782, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 50207 + }, + { + "epoch": 2.7192781954887217, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 50208 + }, + { + "epoch": 2.719338345864662, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 50209 + }, + { + "epoch": 2.7193984962406015, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 50210 + }, + { + "epoch": 2.719458646616541, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 50211 + }, + { + "epoch": 2.7195187969924812, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 50212 + }, + { + "epoch": 2.7195789473684213, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 50213 + }, + { + "epoch": 2.719639097744361, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 50214 + }, + { + "epoch": 2.7196992481203006, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 50215 + }, + { + "epoch": 2.7197593984962407, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 50216 + }, + { + "epoch": 2.7198195488721804, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 50217 + }, + { + "epoch": 2.7198796992481205, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 50218 + }, + { + "epoch": 2.71993984962406, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 50219 + }, + { + "epoch": 2.7199999999999998, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 50220 + }, + { + "epoch": 2.72006015037594, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 50221 + }, + { + "epoch": 2.72012030075188, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 50222 + }, + { + "epoch": 2.7201804511278196, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 50223 + }, + { + "epoch": 2.7202406015037592, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 50224 + }, + { + "epoch": 2.7203007518796993, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 50225 + }, + { + "epoch": 2.720360902255639, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 50226 + }, + { + "epoch": 2.720421052631579, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 50227 + }, + { + "epoch": 2.7204812030075187, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 50228 + }, + { + "epoch": 2.7205413533834584, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 50229 + }, + { + "epoch": 2.7206015037593985, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 50230 + }, + { + "epoch": 2.7206616541353386, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 50231 + }, + { + "epoch": 2.720721804511278, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 50232 + }, + { + "epoch": 2.720781954887218, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 50233 + }, + { + "epoch": 2.720842105263158, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 50234 + }, + { + "epoch": 2.7209022556390976, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 50235 + }, + { + "epoch": 2.7209624060150377, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 50236 + }, + { + "epoch": 2.7210225563909773, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 50237 + }, + { + "epoch": 2.7210827067669174, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 50238 + }, + { + "epoch": 2.721142857142857, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 50239 + }, + { + "epoch": 2.721203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 50240 + }, + { + "epoch": 2.721263157894737, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 50241 + }, + { + "epoch": 2.7213233082706765, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 50242 + }, + { + "epoch": 2.7213834586466166, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 50243 + }, + { + "epoch": 2.7214436090225567, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.1448, + "step": 50244 + }, + { + "epoch": 2.7215037593984963, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 50245 + }, + { + "epoch": 2.721563909774436, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 50246 + }, + { + "epoch": 2.721624060150376, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 50247 + }, + { + "epoch": 2.7216842105263157, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9357, + "step": 50248 + }, + { + "epoch": 2.721744360902256, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9323, + "step": 50249 + }, + { + "epoch": 2.7218045112781954, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.8906, + "step": 50250 + }, + { + "epoch": 2.721864661654135, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 50251 + }, + { + "epoch": 2.721924812030075, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 50252 + }, + { + "epoch": 2.7219849624060153, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 50253 + }, + { + "epoch": 2.722045112781955, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 50254 + }, + { + "epoch": 2.7221052631578946, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 50255 + }, + { + "epoch": 2.7221654135338347, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 50256 + }, + { + "epoch": 2.7222255639097743, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 50257 + }, + { + "epoch": 2.7222857142857144, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 50258 + }, + { + "epoch": 2.722345864661654, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 50259 + }, + { + "epoch": 2.7224060150375937, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 50260 + }, + { + "epoch": 2.722466165413534, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 50261 + }, + { + "epoch": 2.722526315789474, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 50262 + }, + { + "epoch": 2.7225864661654136, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 50263 + }, + { + "epoch": 2.722646616541353, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 50264 + }, + { + "epoch": 2.7227067669172933, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 50265 + }, + { + "epoch": 2.722766917293233, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 50266 + }, + { + "epoch": 2.722827067669173, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 50267 + }, + { + "epoch": 2.7228872180451127, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 50268 + }, + { + "epoch": 2.722947368421053, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 50269 + }, + { + "epoch": 2.7230075187969924, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 50270 + }, + { + "epoch": 2.7230676691729325, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0641, + "step": 50271 + }, + { + "epoch": 2.723127819548872, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 50272 + }, + { + "epoch": 2.723187969924812, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9164, + "step": 50273 + }, + { + "epoch": 2.723248120300752, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 50274 + }, + { + "epoch": 2.7233082706766916, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 50275 + }, + { + "epoch": 2.7233684210526317, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 50276 + }, + { + "epoch": 2.7234285714285713, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 50277 + }, + { + "epoch": 2.7234887218045114, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 50278 + }, + { + "epoch": 2.723548872180451, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 50279 + }, + { + "epoch": 2.723609022556391, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9451, + "step": 50280 + }, + { + "epoch": 2.723669172932331, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 50281 + }, + { + "epoch": 2.7237293233082704, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1494, + "step": 50282 + }, + { + "epoch": 2.7237894736842105, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 50283 + }, + { + "epoch": 2.7238496240601506, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 50284 + }, + { + "epoch": 2.7239097744360903, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 50285 + }, + { + "epoch": 2.72396992481203, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9307, + "step": 50286 + }, + { + "epoch": 2.72403007518797, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 50287 + }, + { + "epoch": 2.7240902255639097, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 50288 + }, + { + "epoch": 2.7241503759398498, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 50289 + }, + { + "epoch": 2.7242105263157894, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 50290 + }, + { + "epoch": 2.724270676691729, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 50291 + }, + { + "epoch": 2.724330827067669, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 50292 + }, + { + "epoch": 2.7243909774436093, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 50293 + }, + { + "epoch": 2.724451127819549, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 50294 + }, + { + "epoch": 2.7245112781954886, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 50295 + }, + { + "epoch": 2.7245714285714286, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 50296 + }, + { + "epoch": 2.7246315789473683, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 50297 + }, + { + "epoch": 2.7246917293233084, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 50298 + }, + { + "epoch": 2.724751879699248, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 50299 + }, + { + "epoch": 2.724812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 50300 + }, + { + "epoch": 2.724872180451128, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 50301 + }, + { + "epoch": 2.724932330827068, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 50302 + }, + { + "epoch": 2.7249924812030075, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 50303 + }, + { + "epoch": 2.725052631578947, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 50304 + }, + { + "epoch": 2.7251127819548873, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2662, + "step": 50305 + }, + { + "epoch": 2.725172932330827, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 50306 + }, + { + "epoch": 2.725233082706767, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9482, + "step": 50307 + }, + { + "epoch": 2.7252932330827067, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 50308 + }, + { + "epoch": 2.7253533834586467, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 50309 + }, + { + "epoch": 2.7254135338345864, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0832, + "step": 50310 + }, + { + "epoch": 2.7254736842105265, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 50311 + }, + { + "epoch": 2.725533834586466, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 50312 + }, + { + "epoch": 2.725593984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 50313 + }, + { + "epoch": 2.725654135338346, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 50314 + }, + { + "epoch": 2.725714285714286, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1689, + "step": 50315 + }, + { + "epoch": 2.7257744360902256, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 50316 + }, + { + "epoch": 2.7258345864661653, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 50317 + }, + { + "epoch": 2.7258947368421054, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 50318 + }, + { + "epoch": 2.725954887218045, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 50319 + }, + { + "epoch": 2.726015037593985, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9218, + "step": 50320 + }, + { + "epoch": 2.7260751879699248, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 50321 + }, + { + "epoch": 2.7261353383458644, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 50322 + }, + { + "epoch": 2.7261954887218045, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 50323 + }, + { + "epoch": 2.7262556390977446, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 50324 + }, + { + "epoch": 2.7263157894736842, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9366, + "step": 50325 + }, + { + "epoch": 2.726375939849624, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9344, + "step": 50326 + }, + { + "epoch": 2.726436090225564, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 50327 + }, + { + "epoch": 2.7264962406015036, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 50328 + }, + { + "epoch": 2.7265563909774437, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 50329 + }, + { + "epoch": 2.7266165413533834, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 50330 + }, + { + "epoch": 2.726676691729323, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 50331 + }, + { + "epoch": 2.726736842105263, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 50332 + }, + { + "epoch": 2.726796992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.934, + "step": 50333 + }, + { + "epoch": 2.726857142857143, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 50334 + }, + { + "epoch": 2.7269172932330825, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 50335 + }, + { + "epoch": 2.7269774436090226, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 50336 + }, + { + "epoch": 2.7270375939849623, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 50337 + }, + { + "epoch": 2.7270977443609024, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1991, + "step": 50338 + }, + { + "epoch": 2.727157894736842, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 50339 + }, + { + "epoch": 2.727218045112782, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 50340 + }, + { + "epoch": 2.7272781954887217, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9362, + "step": 50341 + }, + { + "epoch": 2.727338345864662, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 50342 + }, + { + "epoch": 2.7273984962406015, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 50343 + }, + { + "epoch": 2.727458646616541, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 50344 + }, + { + "epoch": 2.7275187969924812, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 50345 + }, + { + "epoch": 2.7275789473684213, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 50346 + }, + { + "epoch": 2.727639097744361, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 50347 + }, + { + "epoch": 2.7276992481203006, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 50348 + }, + { + "epoch": 2.7277593984962407, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 50349 + }, + { + "epoch": 2.7278195488721804, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 50350 + }, + { + "epoch": 2.7278796992481205, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 50351 + }, + { + "epoch": 2.72793984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 50352 + }, + { + "epoch": 2.7279999999999998, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 50353 + }, + { + "epoch": 2.72806015037594, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 50354 + }, + { + "epoch": 2.72812030075188, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 50355 + }, + { + "epoch": 2.7281804511278196, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 50356 + }, + { + "epoch": 2.7282406015037592, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 50357 + }, + { + "epoch": 2.7283007518796993, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 50358 + }, + { + "epoch": 2.728360902255639, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 50359 + }, + { + "epoch": 2.728421052631579, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 50360 + }, + { + "epoch": 2.7284812030075187, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 50361 + }, + { + "epoch": 2.7285413533834584, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 50362 + }, + { + "epoch": 2.7286015037593985, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 50363 + }, + { + "epoch": 2.7286616541353386, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1131, + "step": 50364 + }, + { + "epoch": 2.728721804511278, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 50365 + }, + { + "epoch": 2.728781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 50366 + }, + { + "epoch": 2.728842105263158, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1917, + "step": 50367 + }, + { + "epoch": 2.7289022556390976, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 50368 + }, + { + "epoch": 2.7289624060150377, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 50369 + }, + { + "epoch": 2.7290225563909773, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 50370 + }, + { + "epoch": 2.7290827067669174, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 50371 + }, + { + "epoch": 2.729142857142857, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1565, + "step": 50372 + }, + { + "epoch": 2.729203007518797, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 50373 + }, + { + "epoch": 2.729263157894737, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9093, + "step": 50374 + }, + { + "epoch": 2.7293233082706765, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 50375 + }, + { + "epoch": 2.7293834586466166, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 50376 + }, + { + "epoch": 2.7294436090225562, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9417, + "step": 50377 + }, + { + "epoch": 2.7295037593984963, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 50378 + }, + { + "epoch": 2.729563909774436, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 50379 + }, + { + "epoch": 2.729624060150376, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 50380 + }, + { + "epoch": 2.7296842105263157, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 50381 + }, + { + "epoch": 2.729744360902256, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 50382 + }, + { + "epoch": 2.7298045112781955, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 50383 + }, + { + "epoch": 2.729864661654135, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 50384 + }, + { + "epoch": 2.729924812030075, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 50385 + }, + { + "epoch": 2.7299849624060153, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 50386 + }, + { + "epoch": 2.730045112781955, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 50387 + }, + { + "epoch": 2.7301052631578946, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 50388 + }, + { + "epoch": 2.7301654135338347, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 50389 + }, + { + "epoch": 2.7302255639097743, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 50390 + }, + { + "epoch": 2.7302857142857144, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 50391 + }, + { + "epoch": 2.730345864661654, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 50392 + }, + { + "epoch": 2.7304060150375937, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 50393 + }, + { + "epoch": 2.730466165413534, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 50394 + }, + { + "epoch": 2.730526315789474, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1948, + "step": 50395 + }, + { + "epoch": 2.7305864661654136, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9064, + "step": 50396 + }, + { + "epoch": 2.730646616541353, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1771, + "step": 50397 + }, + { + "epoch": 2.7307067669172933, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 50398 + }, + { + "epoch": 2.730766917293233, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 50399 + }, + { + "epoch": 2.730827067669173, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 50400 + }, + { + "epoch": 2.7308872180451127, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 50401 + }, + { + "epoch": 2.730947368421053, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 50402 + }, + { + "epoch": 2.7310075187969924, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 50403 + }, + { + "epoch": 2.7310676691729325, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 50404 + }, + { + "epoch": 2.731127819548872, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 50405 + }, + { + "epoch": 2.731187969924812, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 50406 + }, + { + "epoch": 2.731248120300752, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 50407 + }, + { + "epoch": 2.7313082706766916, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 50408 + }, + { + "epoch": 2.7313684210526317, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 50409 + }, + { + "epoch": 2.7314285714285713, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 50410 + }, + { + "epoch": 2.7314887218045114, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 50411 + }, + { + "epoch": 2.731548872180451, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 50412 + }, + { + "epoch": 2.731609022556391, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 50413 + }, + { + "epoch": 2.731669172932331, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 50414 + }, + { + "epoch": 2.7317293233082705, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 50415 + }, + { + "epoch": 2.7317894736842105, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 50416 + }, + { + "epoch": 2.7318496240601506, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 50417 + }, + { + "epoch": 2.7319097744360903, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 50418 + }, + { + "epoch": 2.73196992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 50419 + }, + { + "epoch": 2.73203007518797, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 50420 + }, + { + "epoch": 2.7320902255639097, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 50421 + }, + { + "epoch": 2.7321503759398498, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9514, + "step": 50422 + }, + { + "epoch": 2.7322105263157894, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 50423 + }, + { + "epoch": 2.732270676691729, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 50424 + }, + { + "epoch": 2.732330827067669, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 50425 + }, + { + "epoch": 2.7323909774436093, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 50426 + }, + { + "epoch": 2.732451127819549, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 50427 + }, + { + "epoch": 2.7325112781954886, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 50428 + }, + { + "epoch": 2.7325714285714287, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 50429 + }, + { + "epoch": 2.7326315789473683, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 50430 + }, + { + "epoch": 2.7326917293233084, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9491, + "step": 50431 + }, + { + "epoch": 2.732751879699248, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 50432 + }, + { + "epoch": 2.7328120300751877, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 50433 + }, + { + "epoch": 2.732872180451128, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 50434 + }, + { + "epoch": 2.732932330827068, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 50435 + }, + { + "epoch": 2.7329924812030075, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 50436 + }, + { + "epoch": 2.733052631578947, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 50437 + }, + { + "epoch": 2.7331127819548873, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 50438 + }, + { + "epoch": 2.733172932330827, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 50439 + }, + { + "epoch": 2.733233082706767, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 50440 + }, + { + "epoch": 2.7332932330827067, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 50441 + }, + { + "epoch": 2.7333533834586468, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 50442 + }, + { + "epoch": 2.7334135338345864, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 50443 + }, + { + "epoch": 2.7334736842105265, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 50444 + }, + { + "epoch": 2.733533834586466, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 50445 + }, + { + "epoch": 2.733593984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 50446 + }, + { + "epoch": 2.733654135338346, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 50447 + }, + { + "epoch": 2.733714285714286, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 50448 + }, + { + "epoch": 2.7337744360902256, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 50449 + }, + { + "epoch": 2.7338345864661653, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 50450 + }, + { + "epoch": 2.7338947368421054, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 50451 + }, + { + "epoch": 2.733954887218045, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 50452 + }, + { + "epoch": 2.734015037593985, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 50453 + }, + { + "epoch": 2.7340751879699248, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 50454 + }, + { + "epoch": 2.7341353383458644, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 50455 + }, + { + "epoch": 2.7341954887218045, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 50456 + }, + { + "epoch": 2.7342556390977446, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 50457 + }, + { + "epoch": 2.7343157894736843, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 50458 + }, + { + "epoch": 2.734375939849624, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.891, + "step": 50459 + }, + { + "epoch": 2.734436090225564, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 50460 + }, + { + "epoch": 2.7344962406015036, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 50461 + }, + { + "epoch": 2.7345563909774437, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 50462 + }, + { + "epoch": 2.7346165413533834, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 50463 + }, + { + "epoch": 2.734676691729323, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 50464 + }, + { + "epoch": 2.734736842105263, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 50465 + }, + { + "epoch": 2.7347969924812032, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 50466 + }, + { + "epoch": 2.734857142857143, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 50467 + }, + { + "epoch": 2.7349172932330825, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 50468 + }, + { + "epoch": 2.7349774436090226, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 50469 + }, + { + "epoch": 2.7350375939849623, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 50470 + }, + { + "epoch": 2.7350977443609024, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 50471 + }, + { + "epoch": 2.735157894736842, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 50472 + }, + { + "epoch": 2.735218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9302, + "step": 50473 + }, + { + "epoch": 2.7352781954887218, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 50474 + }, + { + "epoch": 2.735338345864662, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 50475 + }, + { + "epoch": 2.7353984962406015, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 50476 + }, + { + "epoch": 2.735458646616541, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 50477 + }, + { + "epoch": 2.7355187969924812, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9236, + "step": 50478 + }, + { + "epoch": 2.735578947368421, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 50479 + }, + { + "epoch": 2.735639097744361, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 50480 + }, + { + "epoch": 2.7356992481203006, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 50481 + }, + { + "epoch": 2.7357593984962407, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 50482 + }, + { + "epoch": 2.7358195488721804, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9281, + "step": 50483 + }, + { + "epoch": 2.7358796992481205, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9206, + "step": 50484 + }, + { + "epoch": 2.73593984962406, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 50485 + }, + { + "epoch": 2.7359999999999998, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 50486 + }, + { + "epoch": 2.73606015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9282, + "step": 50487 + }, + { + "epoch": 2.73612030075188, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9372, + "step": 50488 + }, + { + "epoch": 2.7361804511278196, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 50489 + }, + { + "epoch": 2.7362406015037593, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 50490 + }, + { + "epoch": 2.7363007518796993, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 50491 + }, + { + "epoch": 2.736360902255639, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 50492 + }, + { + "epoch": 2.736421052631579, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 50493 + }, + { + "epoch": 2.7364812030075187, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 50494 + }, + { + "epoch": 2.7365413533834584, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 50495 + }, + { + "epoch": 2.7366015037593985, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 50496 + }, + { + "epoch": 2.7366616541353386, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0953, + "step": 50497 + }, + { + "epoch": 2.7367218045112782, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 50498 + }, + { + "epoch": 2.736781954887218, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 50499 + }, + { + "epoch": 2.736842105263158, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1761, + "step": 50500 + }, + { + "epoch": 2.7369022556390976, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 50501 + }, + { + "epoch": 2.7369624060150377, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9073, + "step": 50502 + }, + { + "epoch": 2.7370225563909774, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 50503 + }, + { + "epoch": 2.7370827067669175, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9138, + "step": 50504 + }, + { + "epoch": 2.737142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 50505 + }, + { + "epoch": 2.737203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 50506 + }, + { + "epoch": 2.737263157894737, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 50507 + }, + { + "epoch": 2.7373233082706765, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 50508 + }, + { + "epoch": 2.7373834586466166, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 50509 + }, + { + "epoch": 2.7374436090225562, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 50510 + }, + { + "epoch": 2.7375037593984963, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 50511 + }, + { + "epoch": 2.737563909774436, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 50512 + }, + { + "epoch": 2.737624060150376, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 50513 + }, + { + "epoch": 2.7376842105263157, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 50514 + }, + { + "epoch": 2.737744360902256, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 50515 + }, + { + "epoch": 2.7378045112781955, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 50516 + }, + { + "epoch": 2.737864661654135, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 50517 + }, + { + "epoch": 2.737924812030075, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 50518 + }, + { + "epoch": 2.7379849624060153, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 50519 + }, + { + "epoch": 2.738045112781955, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 50520 + }, + { + "epoch": 2.7381052631578946, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 50521 + }, + { + "epoch": 2.7381654135338347, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 50522 + }, + { + "epoch": 2.7382255639097743, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 50523 + }, + { + "epoch": 2.7382857142857144, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 50524 + }, + { + "epoch": 2.738345864661654, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 50525 + }, + { + "epoch": 2.7384060150375937, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 50526 + }, + { + "epoch": 2.738466165413534, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 50527 + }, + { + "epoch": 2.738526315789474, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.2508, + "step": 50528 + }, + { + "epoch": 2.7385864661654136, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 50529 + }, + { + "epoch": 2.738646616541353, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 50530 + }, + { + "epoch": 2.7387067669172933, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 50531 + }, + { + "epoch": 2.738766917293233, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 50532 + }, + { + "epoch": 2.738827067669173, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 50533 + }, + { + "epoch": 2.7388872180451127, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 50534 + }, + { + "epoch": 2.738947368421053, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.2281, + "step": 50535 + }, + { + "epoch": 2.7390075187969924, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 50536 + }, + { + "epoch": 2.7390676691729325, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 50537 + }, + { + "epoch": 2.739127819548872, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 50538 + }, + { + "epoch": 2.739187969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.135, + "step": 50539 + }, + { + "epoch": 2.739248120300752, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 50540 + }, + { + "epoch": 2.7393082706766916, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 50541 + }, + { + "epoch": 2.7393684210526317, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 50542 + }, + { + "epoch": 2.7394285714285713, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 50543 + }, + { + "epoch": 2.7394887218045114, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0848, + "step": 50544 + }, + { + "epoch": 2.739548872180451, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 50545 + }, + { + "epoch": 2.739609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 50546 + }, + { + "epoch": 2.739669172932331, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 50547 + }, + { + "epoch": 2.7397293233082705, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 50548 + }, + { + "epoch": 2.7397894736842106, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 50549 + }, + { + "epoch": 2.7398496240601506, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 50550 + }, + { + "epoch": 2.7399097744360903, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 50551 + }, + { + "epoch": 2.73996992481203, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 50552 + }, + { + "epoch": 2.74003007518797, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 50553 + }, + { + "epoch": 2.7400902255639097, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 50554 + }, + { + "epoch": 2.74015037593985, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 50555 + }, + { + "epoch": 2.7402105263157894, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 50556 + }, + { + "epoch": 2.740270676691729, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 50557 + }, + { + "epoch": 2.740330827067669, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 50558 + }, + { + "epoch": 2.7403909774436093, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 50559 + }, + { + "epoch": 2.740451127819549, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 50560 + }, + { + "epoch": 2.7405112781954886, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1417, + "step": 50561 + }, + { + "epoch": 2.7405714285714287, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 50562 + }, + { + "epoch": 2.7406315789473683, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 50563 + }, + { + "epoch": 2.7406917293233084, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 50564 + }, + { + "epoch": 2.740751879699248, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 50565 + }, + { + "epoch": 2.7408120300751877, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 50566 + }, + { + "epoch": 2.740872180451128, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9514, + "step": 50567 + }, + { + "epoch": 2.740932330827068, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 50568 + }, + { + "epoch": 2.7409924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 50569 + }, + { + "epoch": 2.741052631578947, + "grad_norm": 1.859375, + "learning_rate": 3e-05, + "loss": 1.226, + "step": 50570 + }, + { + "epoch": 2.7411127819548873, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 50571 + }, + { + "epoch": 2.741172932330827, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 50572 + }, + { + "epoch": 2.741233082706767, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 50573 + }, + { + "epoch": 2.7412932330827067, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.095, + "step": 50574 + }, + { + "epoch": 2.7413533834586468, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 50575 + }, + { + "epoch": 2.7414135338345864, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 50576 + }, + { + "epoch": 2.7414736842105265, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 50577 + }, + { + "epoch": 2.741533834586466, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 50578 + }, + { + "epoch": 2.741593984962406, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 50579 + }, + { + "epoch": 2.741654135338346, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 50580 + }, + { + "epoch": 2.7417142857142855, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 50581 + }, + { + "epoch": 2.7417744360902256, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 50582 + }, + { + "epoch": 2.7418345864661653, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 50583 + }, + { + "epoch": 2.7418947368421054, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 50584 + }, + { + "epoch": 2.741954887218045, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 50585 + }, + { + "epoch": 2.742015037593985, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1061, + "step": 50586 + }, + { + "epoch": 2.7420751879699248, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 50587 + }, + { + "epoch": 2.7421353383458644, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 50588 + }, + { + "epoch": 2.7421954887218045, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 50589 + }, + { + "epoch": 2.7422556390977446, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 50590 + }, + { + "epoch": 2.7423157894736843, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 50591 + }, + { + "epoch": 2.742375939849624, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 50592 + }, + { + "epoch": 2.742436090225564, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 50593 + }, + { + "epoch": 2.7424962406015037, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 50594 + }, + { + "epoch": 2.7425563909774437, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 50595 + }, + { + "epoch": 2.7426165413533834, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 50596 + }, + { + "epoch": 2.742676691729323, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 50597 + }, + { + "epoch": 2.742736842105263, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 50598 + }, + { + "epoch": 2.7427969924812032, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 50599 + }, + { + "epoch": 2.742857142857143, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 50600 + }, + { + "epoch": 2.7429172932330825, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 50601 + }, + { + "epoch": 2.7429774436090226, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 50602 + }, + { + "epoch": 2.7430375939849623, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 50603 + }, + { + "epoch": 2.7430977443609024, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 50604 + }, + { + "epoch": 2.743157894736842, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 50605 + }, + { + "epoch": 2.743218045112782, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 50606 + }, + { + "epoch": 2.7432781954887218, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 50607 + }, + { + "epoch": 2.743338345864662, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9408, + "step": 50608 + }, + { + "epoch": 2.7433984962406015, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 50609 + }, + { + "epoch": 2.743458646616541, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9253, + "step": 50610 + }, + { + "epoch": 2.7435187969924812, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 50611 + }, + { + "epoch": 2.743578947368421, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 50612 + }, + { + "epoch": 2.743639097744361, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 50613 + }, + { + "epoch": 2.7436992481203006, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 50614 + }, + { + "epoch": 2.7437593984962407, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.2396, + "step": 50615 + }, + { + "epoch": 2.7438195488721804, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 50616 + }, + { + "epoch": 2.7438796992481205, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 50617 + }, + { + "epoch": 2.74393984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 50618 + }, + { + "epoch": 2.7439999999999998, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 50619 + }, + { + "epoch": 2.74406015037594, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 50620 + }, + { + "epoch": 2.74412030075188, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 50621 + }, + { + "epoch": 2.7441804511278196, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 50622 + }, + { + "epoch": 2.7442406015037593, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 50623 + }, + { + "epoch": 2.7443007518796994, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 50624 + }, + { + "epoch": 2.744360902255639, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 50625 + }, + { + "epoch": 2.744421052631579, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 50626 + }, + { + "epoch": 2.7444812030075187, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 50627 + }, + { + "epoch": 2.7445413533834584, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 50628 + }, + { + "epoch": 2.7446015037593985, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 50629 + }, + { + "epoch": 2.7446616541353386, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 50630 + }, + { + "epoch": 2.7447218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 50631 + }, + { + "epoch": 2.744781954887218, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 50632 + }, + { + "epoch": 2.744842105263158, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 50633 + }, + { + "epoch": 2.7449022556390976, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 50634 + }, + { + "epoch": 2.7449624060150377, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 50635 + }, + { + "epoch": 2.7450225563909774, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 50636 + }, + { + "epoch": 2.7450827067669175, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 50637 + }, + { + "epoch": 2.745142857142857, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 50638 + }, + { + "epoch": 2.745203007518797, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 50639 + }, + { + "epoch": 2.745263157894737, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 50640 + }, + { + "epoch": 2.7453233082706765, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 50641 + }, + { + "epoch": 2.7453834586466166, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 50642 + }, + { + "epoch": 2.7454436090225562, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9263, + "step": 50643 + }, + { + "epoch": 2.7455037593984963, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 50644 + }, + { + "epoch": 2.745563909774436, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 50645 + }, + { + "epoch": 2.745624060150376, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 50646 + }, + { + "epoch": 2.7456842105263157, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1393, + "step": 50647 + }, + { + "epoch": 2.745744360902256, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 50648 + }, + { + "epoch": 2.7458045112781955, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 50649 + }, + { + "epoch": 2.745864661654135, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 50650 + }, + { + "epoch": 2.745924812030075, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 50651 + }, + { + "epoch": 2.7459849624060153, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9347, + "step": 50652 + }, + { + "epoch": 2.746045112781955, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9212, + "step": 50653 + }, + { + "epoch": 2.7461052631578946, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 50654 + }, + { + "epoch": 2.7461654135338347, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 50655 + }, + { + "epoch": 2.7462255639097743, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 50656 + }, + { + "epoch": 2.7462857142857144, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 50657 + }, + { + "epoch": 2.746345864661654, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9001, + "step": 50658 + }, + { + "epoch": 2.7464060150375937, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 50659 + }, + { + "epoch": 2.746466165413534, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 50660 + }, + { + "epoch": 2.746526315789474, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 50661 + }, + { + "epoch": 2.7465864661654136, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 50662 + }, + { + "epoch": 2.7466466165413532, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 50663 + }, + { + "epoch": 2.7467067669172933, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 50664 + }, + { + "epoch": 2.746766917293233, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 50665 + }, + { + "epoch": 2.746827067669173, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 50666 + }, + { + "epoch": 2.7468872180451127, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 50667 + }, + { + "epoch": 2.7469473684210524, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 50668 + }, + { + "epoch": 2.7470075187969925, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 50669 + }, + { + "epoch": 2.7470676691729325, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 50670 + }, + { + "epoch": 2.747127819548872, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 50671 + }, + { + "epoch": 2.747187969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 50672 + }, + { + "epoch": 2.747248120300752, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 50673 + }, + { + "epoch": 2.7473082706766916, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0732, + "step": 50674 + }, + { + "epoch": 2.7473684210526317, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 50675 + }, + { + "epoch": 2.7474285714285713, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 50676 + }, + { + "epoch": 2.7474887218045114, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 50677 + }, + { + "epoch": 2.747548872180451, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 50678 + }, + { + "epoch": 2.747609022556391, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 50679 + }, + { + "epoch": 2.747669172932331, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 50680 + }, + { + "epoch": 2.7477293233082705, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 50681 + }, + { + "epoch": 2.7477894736842106, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 50682 + }, + { + "epoch": 2.7478496240601507, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2018, + "step": 50683 + }, + { + "epoch": 2.7479097744360903, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 50684 + }, + { + "epoch": 2.74796992481203, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 50685 + }, + { + "epoch": 2.74803007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 50686 + }, + { + "epoch": 2.7480902255639097, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 50687 + }, + { + "epoch": 2.74815037593985, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 50688 + }, + { + "epoch": 2.7482105263157894, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 50689 + }, + { + "epoch": 2.748270676691729, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.8528, + "step": 50690 + }, + { + "epoch": 2.748330827067669, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.106, + "step": 50691 + }, + { + "epoch": 2.7483909774436093, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 50692 + }, + { + "epoch": 2.748451127819549, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 50693 + }, + { + "epoch": 2.7485112781954886, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9259, + "step": 50694 + }, + { + "epoch": 2.7485714285714287, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 50695 + }, + { + "epoch": 2.7486315789473683, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 50696 + }, + { + "epoch": 2.7486917293233084, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 50697 + }, + { + "epoch": 2.748751879699248, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 50698 + }, + { + "epoch": 2.7488120300751877, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 50699 + }, + { + "epoch": 2.748872180451128, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 50700 + }, + { + "epoch": 2.748932330827068, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 50701 + }, + { + "epoch": 2.7489924812030075, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 50702 + }, + { + "epoch": 2.749052631578947, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 50703 + }, + { + "epoch": 2.7491127819548873, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9315, + "step": 50704 + }, + { + "epoch": 2.749172932330827, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 50705 + }, + { + "epoch": 2.749233082706767, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 50706 + }, + { + "epoch": 2.7492932330827067, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 50707 + }, + { + "epoch": 2.7493533834586468, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 50708 + }, + { + "epoch": 2.7494135338345864, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 50709 + }, + { + "epoch": 2.7494736842105265, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1946, + "step": 50710 + }, + { + "epoch": 2.749533834586466, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 50711 + }, + { + "epoch": 2.749593984962406, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 50712 + }, + { + "epoch": 2.749654135338346, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 50713 + }, + { + "epoch": 2.7497142857142856, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 50714 + }, + { + "epoch": 2.7497744360902256, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 50715 + }, + { + "epoch": 2.7498345864661653, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 50716 + }, + { + "epoch": 2.7498947368421054, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 50717 + }, + { + "epoch": 2.749954887218045, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 50718 + }, + { + "epoch": 2.750015037593985, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 50719 + }, + { + "epoch": 2.750075187969925, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 50720 + }, + { + "epoch": 2.7501353383458644, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 50721 + }, + { + "epoch": 2.7501954887218045, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 50722 + }, + { + "epoch": 2.7502556390977446, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 50723 + }, + { + "epoch": 2.7503157894736843, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 50724 + }, + { + "epoch": 2.750375939849624, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 50725 + }, + { + "epoch": 2.750436090225564, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9373, + "step": 50726 + }, + { + "epoch": 2.7504962406015037, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 50727 + }, + { + "epoch": 2.7505563909774438, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9406, + "step": 50728 + }, + { + "epoch": 2.7506165413533834, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 50729 + }, + { + "epoch": 2.750676691729323, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 50730 + }, + { + "epoch": 2.750736842105263, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 50731 + }, + { + "epoch": 2.7507969924812032, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 50732 + }, + { + "epoch": 2.750857142857143, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.914, + "step": 50733 + }, + { + "epoch": 2.7509172932330825, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 50734 + }, + { + "epoch": 2.7509774436090226, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 50735 + }, + { + "epoch": 2.7510375939849623, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 50736 + }, + { + "epoch": 2.7510977443609024, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 50737 + }, + { + "epoch": 2.751157894736842, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 50738 + }, + { + "epoch": 2.751218045112782, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 50739 + }, + { + "epoch": 2.7512781954887218, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 50740 + }, + { + "epoch": 2.751338345864662, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 50741 + }, + { + "epoch": 2.7513984962406015, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9308, + "step": 50742 + }, + { + "epoch": 2.751458646616541, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 50743 + }, + { + "epoch": 2.7515187969924813, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 50744 + }, + { + "epoch": 2.751578947368421, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 50745 + }, + { + "epoch": 2.751639097744361, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 50746 + }, + { + "epoch": 2.7516992481203006, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 50747 + }, + { + "epoch": 2.7517593984962407, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 50748 + }, + { + "epoch": 2.7518195488721804, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9131, + "step": 50749 + }, + { + "epoch": 2.7518796992481205, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 50750 + }, + { + "epoch": 2.75193984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 50751 + }, + { + "epoch": 2.752, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9413, + "step": 50752 + }, + { + "epoch": 2.75206015037594, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 50753 + }, + { + "epoch": 2.75212030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 50754 + }, + { + "epoch": 2.7521804511278196, + "grad_norm": 0.49609375, + "learning_rate": 3e-05, + "loss": 0.9373, + "step": 50755 + }, + { + "epoch": 2.7522406015037593, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 50756 + }, + { + "epoch": 2.7523007518796994, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 50757 + }, + { + "epoch": 2.752360902255639, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 50758 + }, + { + "epoch": 2.752421052631579, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9235, + "step": 50759 + }, + { + "epoch": 2.7524812030075188, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 50760 + }, + { + "epoch": 2.7525413533834584, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 50761 + }, + { + "epoch": 2.7526015037593985, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 50762 + }, + { + "epoch": 2.7526616541353386, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 50763 + }, + { + "epoch": 2.7527218045112782, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 50764 + }, + { + "epoch": 2.752781954887218, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 50765 + }, + { + "epoch": 2.752842105263158, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 50766 + }, + { + "epoch": 2.7529022556390976, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 50767 + }, + { + "epoch": 2.7529624060150377, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.8829, + "step": 50768 + }, + { + "epoch": 2.7530225563909774, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 50769 + }, + { + "epoch": 2.753082706766917, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 50770 + }, + { + "epoch": 2.753142857142857, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 50771 + }, + { + "epoch": 2.753203007518797, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 50772 + }, + { + "epoch": 2.753263157894737, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 50773 + }, + { + "epoch": 2.7533233082706765, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 50774 + }, + { + "epoch": 2.7533834586466166, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 50775 + }, + { + "epoch": 2.7534436090225562, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 50776 + }, + { + "epoch": 2.7535037593984963, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 50777 + }, + { + "epoch": 2.753563909774436, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 50778 + }, + { + "epoch": 2.753624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 50779 + }, + { + "epoch": 2.7536842105263157, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 50780 + }, + { + "epoch": 2.753744360902256, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 50781 + }, + { + "epoch": 2.7538045112781955, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 50782 + }, + { + "epoch": 2.753864661654135, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 50783 + }, + { + "epoch": 2.753924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 50784 + }, + { + "epoch": 2.7539849624060153, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1781, + "step": 50785 + }, + { + "epoch": 2.754045112781955, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1159, + "step": 50786 + }, + { + "epoch": 2.7541052631578946, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 50787 + }, + { + "epoch": 2.7541654135338347, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 50788 + }, + { + "epoch": 2.7542255639097744, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 50789 + }, + { + "epoch": 2.7542857142857144, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 50790 + }, + { + "epoch": 2.754345864661654, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9378, + "step": 50791 + }, + { + "epoch": 2.7544060150375937, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 50792 + }, + { + "epoch": 2.754466165413534, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 50793 + }, + { + "epoch": 2.754526315789474, + "grad_norm": 1.484375, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 50794 + }, + { + "epoch": 2.7545864661654136, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1788, + "step": 50795 + }, + { + "epoch": 2.7546466165413532, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 50796 + }, + { + "epoch": 2.7547067669172933, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 50797 + }, + { + "epoch": 2.754766917293233, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 50798 + }, + { + "epoch": 2.754827067669173, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 50799 + }, + { + "epoch": 2.7548872180451127, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 50800 + }, + { + "epoch": 2.7549473684210524, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 50801 + }, + { + "epoch": 2.7550075187969925, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 50802 + }, + { + "epoch": 2.7550676691729326, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 50803 + }, + { + "epoch": 2.755127819548872, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 50804 + }, + { + "epoch": 2.755187969924812, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 50805 + }, + { + "epoch": 2.755248120300752, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 50806 + }, + { + "epoch": 2.7553082706766916, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 50807 + }, + { + "epoch": 2.7553684210526317, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9456, + "step": 50808 + }, + { + "epoch": 2.7554285714285713, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 50809 + }, + { + "epoch": 2.7554887218045114, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 50810 + }, + { + "epoch": 2.755548872180451, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 50811 + }, + { + "epoch": 2.755609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 50812 + }, + { + "epoch": 2.755669172932331, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 50813 + }, + { + "epoch": 2.7557293233082705, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 50814 + }, + { + "epoch": 2.7557894736842106, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 50815 + }, + { + "epoch": 2.75584962406015, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 50816 + }, + { + "epoch": 2.7559097744360903, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 50817 + }, + { + "epoch": 2.75596992481203, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 50818 + }, + { + "epoch": 2.75603007518797, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1095, + "step": 50819 + }, + { + "epoch": 2.7560902255639097, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 50820 + }, + { + "epoch": 2.75615037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 50821 + }, + { + "epoch": 2.7562105263157894, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 50822 + }, + { + "epoch": 2.756270676691729, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 50823 + }, + { + "epoch": 2.756330827067669, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 50824 + }, + { + "epoch": 2.7563909774436093, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 50825 + }, + { + "epoch": 2.756451127819549, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 50826 + }, + { + "epoch": 2.7565112781954886, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 50827 + }, + { + "epoch": 2.7565714285714287, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9341, + "step": 50828 + }, + { + "epoch": 2.7566315789473683, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 50829 + }, + { + "epoch": 2.7566917293233084, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 50830 + }, + { + "epoch": 2.756751879699248, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.1332, + "step": 50831 + }, + { + "epoch": 2.7568120300751877, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 50832 + }, + { + "epoch": 2.756872180451128, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 50833 + }, + { + "epoch": 2.756932330827068, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.954, + "step": 50834 + }, + { + "epoch": 2.7569924812030076, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 50835 + }, + { + "epoch": 2.757052631578947, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 50836 + }, + { + "epoch": 2.7571127819548873, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 50837 + }, + { + "epoch": 2.757172932330827, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 50838 + }, + { + "epoch": 2.757233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 50839 + }, + { + "epoch": 2.7572932330827067, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 50840 + }, + { + "epoch": 2.757353383458647, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 50841 + }, + { + "epoch": 2.7574135338345864, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 50842 + }, + { + "epoch": 2.7574736842105265, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 50843 + }, + { + "epoch": 2.757533834586466, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 50844 + }, + { + "epoch": 2.757593984962406, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 50845 + }, + { + "epoch": 2.757654135338346, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 50846 + }, + { + "epoch": 2.7577142857142856, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 50847 + }, + { + "epoch": 2.7577744360902257, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 50848 + }, + { + "epoch": 2.7578345864661653, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 50849 + }, + { + "epoch": 2.7578947368421054, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 50850 + }, + { + "epoch": 2.757954887218045, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 50851 + }, + { + "epoch": 2.758015037593985, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 50852 + }, + { + "epoch": 2.758075187969925, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 50853 + }, + { + "epoch": 2.7581353383458644, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 50854 + }, + { + "epoch": 2.7581954887218045, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 50855 + }, + { + "epoch": 2.7582556390977446, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 50856 + }, + { + "epoch": 2.7583157894736843, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 50857 + }, + { + "epoch": 2.758375939849624, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 50858 + }, + { + "epoch": 2.758436090225564, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 50859 + }, + { + "epoch": 2.7584962406015037, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 50860 + }, + { + "epoch": 2.7585563909774438, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 50861 + }, + { + "epoch": 2.7586165413533834, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 50862 + }, + { + "epoch": 2.758676691729323, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 50863 + }, + { + "epoch": 2.758736842105263, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 50864 + }, + { + "epoch": 2.7587969924812032, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 50865 + }, + { + "epoch": 2.758857142857143, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 50866 + }, + { + "epoch": 2.7589172932330825, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 50867 + }, + { + "epoch": 2.7589774436090226, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9053, + "step": 50868 + }, + { + "epoch": 2.7590375939849623, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 50869 + }, + { + "epoch": 2.7590977443609024, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 50870 + }, + { + "epoch": 2.759157894736842, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 50871 + }, + { + "epoch": 2.7592180451127817, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 50872 + }, + { + "epoch": 2.7592781954887218, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 50873 + }, + { + "epoch": 2.759338345864662, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 50874 + }, + { + "epoch": 2.7593984962406015, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1558, + "step": 50875 + }, + { + "epoch": 2.759458646616541, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 50876 + }, + { + "epoch": 2.7595187969924813, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 50877 + }, + { + "epoch": 2.759578947368421, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9241, + "step": 50878 + }, + { + "epoch": 2.759639097744361, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 50879 + }, + { + "epoch": 2.7596992481203007, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 50880 + }, + { + "epoch": 2.7597593984962407, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0897, + "step": 50881 + }, + { + "epoch": 2.7598195488721804, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 50882 + }, + { + "epoch": 2.7598796992481205, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 50883 + }, + { + "epoch": 2.75993984962406, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 50884 + }, + { + "epoch": 2.76, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0838, + "step": 50885 + }, + { + "epoch": 2.76006015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.898, + "step": 50886 + }, + { + "epoch": 2.76012030075188, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 50887 + }, + { + "epoch": 2.7601804511278196, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 50888 + }, + { + "epoch": 2.7602406015037593, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 50889 + }, + { + "epoch": 2.7603007518796994, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 50890 + }, + { + "epoch": 2.760360902255639, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 50891 + }, + { + "epoch": 2.760421052631579, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 50892 + }, + { + "epoch": 2.7604812030075188, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 0.9514, + "step": 50893 + }, + { + "epoch": 2.7605413533834584, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 50894 + }, + { + "epoch": 2.7606015037593985, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 50895 + }, + { + "epoch": 2.7606616541353386, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 50896 + }, + { + "epoch": 2.7607218045112782, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 50897 + }, + { + "epoch": 2.760781954887218, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 50898 + }, + { + "epoch": 2.760842105263158, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 50899 + }, + { + "epoch": 2.7609022556390976, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 50900 + }, + { + "epoch": 2.7609624060150377, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 50901 + }, + { + "epoch": 2.7610225563909774, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 50902 + }, + { + "epoch": 2.761082706766917, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9429, + "step": 50903 + }, + { + "epoch": 2.761142857142857, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 50904 + }, + { + "epoch": 2.761203007518797, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 50905 + }, + { + "epoch": 2.761263157894737, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0855, + "step": 50906 + }, + { + "epoch": 2.7613233082706765, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 50907 + }, + { + "epoch": 2.7613834586466166, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 50908 + }, + { + "epoch": 2.7614436090225563, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 50909 + }, + { + "epoch": 2.7615037593984963, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 50910 + }, + { + "epoch": 2.761563909774436, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 50911 + }, + { + "epoch": 2.761624060150376, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.111, + "step": 50912 + }, + { + "epoch": 2.7616842105263157, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 50913 + }, + { + "epoch": 2.761744360902256, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 50914 + }, + { + "epoch": 2.7618045112781955, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 50915 + }, + { + "epoch": 2.761864661654135, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 50916 + }, + { + "epoch": 2.7619248120300752, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9243, + "step": 50917 + }, + { + "epoch": 2.761984962406015, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 50918 + }, + { + "epoch": 2.762045112781955, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 50919 + }, + { + "epoch": 2.7621052631578946, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 50920 + }, + { + "epoch": 2.7621654135338347, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9366, + "step": 50921 + }, + { + "epoch": 2.7622255639097744, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 50922 + }, + { + "epoch": 2.7622857142857145, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 50923 + }, + { + "epoch": 2.762345864661654, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 50924 + }, + { + "epoch": 2.7624060150375938, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 50925 + }, + { + "epoch": 2.762466165413534, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 50926 + }, + { + "epoch": 2.762526315789474, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 50927 + }, + { + "epoch": 2.7625864661654136, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9291, + "step": 50928 + }, + { + "epoch": 2.7626466165413532, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 50929 + }, + { + "epoch": 2.7627067669172933, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 50930 + }, + { + "epoch": 2.762766917293233, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 50931 + }, + { + "epoch": 2.762827067669173, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 50932 + }, + { + "epoch": 2.7628872180451127, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 50933 + }, + { + "epoch": 2.7629473684210524, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 50934 + }, + { + "epoch": 2.7630075187969925, + "grad_norm": 1.6953125, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 50935 + }, + { + "epoch": 2.7630676691729326, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 50936 + }, + { + "epoch": 2.763127819548872, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 50937 + }, + { + "epoch": 2.763187969924812, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9531, + "step": 50938 + }, + { + "epoch": 2.763248120300752, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 50939 + }, + { + "epoch": 2.7633082706766916, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 50940 + }, + { + "epoch": 2.7633684210526317, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 50941 + }, + { + "epoch": 2.7634285714285713, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 50942 + }, + { + "epoch": 2.7634887218045114, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 50943 + }, + { + "epoch": 2.763548872180451, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 50944 + }, + { + "epoch": 2.763609022556391, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 50945 + }, + { + "epoch": 2.763669172932331, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 50946 + }, + { + "epoch": 2.7637293233082705, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 50947 + }, + { + "epoch": 2.7637894736842106, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 50948 + }, + { + "epoch": 2.7638496240601502, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 50949 + }, + { + "epoch": 2.7639097744360903, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 50950 + }, + { + "epoch": 2.76396992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 50951 + }, + { + "epoch": 2.76403007518797, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 50952 + }, + { + "epoch": 2.7640902255639097, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 50953 + }, + { + "epoch": 2.76415037593985, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 50954 + }, + { + "epoch": 2.7642105263157895, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 50955 + }, + { + "epoch": 2.764270676691729, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 50956 + }, + { + "epoch": 2.764330827067669, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9345, + "step": 50957 + }, + { + "epoch": 2.7643909774436093, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 50958 + }, + { + "epoch": 2.764451127819549, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 50959 + }, + { + "epoch": 2.7645112781954886, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1881, + "step": 50960 + }, + { + "epoch": 2.7645714285714287, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 50961 + }, + { + "epoch": 2.7646315789473683, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1379, + "step": 50962 + }, + { + "epoch": 2.7646917293233084, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 50963 + }, + { + "epoch": 2.764751879699248, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9499, + "step": 50964 + }, + { + "epoch": 2.7648120300751877, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 50965 + }, + { + "epoch": 2.764872180451128, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 50966 + }, + { + "epoch": 2.764932330827068, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 50967 + }, + { + "epoch": 2.7649924812030076, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 50968 + }, + { + "epoch": 2.765052631578947, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 50969 + }, + { + "epoch": 2.7651127819548873, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 50970 + }, + { + "epoch": 2.765172932330827, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 50971 + }, + { + "epoch": 2.765233082706767, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 50972 + }, + { + "epoch": 2.7652932330827067, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9586, + "step": 50973 + }, + { + "epoch": 2.7653533834586463, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 50974 + }, + { + "epoch": 2.7654135338345864, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1916, + "step": 50975 + }, + { + "epoch": 2.7654736842105265, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 50976 + }, + { + "epoch": 2.765533834586466, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 50977 + }, + { + "epoch": 2.765593984962406, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 50978 + }, + { + "epoch": 2.765654135338346, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 50979 + }, + { + "epoch": 2.7657142857142856, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 50980 + }, + { + "epoch": 2.7657744360902257, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9285, + "step": 50981 + }, + { + "epoch": 2.7658345864661653, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 50982 + }, + { + "epoch": 2.7658947368421054, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 50983 + }, + { + "epoch": 2.765954887218045, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 50984 + }, + { + "epoch": 2.766015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 50985 + }, + { + "epoch": 2.766075187969925, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 50986 + }, + { + "epoch": 2.7661353383458644, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 50987 + }, + { + "epoch": 2.7661954887218045, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 50988 + }, + { + "epoch": 2.7662556390977446, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 50989 + }, + { + "epoch": 2.7663157894736843, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 50990 + }, + { + "epoch": 2.766375939849624, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 50991 + }, + { + "epoch": 2.766436090225564, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 50992 + }, + { + "epoch": 2.7664962406015037, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 50993 + }, + { + "epoch": 2.7665563909774438, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 50994 + }, + { + "epoch": 2.7666165413533834, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 50995 + }, + { + "epoch": 2.766676691729323, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 50996 + }, + { + "epoch": 2.766736842105263, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 50997 + }, + { + "epoch": 2.7667969924812033, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 50998 + }, + { + "epoch": 2.766857142857143, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 50999 + }, + { + "epoch": 2.7669172932330826, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 51000 + }, + { + "epoch": 2.7669774436090226, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 51001 + }, + { + "epoch": 2.7670375939849623, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 51002 + }, + { + "epoch": 2.7670977443609024, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 51003 + }, + { + "epoch": 2.767157894736842, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 51004 + }, + { + "epoch": 2.7672180451127817, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 51005 + }, + { + "epoch": 2.767278195488722, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 51006 + }, + { + "epoch": 2.767338345864662, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 51007 + }, + { + "epoch": 2.7673984962406015, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 51008 + }, + { + "epoch": 2.767458646616541, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 51009 + }, + { + "epoch": 2.7675187969924813, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 51010 + }, + { + "epoch": 2.767578947368421, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 51011 + }, + { + "epoch": 2.767639097744361, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 51012 + }, + { + "epoch": 2.7676992481203007, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 51013 + }, + { + "epoch": 2.7677593984962408, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 51014 + }, + { + "epoch": 2.7678195488721804, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 51015 + }, + { + "epoch": 2.7678796992481205, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 51016 + }, + { + "epoch": 2.76793984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 51017 + }, + { + "epoch": 2.768, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 51018 + }, + { + "epoch": 2.76806015037594, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 51019 + }, + { + "epoch": 2.7681203007518795, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 51020 + }, + { + "epoch": 2.7681804511278196, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 51021 + }, + { + "epoch": 2.7682406015037593, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 51022 + }, + { + "epoch": 2.7683007518796994, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 51023 + }, + { + "epoch": 2.768360902255639, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 51024 + }, + { + "epoch": 2.768421052631579, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 51025 + }, + { + "epoch": 2.7684812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 51026 + }, + { + "epoch": 2.7685413533834584, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 51027 + }, + { + "epoch": 2.7686015037593985, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9417, + "step": 51028 + }, + { + "epoch": 2.7686616541353386, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 51029 + }, + { + "epoch": 2.7687218045112783, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 51030 + }, + { + "epoch": 2.768781954887218, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 51031 + }, + { + "epoch": 2.768842105263158, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 51032 + }, + { + "epoch": 2.7689022556390976, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 51033 + }, + { + "epoch": 2.7689624060150377, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 51034 + }, + { + "epoch": 2.7690225563909774, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 51035 + }, + { + "epoch": 2.769082706766917, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 51036 + }, + { + "epoch": 2.769142857142857, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9298, + "step": 51037 + }, + { + "epoch": 2.769203007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 51038 + }, + { + "epoch": 2.769263157894737, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 51039 + }, + { + "epoch": 2.7693233082706765, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 51040 + }, + { + "epoch": 2.7693834586466166, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 51041 + }, + { + "epoch": 2.7694436090225563, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 51042 + }, + { + "epoch": 2.7695037593984964, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 51043 + }, + { + "epoch": 2.769563909774436, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 51044 + }, + { + "epoch": 2.769624060150376, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 51045 + }, + { + "epoch": 2.7696842105263157, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 51046 + }, + { + "epoch": 2.769744360902256, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 51047 + }, + { + "epoch": 2.7698045112781955, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 51048 + }, + { + "epoch": 2.769864661654135, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 51049 + }, + { + "epoch": 2.7699248120300752, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 51050 + }, + { + "epoch": 2.769984962406015, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.8854, + "step": 51051 + }, + { + "epoch": 2.770045112781955, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 51052 + }, + { + "epoch": 2.7701052631578946, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 51053 + }, + { + "epoch": 2.7701654135338347, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 51054 + }, + { + "epoch": 2.7702255639097744, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 51055 + }, + { + "epoch": 2.7702857142857145, + "grad_norm": 1.5625, + "learning_rate": 3e-05, + "loss": 1.2784, + "step": 51056 + }, + { + "epoch": 2.770345864661654, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 51057 + }, + { + "epoch": 2.7704060150375938, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 51058 + }, + { + "epoch": 2.770466165413534, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 51059 + }, + { + "epoch": 2.770526315789474, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 51060 + }, + { + "epoch": 2.7705864661654136, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 51061 + }, + { + "epoch": 2.7706466165413532, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 51062 + }, + { + "epoch": 2.7707067669172933, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 51063 + }, + { + "epoch": 2.770766917293233, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 51064 + }, + { + "epoch": 2.770827067669173, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.8924, + "step": 51065 + }, + { + "epoch": 2.7708872180451127, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 51066 + }, + { + "epoch": 2.7709473684210524, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 51067 + }, + { + "epoch": 2.7710075187969925, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 51068 + }, + { + "epoch": 2.7710676691729326, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 51069 + }, + { + "epoch": 2.771127819548872, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9346, + "step": 51070 + }, + { + "epoch": 2.771187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 51071 + }, + { + "epoch": 2.771248120300752, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 51072 + }, + { + "epoch": 2.7713082706766916, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 51073 + }, + { + "epoch": 2.7713684210526317, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9282, + "step": 51074 + }, + { + "epoch": 2.7714285714285714, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 51075 + }, + { + "epoch": 2.771488721804511, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 51076 + }, + { + "epoch": 2.771548872180451, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 51077 + }, + { + "epoch": 2.771609022556391, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 51078 + }, + { + "epoch": 2.771669172932331, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 51079 + }, + { + "epoch": 2.7717293233082705, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 51080 + }, + { + "epoch": 2.7717894736842106, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 51081 + }, + { + "epoch": 2.7718496240601502, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 51082 + }, + { + "epoch": 2.7719097744360903, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 51083 + }, + { + "epoch": 2.77196992481203, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 51084 + }, + { + "epoch": 2.77203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 51085 + }, + { + "epoch": 2.7720902255639097, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9501, + "step": 51086 + }, + { + "epoch": 2.77215037593985, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 51087 + }, + { + "epoch": 2.7722105263157895, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 51088 + }, + { + "epoch": 2.772270676691729, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 51089 + }, + { + "epoch": 2.772330827067669, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 51090 + }, + { + "epoch": 2.7723909774436093, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 51091 + }, + { + "epoch": 2.772451127819549, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 51092 + }, + { + "epoch": 2.7725112781954886, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 51093 + }, + { + "epoch": 2.7725714285714287, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 51094 + }, + { + "epoch": 2.7726315789473683, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 51095 + }, + { + "epoch": 2.7726917293233084, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1502, + "step": 51096 + }, + { + "epoch": 2.772751879699248, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 51097 + }, + { + "epoch": 2.7728120300751877, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 51098 + }, + { + "epoch": 2.772872180451128, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 51099 + }, + { + "epoch": 2.772932330827068, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 51100 + }, + { + "epoch": 2.7729924812030076, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 51101 + }, + { + "epoch": 2.773052631578947, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9509, + "step": 51102 + }, + { + "epoch": 2.7731127819548873, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 51103 + }, + { + "epoch": 2.773172932330827, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 51104 + }, + { + "epoch": 2.773233082706767, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 51105 + }, + { + "epoch": 2.7732932330827067, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 51106 + }, + { + "epoch": 2.7733533834586463, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9204, + "step": 51107 + }, + { + "epoch": 2.7734135338345864, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 51108 + }, + { + "epoch": 2.7734736842105265, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 51109 + }, + { + "epoch": 2.773533834586466, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 51110 + }, + { + "epoch": 2.773593984962406, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 51111 + }, + { + "epoch": 2.773654135338346, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 51112 + }, + { + "epoch": 2.7737142857142856, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 51113 + }, + { + "epoch": 2.7737744360902257, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9451, + "step": 51114 + }, + { + "epoch": 2.7738345864661653, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 51115 + }, + { + "epoch": 2.7738947368421054, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 51116 + }, + { + "epoch": 2.773954887218045, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.8748, + "step": 51117 + }, + { + "epoch": 2.774015037593985, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 51118 + }, + { + "epoch": 2.774075187969925, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.956, + "step": 51119 + }, + { + "epoch": 2.7741353383458645, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 51120 + }, + { + "epoch": 2.7741954887218045, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 51121 + }, + { + "epoch": 2.774255639097744, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 51122 + }, + { + "epoch": 2.7743157894736843, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 51123 + }, + { + "epoch": 2.774375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 51124 + }, + { + "epoch": 2.774436090225564, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 51125 + }, + { + "epoch": 2.7744962406015037, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 51126 + }, + { + "epoch": 2.7745563909774438, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 51127 + }, + { + "epoch": 2.7746165413533834, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 51128 + }, + { + "epoch": 2.774676691729323, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 51129 + }, + { + "epoch": 2.774736842105263, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 51130 + }, + { + "epoch": 2.7747969924812033, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 51131 + }, + { + "epoch": 2.774857142857143, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 51132 + }, + { + "epoch": 2.7749172932330826, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 51133 + }, + { + "epoch": 2.7749774436090227, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1815, + "step": 51134 + }, + { + "epoch": 2.7750375939849623, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 51135 + }, + { + "epoch": 2.7750977443609024, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 51136 + }, + { + "epoch": 2.775157894736842, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 51137 + }, + { + "epoch": 2.7752180451127817, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 51138 + }, + { + "epoch": 2.775278195488722, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 51139 + }, + { + "epoch": 2.775338345864662, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 51140 + }, + { + "epoch": 2.7753984962406015, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 51141 + }, + { + "epoch": 2.775458646616541, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 51142 + }, + { + "epoch": 2.7755187969924813, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 51143 + }, + { + "epoch": 2.775578947368421, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 51144 + }, + { + "epoch": 2.775639097744361, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 51145 + }, + { + "epoch": 2.7756992481203007, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 51146 + }, + { + "epoch": 2.7757593984962408, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 51147 + }, + { + "epoch": 2.7758195488721804, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 51148 + }, + { + "epoch": 2.7758796992481205, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 51149 + }, + { + "epoch": 2.77593984962406, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 51150 + }, + { + "epoch": 2.776, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 51151 + }, + { + "epoch": 2.77606015037594, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 51152 + }, + { + "epoch": 2.7761203007518795, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9336, + "step": 51153 + }, + { + "epoch": 2.7761804511278196, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 51154 + }, + { + "epoch": 2.7762406015037593, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 51155 + }, + { + "epoch": 2.7763007518796994, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 51156 + }, + { + "epoch": 2.776360902255639, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 51157 + }, + { + "epoch": 2.776421052631579, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1248, + "step": 51158 + }, + { + "epoch": 2.7764812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 51159 + }, + { + "epoch": 2.7765413533834584, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 51160 + }, + { + "epoch": 2.7766015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 51161 + }, + { + "epoch": 2.7766616541353386, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 51162 + }, + { + "epoch": 2.7767218045112783, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9351, + "step": 51163 + }, + { + "epoch": 2.776781954887218, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 51164 + }, + { + "epoch": 2.776842105263158, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 51165 + }, + { + "epoch": 2.7769022556390976, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9081, + "step": 51166 + }, + { + "epoch": 2.7769624060150377, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 51167 + }, + { + "epoch": 2.7770225563909774, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 51168 + }, + { + "epoch": 2.777082706766917, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 51169 + }, + { + "epoch": 2.777142857142857, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 51170 + }, + { + "epoch": 2.7772030075187972, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 51171 + }, + { + "epoch": 2.777263157894737, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9146, + "step": 51172 + }, + { + "epoch": 2.7773233082706765, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 51173 + }, + { + "epoch": 2.7773834586466166, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 51174 + }, + { + "epoch": 2.7774436090225563, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 51175 + }, + { + "epoch": 2.7775037593984964, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1655, + "step": 51176 + }, + { + "epoch": 2.777563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 51177 + }, + { + "epoch": 2.777624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9374, + "step": 51178 + }, + { + "epoch": 2.7776842105263158, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 51179 + }, + { + "epoch": 2.777744360902256, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 51180 + }, + { + "epoch": 2.7778045112781955, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 51181 + }, + { + "epoch": 2.777864661654135, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 51182 + }, + { + "epoch": 2.7779248120300752, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 51183 + }, + { + "epoch": 2.777984962406015, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 51184 + }, + { + "epoch": 2.778045112781955, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 51185 + }, + { + "epoch": 2.7781052631578946, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 51186 + }, + { + "epoch": 2.7781654135338347, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 51187 + }, + { + "epoch": 2.7782255639097744, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1596, + "step": 51188 + }, + { + "epoch": 2.7782857142857145, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 51189 + }, + { + "epoch": 2.778345864661654, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 51190 + }, + { + "epoch": 2.7784060150375938, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 51191 + }, + { + "epoch": 2.778466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.188, + "step": 51192 + }, + { + "epoch": 2.778526315789474, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 51193 + }, + { + "epoch": 2.7785864661654136, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 51194 + }, + { + "epoch": 2.7786466165413533, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 51195 + }, + { + "epoch": 2.7787067669172933, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1522, + "step": 51196 + }, + { + "epoch": 2.778766917293233, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 51197 + }, + { + "epoch": 2.778827067669173, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 51198 + }, + { + "epoch": 2.7788872180451127, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 51199 + }, + { + "epoch": 2.7789473684210524, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 51200 + }, + { + "epoch": 2.7790075187969925, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 51201 + }, + { + "epoch": 2.7790676691729326, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 51202 + }, + { + "epoch": 2.7791278195488722, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 51203 + }, + { + "epoch": 2.779187969924812, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.8912, + "step": 51204 + }, + { + "epoch": 2.779248120300752, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 51205 + }, + { + "epoch": 2.7793082706766916, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9509, + "step": 51206 + }, + { + "epoch": 2.7793684210526317, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 51207 + }, + { + "epoch": 2.7794285714285714, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 51208 + }, + { + "epoch": 2.779488721804511, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9403, + "step": 51209 + }, + { + "epoch": 2.779548872180451, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 51210 + }, + { + "epoch": 2.779609022556391, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9395, + "step": 51211 + }, + { + "epoch": 2.779669172932331, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 51212 + }, + { + "epoch": 2.7797293233082705, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 51213 + }, + { + "epoch": 2.7797894736842106, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 51214 + }, + { + "epoch": 2.7798496240601502, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 51215 + }, + { + "epoch": 2.7799097744360903, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 51216 + }, + { + "epoch": 2.77996992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 51217 + }, + { + "epoch": 2.78003007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 51218 + }, + { + "epoch": 2.7800902255639097, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 51219 + }, + { + "epoch": 2.78015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 51220 + }, + { + "epoch": 2.7802105263157895, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 51221 + }, + { + "epoch": 2.780270676691729, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 51222 + }, + { + "epoch": 2.780330827067669, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 51223 + }, + { + "epoch": 2.780390977443609, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 51224 + }, + { + "epoch": 2.780451127819549, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 51225 + }, + { + "epoch": 2.7805112781954886, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 51226 + }, + { + "epoch": 2.7805714285714287, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 51227 + }, + { + "epoch": 2.7806315789473683, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 51228 + }, + { + "epoch": 2.7806917293233084, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 51229 + }, + { + "epoch": 2.780751879699248, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 51230 + }, + { + "epoch": 2.7808120300751877, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 51231 + }, + { + "epoch": 2.780872180451128, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.2281, + "step": 51232 + }, + { + "epoch": 2.780932330827068, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 51233 + }, + { + "epoch": 2.7809924812030076, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 51234 + }, + { + "epoch": 2.781052631578947, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 51235 + }, + { + "epoch": 2.7811127819548873, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 51236 + }, + { + "epoch": 2.781172932330827, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 51237 + }, + { + "epoch": 2.781233082706767, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 51238 + }, + { + "epoch": 2.7812932330827067, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 51239 + }, + { + "epoch": 2.7813533834586464, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 51240 + }, + { + "epoch": 2.7814135338345864, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 51241 + }, + { + "epoch": 2.7814736842105265, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 51242 + }, + { + "epoch": 2.781533834586466, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 51243 + }, + { + "epoch": 2.781593984962406, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 51244 + }, + { + "epoch": 2.781654135338346, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 51245 + }, + { + "epoch": 2.7817142857142856, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 51246 + }, + { + "epoch": 2.7817744360902257, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 51247 + }, + { + "epoch": 2.7818345864661653, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 51248 + }, + { + "epoch": 2.7818947368421054, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 51249 + }, + { + "epoch": 2.781954887218045, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 51250 + }, + { + "epoch": 2.782015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 51251 + }, + { + "epoch": 2.782075187969925, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 51252 + }, + { + "epoch": 2.7821353383458645, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 51253 + }, + { + "epoch": 2.7821954887218046, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 51254 + }, + { + "epoch": 2.782255639097744, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 51255 + }, + { + "epoch": 2.7823157894736843, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 51256 + }, + { + "epoch": 2.782375939849624, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 51257 + }, + { + "epoch": 2.782436090225564, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 51258 + }, + { + "epoch": 2.7824962406015037, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9299, + "step": 51259 + }, + { + "epoch": 2.782556390977444, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 51260 + }, + { + "epoch": 2.7826165413533834, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 51261 + }, + { + "epoch": 2.782676691729323, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 51262 + }, + { + "epoch": 2.782736842105263, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 51263 + }, + { + "epoch": 2.7827969924812033, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1122, + "step": 51264 + }, + { + "epoch": 2.782857142857143, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 51265 + }, + { + "epoch": 2.7829172932330826, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 51266 + }, + { + "epoch": 2.7829774436090227, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 51267 + }, + { + "epoch": 2.7830375939849623, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 51268 + }, + { + "epoch": 2.7830977443609024, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 51269 + }, + { + "epoch": 2.783157894736842, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 51270 + }, + { + "epoch": 2.7832180451127817, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 51271 + }, + { + "epoch": 2.783278195488722, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 51272 + }, + { + "epoch": 2.783338345864662, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1544, + "step": 51273 + }, + { + "epoch": 2.7833984962406015, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 51274 + }, + { + "epoch": 2.783458646616541, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 51275 + }, + { + "epoch": 2.7835187969924813, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 51276 + }, + { + "epoch": 2.783578947368421, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 51277 + }, + { + "epoch": 2.783639097744361, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 51278 + }, + { + "epoch": 2.7836992481203007, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9061, + "step": 51279 + }, + { + "epoch": 2.7837593984962408, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 51280 + }, + { + "epoch": 2.7838195488721804, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 51281 + }, + { + "epoch": 2.7838796992481205, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 51282 + }, + { + "epoch": 2.78393984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 51283 + }, + { + "epoch": 2.784, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9112, + "step": 51284 + }, + { + "epoch": 2.78406015037594, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 51285 + }, + { + "epoch": 2.7841203007518796, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 51286 + }, + { + "epoch": 2.7841804511278196, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 51287 + }, + { + "epoch": 2.7842406015037593, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 51288 + }, + { + "epoch": 2.7843007518796994, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 51289 + }, + { + "epoch": 2.784360902255639, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2198, + "step": 51290 + }, + { + "epoch": 2.784421052631579, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9176, + "step": 51291 + }, + { + "epoch": 2.784481203007519, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 51292 + }, + { + "epoch": 2.7845413533834584, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0999, + "step": 51293 + }, + { + "epoch": 2.7846015037593985, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 51294 + }, + { + "epoch": 2.7846616541353386, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 51295 + }, + { + "epoch": 2.7847218045112783, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 51296 + }, + { + "epoch": 2.784781954887218, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 51297 + }, + { + "epoch": 2.784842105263158, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 51298 + }, + { + "epoch": 2.7849022556390977, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 51299 + }, + { + "epoch": 2.7849624060150378, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 51300 + }, + { + "epoch": 2.7850225563909774, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 51301 + }, + { + "epoch": 2.785082706766917, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 51302 + }, + { + "epoch": 2.785142857142857, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 51303 + }, + { + "epoch": 2.7852030075187972, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 51304 + }, + { + "epoch": 2.785263157894737, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 51305 + }, + { + "epoch": 2.7853233082706765, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 51306 + }, + { + "epoch": 2.7853834586466166, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 51307 + }, + { + "epoch": 2.7854436090225563, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 51308 + }, + { + "epoch": 2.7855037593984964, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 51309 + }, + { + "epoch": 2.785563909774436, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 51310 + }, + { + "epoch": 2.7856240601503757, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 51311 + }, + { + "epoch": 2.7856842105263158, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 51312 + }, + { + "epoch": 2.785744360902256, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 51313 + }, + { + "epoch": 2.7858045112781955, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 51314 + }, + { + "epoch": 2.785864661654135, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 51315 + }, + { + "epoch": 2.7859248120300752, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 51316 + }, + { + "epoch": 2.785984962406015, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 51317 + }, + { + "epoch": 2.786045112781955, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 51318 + }, + { + "epoch": 2.7861052631578946, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9527, + "step": 51319 + }, + { + "epoch": 2.7861654135338347, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 51320 + }, + { + "epoch": 2.7862255639097744, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 51321 + }, + { + "epoch": 2.7862857142857145, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 51322 + }, + { + "epoch": 2.786345864661654, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 51323 + }, + { + "epoch": 2.7864060150375938, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 51324 + }, + { + "epoch": 2.786466165413534, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 51325 + }, + { + "epoch": 2.7865263157894735, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 51326 + }, + { + "epoch": 2.7865864661654136, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 51327 + }, + { + "epoch": 2.7866466165413533, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 51328 + }, + { + "epoch": 2.7867067669172934, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 51329 + }, + { + "epoch": 2.786766917293233, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9304, + "step": 51330 + }, + { + "epoch": 2.786827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 51331 + }, + { + "epoch": 2.7868872180451127, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 51332 + }, + { + "epoch": 2.7869473684210524, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 51333 + }, + { + "epoch": 2.7870075187969925, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 51334 + }, + { + "epoch": 2.7870676691729326, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 51335 + }, + { + "epoch": 2.7871278195488722, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 51336 + }, + { + "epoch": 2.787187969924812, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 51337 + }, + { + "epoch": 2.787248120300752, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 51338 + }, + { + "epoch": 2.7873082706766916, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 51339 + }, + { + "epoch": 2.7873684210526317, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 51340 + }, + { + "epoch": 2.7874285714285714, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 51341 + }, + { + "epoch": 2.787488721804511, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 51342 + }, + { + "epoch": 2.787548872180451, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9323, + "step": 51343 + }, + { + "epoch": 2.787609022556391, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 51344 + }, + { + "epoch": 2.787669172932331, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 51345 + }, + { + "epoch": 2.7877293233082705, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 51346 + }, + { + "epoch": 2.7877894736842106, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.2003, + "step": 51347 + }, + { + "epoch": 2.7878496240601502, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 51348 + }, + { + "epoch": 2.7879097744360903, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 51349 + }, + { + "epoch": 2.78796992481203, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 51350 + }, + { + "epoch": 2.78803007518797, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 51351 + }, + { + "epoch": 2.7880902255639097, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 51352 + }, + { + "epoch": 2.78815037593985, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 51353 + }, + { + "epoch": 2.7882105263157895, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 51354 + }, + { + "epoch": 2.788270676691729, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 51355 + }, + { + "epoch": 2.788330827067669, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 51356 + }, + { + "epoch": 2.788390977443609, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 51357 + }, + { + "epoch": 2.788451127819549, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 51358 + }, + { + "epoch": 2.7885112781954886, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 51359 + }, + { + "epoch": 2.7885714285714287, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 51360 + }, + { + "epoch": 2.7886315789473684, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 51361 + }, + { + "epoch": 2.7886917293233084, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 51362 + }, + { + "epoch": 2.788751879699248, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 51363 + }, + { + "epoch": 2.7888120300751877, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 51364 + }, + { + "epoch": 2.788872180451128, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 51365 + }, + { + "epoch": 2.788932330827068, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 51366 + }, + { + "epoch": 2.7889924812030076, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 51367 + }, + { + "epoch": 2.7890526315789472, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9189, + "step": 51368 + }, + { + "epoch": 2.7891127819548873, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 51369 + }, + { + "epoch": 2.789172932330827, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 51370 + }, + { + "epoch": 2.789233082706767, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 51371 + }, + { + "epoch": 2.7892932330827067, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 51372 + }, + { + "epoch": 2.7893533834586464, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 51373 + }, + { + "epoch": 2.7894135338345865, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 51374 + }, + { + "epoch": 2.7894736842105265, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 51375 + }, + { + "epoch": 2.789533834586466, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 51376 + }, + { + "epoch": 2.789593984962406, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 51377 + }, + { + "epoch": 2.789654135338346, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 51378 + }, + { + "epoch": 2.7897142857142856, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9427, + "step": 51379 + }, + { + "epoch": 2.7897744360902257, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 51380 + }, + { + "epoch": 2.7898345864661653, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 51381 + }, + { + "epoch": 2.7898947368421054, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 51382 + }, + { + "epoch": 2.789954887218045, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 51383 + }, + { + "epoch": 2.790015037593985, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 51384 + }, + { + "epoch": 2.790075187969925, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 51385 + }, + { + "epoch": 2.7901353383458645, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1738, + "step": 51386 + }, + { + "epoch": 2.7901954887218046, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 51387 + }, + { + "epoch": 2.790255639097744, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9487, + "step": 51388 + }, + { + "epoch": 2.7903157894736843, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 51389 + }, + { + "epoch": 2.790375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 51390 + }, + { + "epoch": 2.790436090225564, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 51391 + }, + { + "epoch": 2.7904962406015037, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.2097, + "step": 51392 + }, + { + "epoch": 2.790556390977444, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 51393 + }, + { + "epoch": 2.7906165413533834, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 51394 + }, + { + "epoch": 2.790676691729323, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 51395 + }, + { + "epoch": 2.790736842105263, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 51396 + }, + { + "epoch": 2.7907969924812033, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 51397 + }, + { + "epoch": 2.790857142857143, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 51398 + }, + { + "epoch": 2.7909172932330826, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 51399 + }, + { + "epoch": 2.7909774436090227, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 51400 + }, + { + "epoch": 2.7910375939849623, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 51401 + }, + { + "epoch": 2.7910977443609024, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 51402 + }, + { + "epoch": 2.791157894736842, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 51403 + }, + { + "epoch": 2.7912180451127817, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 51404 + }, + { + "epoch": 2.791278195488722, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 51405 + }, + { + "epoch": 2.791338345864662, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 51406 + }, + { + "epoch": 2.7913984962406015, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 51407 + }, + { + "epoch": 2.791458646616541, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 51408 + }, + { + "epoch": 2.7915187969924813, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 51409 + }, + { + "epoch": 2.791578947368421, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 51410 + }, + { + "epoch": 2.791639097744361, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 51411 + }, + { + "epoch": 2.7916992481203007, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 51412 + }, + { + "epoch": 2.7917593984962403, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 51413 + }, + { + "epoch": 2.7918195488721804, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 51414 + }, + { + "epoch": 2.7918796992481205, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 51415 + }, + { + "epoch": 2.79193984962406, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9414, + "step": 51416 + }, + { + "epoch": 2.792, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 51417 + }, + { + "epoch": 2.79206015037594, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 51418 + }, + { + "epoch": 2.7921203007518796, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 51419 + }, + { + "epoch": 2.7921804511278197, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 51420 + }, + { + "epoch": 2.7922406015037593, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 51421 + }, + { + "epoch": 2.7923007518796994, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 51422 + }, + { + "epoch": 2.792360902255639, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 51423 + }, + { + "epoch": 2.792421052631579, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 51424 + }, + { + "epoch": 2.792481203007519, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 51425 + }, + { + "epoch": 2.7925413533834584, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 51426 + }, + { + "epoch": 2.7926015037593985, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 51427 + }, + { + "epoch": 2.7926616541353386, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9053, + "step": 51428 + }, + { + "epoch": 2.7927218045112783, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 51429 + }, + { + "epoch": 2.792781954887218, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 51430 + }, + { + "epoch": 2.792842105263158, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 51431 + }, + { + "epoch": 2.7929022556390977, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 51432 + }, + { + "epoch": 2.7929624060150378, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1372, + "step": 51433 + }, + { + "epoch": 2.7930225563909774, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 51434 + }, + { + "epoch": 2.793082706766917, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 51435 + }, + { + "epoch": 2.793142857142857, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 51436 + }, + { + "epoch": 2.7932030075187972, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 51437 + }, + { + "epoch": 2.793263157894737, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 51438 + }, + { + "epoch": 2.7933233082706765, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 51439 + }, + { + "epoch": 2.7933834586466166, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 51440 + }, + { + "epoch": 2.7934436090225563, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 51441 + }, + { + "epoch": 2.7935037593984964, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 51442 + }, + { + "epoch": 2.793563909774436, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 51443 + }, + { + "epoch": 2.7936240601503757, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 51444 + }, + { + "epoch": 2.7936842105263158, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 51445 + }, + { + "epoch": 2.793744360902256, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 51446 + }, + { + "epoch": 2.7938045112781955, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 51447 + }, + { + "epoch": 2.793864661654135, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 51448 + }, + { + "epoch": 2.7939248120300753, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 51449 + }, + { + "epoch": 2.793984962406015, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 51450 + }, + { + "epoch": 2.794045112781955, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 51451 + }, + { + "epoch": 2.7941052631578946, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 51452 + }, + { + "epoch": 2.7941654135338347, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 51453 + }, + { + "epoch": 2.7942255639097744, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 51454 + }, + { + "epoch": 2.7942857142857145, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 51455 + }, + { + "epoch": 2.794345864661654, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 51456 + }, + { + "epoch": 2.794406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 51457 + }, + { + "epoch": 2.794466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 51458 + }, + { + "epoch": 2.7945263157894735, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9244, + "step": 51459 + }, + { + "epoch": 2.7945864661654136, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9406, + "step": 51460 + }, + { + "epoch": 2.7946466165413533, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 51461 + }, + { + "epoch": 2.7947067669172934, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 51462 + }, + { + "epoch": 2.794766917293233, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1981, + "step": 51463 + }, + { + "epoch": 2.794827067669173, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 51464 + }, + { + "epoch": 2.7948872180451128, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 51465 + }, + { + "epoch": 2.7949473684210524, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 51466 + }, + { + "epoch": 2.7950075187969925, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 51467 + }, + { + "epoch": 2.7950676691729326, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 51468 + }, + { + "epoch": 2.7951278195488722, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 51469 + }, + { + "epoch": 2.795187969924812, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 51470 + }, + { + "epoch": 2.795248120300752, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 51471 + }, + { + "epoch": 2.7953082706766916, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 51472 + }, + { + "epoch": 2.7953684210526317, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 51473 + }, + { + "epoch": 2.7954285714285714, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 51474 + }, + { + "epoch": 2.795488721804511, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 51475 + }, + { + "epoch": 2.795548872180451, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 51476 + }, + { + "epoch": 2.795609022556391, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 51477 + }, + { + "epoch": 2.795669172932331, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 51478 + }, + { + "epoch": 2.7957293233082705, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 51479 + }, + { + "epoch": 2.7957894736842106, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9411, + "step": 51480 + }, + { + "epoch": 2.7958496240601503, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 51481 + }, + { + "epoch": 2.7959097744360903, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 51482 + }, + { + "epoch": 2.79596992481203, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 51483 + }, + { + "epoch": 2.79603007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 51484 + }, + { + "epoch": 2.7960902255639097, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9288, + "step": 51485 + }, + { + "epoch": 2.79615037593985, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 51486 + }, + { + "epoch": 2.7962105263157895, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 51487 + }, + { + "epoch": 2.796270676691729, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 51488 + }, + { + "epoch": 2.7963308270676692, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 51489 + }, + { + "epoch": 2.796390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 51490 + }, + { + "epoch": 2.796451127819549, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1259, + "step": 51491 + }, + { + "epoch": 2.7965112781954886, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 51492 + }, + { + "epoch": 2.7965714285714287, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9369, + "step": 51493 + }, + { + "epoch": 2.7966315789473684, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 51494 + }, + { + "epoch": 2.7966917293233085, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 51495 + }, + { + "epoch": 2.796751879699248, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 51496 + }, + { + "epoch": 2.7968120300751877, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 51497 + }, + { + "epoch": 2.796872180451128, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 51498 + }, + { + "epoch": 2.796932330827068, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 51499 + }, + { + "epoch": 2.7969924812030076, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9481, + "step": 51500 + }, + { + "epoch": 2.7970526315789472, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 51501 + }, + { + "epoch": 2.7971127819548873, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1722, + "step": 51502 + }, + { + "epoch": 2.797172932330827, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 51503 + }, + { + "epoch": 2.797233082706767, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 51504 + }, + { + "epoch": 2.7972932330827067, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 51505 + }, + { + "epoch": 2.7973533834586464, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 51506 + }, + { + "epoch": 2.7974135338345865, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 51507 + }, + { + "epoch": 2.7974736842105266, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 51508 + }, + { + "epoch": 2.797533834586466, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 51509 + }, + { + "epoch": 2.797593984962406, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 51510 + }, + { + "epoch": 2.797654135338346, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 51511 + }, + { + "epoch": 2.7977142857142856, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 51512 + }, + { + "epoch": 2.7977744360902257, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 51513 + }, + { + "epoch": 2.7978345864661653, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 51514 + }, + { + "epoch": 2.797894736842105, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 51515 + }, + { + "epoch": 2.797954887218045, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 51516 + }, + { + "epoch": 2.798015037593985, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9166, + "step": 51517 + }, + { + "epoch": 2.798075187969925, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 51518 + }, + { + "epoch": 2.7981353383458645, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 51519 + }, + { + "epoch": 2.7981954887218046, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 51520 + }, + { + "epoch": 2.798255639097744, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 51521 + }, + { + "epoch": 2.7983157894736843, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 51522 + }, + { + "epoch": 2.798375939849624, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 51523 + }, + { + "epoch": 2.798436090225564, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 51524 + }, + { + "epoch": 2.7984962406015037, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 51525 + }, + { + "epoch": 2.798556390977444, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 51526 + }, + { + "epoch": 2.7986165413533834, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 51527 + }, + { + "epoch": 2.798676691729323, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 51528 + }, + { + "epoch": 2.798736842105263, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 51529 + }, + { + "epoch": 2.7987969924812033, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.2247, + "step": 51530 + }, + { + "epoch": 2.798857142857143, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 51531 + }, + { + "epoch": 2.7989172932330826, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 51532 + }, + { + "epoch": 2.7989774436090227, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 51533 + }, + { + "epoch": 2.7990375939849623, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 51534 + }, + { + "epoch": 2.7990977443609024, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 51535 + }, + { + "epoch": 2.799157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 51536 + }, + { + "epoch": 2.7992180451127817, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 51537 + }, + { + "epoch": 2.799278195488722, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9283, + "step": 51538 + }, + { + "epoch": 2.799338345864662, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 51539 + }, + { + "epoch": 2.7993984962406016, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 51540 + }, + { + "epoch": 2.799458646616541, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 51541 + }, + { + "epoch": 2.7995187969924813, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 51542 + }, + { + "epoch": 2.799578947368421, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 51543 + }, + { + "epoch": 2.799639097744361, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 51544 + }, + { + "epoch": 2.7996992481203007, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 51545 + }, + { + "epoch": 2.7997593984962403, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 51546 + }, + { + "epoch": 2.7998195488721804, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 51547 + }, + { + "epoch": 2.7998796992481205, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 51548 + }, + { + "epoch": 2.79993984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 51549 + }, + { + "epoch": 2.8, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 51550 + }, + { + "epoch": 2.80006015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 51551 + }, + { + "epoch": 2.8001203007518796, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 51552 + }, + { + "epoch": 2.8001804511278197, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 51553 + }, + { + "epoch": 2.8002406015037593, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 51554 + }, + { + "epoch": 2.8003007518796994, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 51555 + }, + { + "epoch": 2.800360902255639, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 51556 + }, + { + "epoch": 2.800421052631579, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 51557 + }, + { + "epoch": 2.800481203007519, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 51558 + }, + { + "epoch": 2.8005413533834584, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 51559 + }, + { + "epoch": 2.8006015037593985, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 51560 + }, + { + "epoch": 2.800661654135338, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 51561 + }, + { + "epoch": 2.8007218045112783, + "grad_norm": 1.7265625, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 51562 + }, + { + "epoch": 2.800781954887218, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 51563 + }, + { + "epoch": 2.800842105263158, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9387, + "step": 51564 + }, + { + "epoch": 2.8009022556390977, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 51565 + }, + { + "epoch": 2.8009624060150378, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 51566 + }, + { + "epoch": 2.8010225563909774, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 51567 + }, + { + "epoch": 2.801082706766917, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 51568 + }, + { + "epoch": 2.801142857142857, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 51569 + }, + { + "epoch": 2.8012030075187972, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9372, + "step": 51570 + }, + { + "epoch": 2.801263157894737, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 51571 + }, + { + "epoch": 2.8013233082706765, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 51572 + }, + { + "epoch": 2.8013834586466166, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 51573 + }, + { + "epoch": 2.8014436090225563, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 51574 + }, + { + "epoch": 2.8015037593984964, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 51575 + }, + { + "epoch": 2.801563909774436, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 51576 + }, + { + "epoch": 2.8016240601503757, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 51577 + }, + { + "epoch": 2.8016842105263158, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 51578 + }, + { + "epoch": 2.801744360902256, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 51579 + }, + { + "epoch": 2.8018045112781955, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 51580 + }, + { + "epoch": 2.801864661654135, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.938, + "step": 51581 + }, + { + "epoch": 2.8019248120300753, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 51582 + }, + { + "epoch": 2.801984962406015, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 51583 + }, + { + "epoch": 2.802045112781955, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9595, + "step": 51584 + }, + { + "epoch": 2.8021052631578947, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 51585 + }, + { + "epoch": 2.8021654135338347, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9331, + "step": 51586 + }, + { + "epoch": 2.8022255639097744, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 51587 + }, + { + "epoch": 2.8022857142857145, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 51588 + }, + { + "epoch": 2.802345864661654, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 51589 + }, + { + "epoch": 2.802406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 51590 + }, + { + "epoch": 2.802466165413534, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 51591 + }, + { + "epoch": 2.8025263157894735, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 51592 + }, + { + "epoch": 2.8025864661654136, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 51593 + }, + { + "epoch": 2.8026466165413533, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.8934, + "step": 51594 + }, + { + "epoch": 2.8027067669172934, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 51595 + }, + { + "epoch": 2.802766917293233, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 51596 + }, + { + "epoch": 2.802827067669173, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 51597 + }, + { + "epoch": 2.8028872180451128, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 51598 + }, + { + "epoch": 2.8029473684210524, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 51599 + }, + { + "epoch": 2.8030075187969925, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 51600 + }, + { + "epoch": 2.8030676691729326, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9346, + "step": 51601 + }, + { + "epoch": 2.8031278195488722, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 51602 + }, + { + "epoch": 2.803187969924812, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 51603 + }, + { + "epoch": 2.803248120300752, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 51604 + }, + { + "epoch": 2.8033082706766916, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 51605 + }, + { + "epoch": 2.8033684210526317, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9152, + "step": 51606 + }, + { + "epoch": 2.8034285714285714, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 51607 + }, + { + "epoch": 2.803488721804511, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 51608 + }, + { + "epoch": 2.803548872180451, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 51609 + }, + { + "epoch": 2.803609022556391, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 51610 + }, + { + "epoch": 2.803669172932331, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 51611 + }, + { + "epoch": 2.8037293233082705, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 51612 + }, + { + "epoch": 2.8037894736842106, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 51613 + }, + { + "epoch": 2.8038496240601503, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 51614 + }, + { + "epoch": 2.8039097744360904, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.8968, + "step": 51615 + }, + { + "epoch": 2.80396992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 51616 + }, + { + "epoch": 2.8040300751879697, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 51617 + }, + { + "epoch": 2.8040902255639097, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 51618 + }, + { + "epoch": 2.80415037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 51619 + }, + { + "epoch": 2.8042105263157895, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 51620 + }, + { + "epoch": 2.804270676691729, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 51621 + }, + { + "epoch": 2.8043308270676692, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 51622 + }, + { + "epoch": 2.804390977443609, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.926, + "step": 51623 + }, + { + "epoch": 2.804451127819549, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 51624 + }, + { + "epoch": 2.8045112781954886, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 51625 + }, + { + "epoch": 2.8045714285714287, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 51626 + }, + { + "epoch": 2.8046315789473684, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 51627 + }, + { + "epoch": 2.8046917293233085, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 51628 + }, + { + "epoch": 2.804751879699248, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9481, + "step": 51629 + }, + { + "epoch": 2.8048120300751878, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 51630 + }, + { + "epoch": 2.804872180451128, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 51631 + }, + { + "epoch": 2.804932330827068, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1196, + "step": 51632 + }, + { + "epoch": 2.8049924812030076, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 51633 + }, + { + "epoch": 2.8050526315789472, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9214, + "step": 51634 + }, + { + "epoch": 2.8051127819548873, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9456, + "step": 51635 + }, + { + "epoch": 2.805172932330827, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 51636 + }, + { + "epoch": 2.805233082706767, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 51637 + }, + { + "epoch": 2.8052932330827067, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 51638 + }, + { + "epoch": 2.8053533834586464, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 51639 + }, + { + "epoch": 2.8054135338345865, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 51640 + }, + { + "epoch": 2.8054736842105266, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 51641 + }, + { + "epoch": 2.805533834586466, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 51642 + }, + { + "epoch": 2.805593984962406, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 51643 + }, + { + "epoch": 2.805654135338346, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 51644 + }, + { + "epoch": 2.8057142857142856, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 51645 + }, + { + "epoch": 2.8057744360902257, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 51646 + }, + { + "epoch": 2.8058345864661653, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 51647 + }, + { + "epoch": 2.805894736842105, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 51648 + }, + { + "epoch": 2.805954887218045, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 51649 + }, + { + "epoch": 2.806015037593985, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 51650 + }, + { + "epoch": 2.806075187969925, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 51651 + }, + { + "epoch": 2.8061353383458645, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 51652 + }, + { + "epoch": 2.8061954887218046, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 51653 + }, + { + "epoch": 2.8062556390977442, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 51654 + }, + { + "epoch": 2.8063157894736843, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 51655 + }, + { + "epoch": 2.806375939849624, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 51656 + }, + { + "epoch": 2.806436090225564, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 51657 + }, + { + "epoch": 2.8064962406015037, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 51658 + }, + { + "epoch": 2.806556390977444, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 51659 + }, + { + "epoch": 2.8066165413533835, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 51660 + }, + { + "epoch": 2.806676691729323, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 51661 + }, + { + "epoch": 2.806736842105263, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 51662 + }, + { + "epoch": 2.806796992481203, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 51663 + }, + { + "epoch": 2.806857142857143, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 51664 + }, + { + "epoch": 2.8069172932330826, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 51665 + }, + { + "epoch": 2.8069774436090227, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 51666 + }, + { + "epoch": 2.8070375939849623, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 51667 + }, + { + "epoch": 2.8070977443609024, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9347, + "step": 51668 + }, + { + "epoch": 2.807157894736842, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 51669 + }, + { + "epoch": 2.8072180451127817, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 51670 + }, + { + "epoch": 2.807278195488722, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 51671 + }, + { + "epoch": 2.807338345864662, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9072, + "step": 51672 + }, + { + "epoch": 2.8073984962406016, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 51673 + }, + { + "epoch": 2.807458646616541, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 51674 + }, + { + "epoch": 2.8075187969924813, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 0.9366, + "step": 51675 + }, + { + "epoch": 2.807578947368421, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 51676 + }, + { + "epoch": 2.807639097744361, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 51677 + }, + { + "epoch": 2.8076992481203007, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9176, + "step": 51678 + }, + { + "epoch": 2.8077593984962403, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.8851, + "step": 51679 + }, + { + "epoch": 2.8078195488721804, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 51680 + }, + { + "epoch": 2.8078796992481205, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 51681 + }, + { + "epoch": 2.80793984962406, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 51682 + }, + { + "epoch": 2.808, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1477, + "step": 51683 + }, + { + "epoch": 2.80806015037594, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 51684 + }, + { + "epoch": 2.8081203007518796, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 51685 + }, + { + "epoch": 2.8081804511278197, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 51686 + }, + { + "epoch": 2.8082406015037593, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 51687 + }, + { + "epoch": 2.8083007518796994, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 51688 + }, + { + "epoch": 2.808360902255639, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 51689 + }, + { + "epoch": 2.808421052631579, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 51690 + }, + { + "epoch": 2.808481203007519, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 51691 + }, + { + "epoch": 2.8085413533834584, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 51692 + }, + { + "epoch": 2.8086015037593985, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 51693 + }, + { + "epoch": 2.808661654135338, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9342, + "step": 51694 + }, + { + "epoch": 2.8087218045112783, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 51695 + }, + { + "epoch": 2.808781954887218, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 51696 + }, + { + "epoch": 2.808842105263158, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9341, + "step": 51697 + }, + { + "epoch": 2.8089022556390977, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 51698 + }, + { + "epoch": 2.8089624060150378, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 51699 + }, + { + "epoch": 2.8090225563909774, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 51700 + }, + { + "epoch": 2.809082706766917, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 51701 + }, + { + "epoch": 2.809142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 51702 + }, + { + "epoch": 2.8092030075187973, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 51703 + }, + { + "epoch": 2.809263157894737, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 51704 + }, + { + "epoch": 2.8093233082706766, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 51705 + }, + { + "epoch": 2.8093834586466166, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 51706 + }, + { + "epoch": 2.8094436090225563, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 51707 + }, + { + "epoch": 2.8095037593984964, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 51708 + }, + { + "epoch": 2.809563909774436, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 51709 + }, + { + "epoch": 2.8096240601503757, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 51710 + }, + { + "epoch": 2.809684210526316, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 51711 + }, + { + "epoch": 2.809744360902256, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 51712 + }, + { + "epoch": 2.8098045112781955, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 51713 + }, + { + "epoch": 2.809864661654135, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 51714 + }, + { + "epoch": 2.8099248120300753, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 51715 + }, + { + "epoch": 2.809984962406015, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 51716 + }, + { + "epoch": 2.810045112781955, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 51717 + }, + { + "epoch": 2.8101052631578947, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 51718 + }, + { + "epoch": 2.8101654135338343, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 51719 + }, + { + "epoch": 2.8102255639097744, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 51720 + }, + { + "epoch": 2.8102857142857145, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 51721 + }, + { + "epoch": 2.810345864661654, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 51722 + }, + { + "epoch": 2.810406015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 51723 + }, + { + "epoch": 2.810466165413534, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 51724 + }, + { + "epoch": 2.8105263157894735, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 51725 + }, + { + "epoch": 2.8105864661654136, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 51726 + }, + { + "epoch": 2.8106466165413533, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 51727 + }, + { + "epoch": 2.8107067669172934, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 51728 + }, + { + "epoch": 2.810766917293233, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 51729 + }, + { + "epoch": 2.810827067669173, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 51730 + }, + { + "epoch": 2.8108872180451128, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 51731 + }, + { + "epoch": 2.8109473684210524, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 51732 + }, + { + "epoch": 2.8110075187969925, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9269, + "step": 51733 + }, + { + "epoch": 2.8110676691729326, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 51734 + }, + { + "epoch": 2.8111278195488723, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 51735 + }, + { + "epoch": 2.811187969924812, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 51736 + }, + { + "epoch": 2.811248120300752, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 51737 + }, + { + "epoch": 2.8113082706766916, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 51738 + }, + { + "epoch": 2.8113684210526317, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 51739 + }, + { + "epoch": 2.8114285714285714, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 51740 + }, + { + "epoch": 2.811488721804511, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9255, + "step": 51741 + }, + { + "epoch": 2.811548872180451, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 51742 + }, + { + "epoch": 2.8116090225563912, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 51743 + }, + { + "epoch": 2.811669172932331, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 51744 + }, + { + "epoch": 2.8117293233082705, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9191, + "step": 51745 + }, + { + "epoch": 2.8117894736842106, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 51746 + }, + { + "epoch": 2.8118496240601503, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 51747 + }, + { + "epoch": 2.8119097744360904, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 51748 + }, + { + "epoch": 2.81196992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 51749 + }, + { + "epoch": 2.8120300751879697, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1243, + "step": 51750 + }, + { + "epoch": 2.8120902255639098, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 51751 + }, + { + "epoch": 2.81215037593985, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 51752 + }, + { + "epoch": 2.8122105263157895, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 51753 + }, + { + "epoch": 2.812270676691729, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9144, + "step": 51754 + }, + { + "epoch": 2.8123308270676692, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 51755 + }, + { + "epoch": 2.812390977443609, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 51756 + }, + { + "epoch": 2.812451127819549, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 51757 + }, + { + "epoch": 2.8125112781954886, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0886, + "step": 51758 + }, + { + "epoch": 2.8125714285714287, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 51759 + }, + { + "epoch": 2.8126315789473684, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9362, + "step": 51760 + }, + { + "epoch": 2.8126917293233085, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 51761 + }, + { + "epoch": 2.812751879699248, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 51762 + }, + { + "epoch": 2.8128120300751878, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9243, + "step": 51763 + }, + { + "epoch": 2.812872180451128, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 51764 + }, + { + "epoch": 2.8129323308270675, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 51765 + }, + { + "epoch": 2.8129924812030076, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 51766 + }, + { + "epoch": 2.8130526315789472, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 51767 + }, + { + "epoch": 2.8131127819548873, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 51768 + }, + { + "epoch": 2.813172932330827, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 51769 + }, + { + "epoch": 2.813233082706767, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 51770 + }, + { + "epoch": 2.8132932330827067, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 51771 + }, + { + "epoch": 2.8133533834586464, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 51772 + }, + { + "epoch": 2.8134135338345865, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 51773 + }, + { + "epoch": 2.8134736842105266, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 51774 + }, + { + "epoch": 2.813533834586466, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 51775 + }, + { + "epoch": 2.813593984962406, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 51776 + }, + { + "epoch": 2.813654135338346, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 51777 + }, + { + "epoch": 2.8137142857142856, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 51778 + }, + { + "epoch": 2.8137744360902257, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9402, + "step": 51779 + }, + { + "epoch": 2.8138345864661654, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 51780 + }, + { + "epoch": 2.813894736842105, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 51781 + }, + { + "epoch": 2.813954887218045, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 51782 + }, + { + "epoch": 2.814015037593985, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 51783 + }, + { + "epoch": 2.814075187969925, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 51784 + }, + { + "epoch": 2.8141353383458645, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 51785 + }, + { + "epoch": 2.8141954887218046, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 51786 + }, + { + "epoch": 2.8142556390977442, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 51787 + }, + { + "epoch": 2.8143157894736843, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 51788 + }, + { + "epoch": 2.814375939849624, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 51789 + }, + { + "epoch": 2.814436090225564, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 51790 + }, + { + "epoch": 2.8144962406015037, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 51791 + }, + { + "epoch": 2.814556390977444, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 51792 + }, + { + "epoch": 2.8146165413533835, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 51793 + }, + { + "epoch": 2.814676691729323, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 51794 + }, + { + "epoch": 2.814736842105263, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 51795 + }, + { + "epoch": 2.814796992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 51796 + }, + { + "epoch": 2.814857142857143, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 51797 + }, + { + "epoch": 2.8149172932330826, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 51798 + }, + { + "epoch": 2.8149774436090227, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 51799 + }, + { + "epoch": 2.8150375939849623, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 51800 + }, + { + "epoch": 2.8150977443609024, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 51801 + }, + { + "epoch": 2.815157894736842, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 51802 + }, + { + "epoch": 2.8152180451127817, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 51803 + }, + { + "epoch": 2.815278195488722, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 51804 + }, + { + "epoch": 2.815338345864662, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 51805 + }, + { + "epoch": 2.8153984962406016, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9444, + "step": 51806 + }, + { + "epoch": 2.815458646616541, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 51807 + }, + { + "epoch": 2.8155187969924813, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 51808 + }, + { + "epoch": 2.815578947368421, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 51809 + }, + { + "epoch": 2.815639097744361, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 51810 + }, + { + "epoch": 2.8156992481203007, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9041, + "step": 51811 + }, + { + "epoch": 2.8157593984962404, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 51812 + }, + { + "epoch": 2.8158195488721804, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 51813 + }, + { + "epoch": 2.8158796992481205, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 51814 + }, + { + "epoch": 2.81593984962406, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 51815 + }, + { + "epoch": 2.816, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 51816 + }, + { + "epoch": 2.81606015037594, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 51817 + }, + { + "epoch": 2.8161203007518796, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 51818 + }, + { + "epoch": 2.8161804511278197, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 51819 + }, + { + "epoch": 2.8162406015037593, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 51820 + }, + { + "epoch": 2.816300751879699, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 51821 + }, + { + "epoch": 2.816360902255639, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 51822 + }, + { + "epoch": 2.816421052631579, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1101, + "step": 51823 + }, + { + "epoch": 2.816481203007519, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.915, + "step": 51824 + }, + { + "epoch": 2.8165413533834585, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 51825 + }, + { + "epoch": 2.8166015037593986, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 51826 + }, + { + "epoch": 2.816661654135338, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1194, + "step": 51827 + }, + { + "epoch": 2.8167218045112783, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 51828 + }, + { + "epoch": 2.816781954887218, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 51829 + }, + { + "epoch": 2.816842105263158, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 51830 + }, + { + "epoch": 2.8169022556390977, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 51831 + }, + { + "epoch": 2.816962406015038, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 51832 + }, + { + "epoch": 2.8170225563909774, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 51833 + }, + { + "epoch": 2.817082706766917, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 51834 + }, + { + "epoch": 2.817142857142857, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1435, + "step": 51835 + }, + { + "epoch": 2.8172030075187973, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 51836 + }, + { + "epoch": 2.817263157894737, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 51837 + }, + { + "epoch": 2.8173233082706766, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 51838 + }, + { + "epoch": 2.8173834586466167, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 51839 + }, + { + "epoch": 2.8174436090225563, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9514, + "step": 51840 + }, + { + "epoch": 2.8175037593984964, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9321, + "step": 51841 + }, + { + "epoch": 2.817563909774436, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 51842 + }, + { + "epoch": 2.8176240601503757, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 51843 + }, + { + "epoch": 2.817684210526316, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 51844 + }, + { + "epoch": 2.817744360902256, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 51845 + }, + { + "epoch": 2.8178045112781955, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 51846 + }, + { + "epoch": 2.817864661654135, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 51847 + }, + { + "epoch": 2.8179248120300753, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 51848 + }, + { + "epoch": 2.817984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 51849 + }, + { + "epoch": 2.818045112781955, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 51850 + }, + { + "epoch": 2.8181052631578947, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 51851 + }, + { + "epoch": 2.8181654135338343, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 51852 + }, + { + "epoch": 2.8182255639097744, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 51853 + }, + { + "epoch": 2.8182857142857145, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 51854 + }, + { + "epoch": 2.818345864661654, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 51855 + }, + { + "epoch": 2.818406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 51856 + }, + { + "epoch": 2.818466165413534, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 51857 + }, + { + "epoch": 2.8185263157894735, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 51858 + }, + { + "epoch": 2.8185864661654136, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 51859 + }, + { + "epoch": 2.8186466165413533, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 51860 + }, + { + "epoch": 2.8187067669172934, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 51861 + }, + { + "epoch": 2.818766917293233, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 51862 + }, + { + "epoch": 2.818827067669173, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 51863 + }, + { + "epoch": 2.8188872180451128, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9143, + "step": 51864 + }, + { + "epoch": 2.8189473684210524, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1114, + "step": 51865 + }, + { + "epoch": 2.8190075187969925, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 51866 + }, + { + "epoch": 2.819067669172932, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 51867 + }, + { + "epoch": 2.8191278195488723, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 51868 + }, + { + "epoch": 2.819187969924812, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 51869 + }, + { + "epoch": 2.819248120300752, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 51870 + }, + { + "epoch": 2.8193082706766917, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 51871 + }, + { + "epoch": 2.8193684210526317, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 51872 + }, + { + "epoch": 2.8194285714285714, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9223, + "step": 51873 + }, + { + "epoch": 2.819488721804511, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 51874 + }, + { + "epoch": 2.819548872180451, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 51875 + }, + { + "epoch": 2.8196090225563912, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 51876 + }, + { + "epoch": 2.819669172932331, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 51877 + }, + { + "epoch": 2.8197293233082705, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.8959, + "step": 51878 + }, + { + "epoch": 2.8197894736842106, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 51879 + }, + { + "epoch": 2.8198496240601503, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 51880 + }, + { + "epoch": 2.8199097744360904, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 51881 + }, + { + "epoch": 2.81996992481203, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 51882 + }, + { + "epoch": 2.8200300751879697, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 51883 + }, + { + "epoch": 2.8200902255639098, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 51884 + }, + { + "epoch": 2.82015037593985, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.91, + "step": 51885 + }, + { + "epoch": 2.8202105263157895, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 51886 + }, + { + "epoch": 2.820270676691729, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 51887 + }, + { + "epoch": 2.8203308270676692, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 51888 + }, + { + "epoch": 2.820390977443609, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 51889 + }, + { + "epoch": 2.820451127819549, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 51890 + }, + { + "epoch": 2.8205112781954886, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 51891 + }, + { + "epoch": 2.8205714285714287, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 51892 + }, + { + "epoch": 2.8206315789473684, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 51893 + }, + { + "epoch": 2.8206917293233085, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 51894 + }, + { + "epoch": 2.820751879699248, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 51895 + }, + { + "epoch": 2.8208120300751878, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1094, + "step": 51896 + }, + { + "epoch": 2.820872180451128, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 51897 + }, + { + "epoch": 2.8209323308270675, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1996, + "step": 51898 + }, + { + "epoch": 2.8209924812030076, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 51899 + }, + { + "epoch": 2.8210526315789473, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 51900 + }, + { + "epoch": 2.8211127819548873, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 51901 + }, + { + "epoch": 2.821172932330827, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 51902 + }, + { + "epoch": 2.821233082706767, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 51903 + }, + { + "epoch": 2.8212932330827067, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0742, + "step": 51904 + }, + { + "epoch": 2.8213533834586464, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2993, + "step": 51905 + }, + { + "epoch": 2.8214135338345865, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 51906 + }, + { + "epoch": 2.8214736842105266, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 51907 + }, + { + "epoch": 2.8215338345864662, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 51908 + }, + { + "epoch": 2.821593984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1312, + "step": 51909 + }, + { + "epoch": 2.821654135338346, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 51910 + }, + { + "epoch": 2.8217142857142856, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 51911 + }, + { + "epoch": 2.8217744360902257, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 51912 + }, + { + "epoch": 2.8218345864661654, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 51913 + }, + { + "epoch": 2.821894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 51914 + }, + { + "epoch": 2.821954887218045, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 51915 + }, + { + "epoch": 2.822015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9349, + "step": 51916 + }, + { + "epoch": 2.822075187969925, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 51917 + }, + { + "epoch": 2.8221353383458645, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 51918 + }, + { + "epoch": 2.8221954887218046, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 51919 + }, + { + "epoch": 2.8222556390977442, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 51920 + }, + { + "epoch": 2.8223157894736843, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9501, + "step": 51921 + }, + { + "epoch": 2.822375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 51922 + }, + { + "epoch": 2.822436090225564, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 51923 + }, + { + "epoch": 2.8224962406015037, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 51924 + }, + { + "epoch": 2.822556390977444, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 51925 + }, + { + "epoch": 2.8226165413533835, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 51926 + }, + { + "epoch": 2.822676691729323, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 51927 + }, + { + "epoch": 2.822736842105263, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 51928 + }, + { + "epoch": 2.822796992481203, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 51929 + }, + { + "epoch": 2.822857142857143, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 51930 + }, + { + "epoch": 2.8229172932330826, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 51931 + }, + { + "epoch": 2.8229774436090227, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 51932 + }, + { + "epoch": 2.8230375939849623, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 51933 + }, + { + "epoch": 2.8230977443609024, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1231, + "step": 51934 + }, + { + "epoch": 2.823157894736842, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0849, + "step": 51935 + }, + { + "epoch": 2.8232180451127817, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 51936 + }, + { + "epoch": 2.823278195488722, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 51937 + }, + { + "epoch": 2.823338345864662, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9322, + "step": 51938 + }, + { + "epoch": 2.8233984962406016, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 51939 + }, + { + "epoch": 2.8234586466165412, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 51940 + }, + { + "epoch": 2.8235187969924813, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 51941 + }, + { + "epoch": 2.823578947368421, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.914, + "step": 51942 + }, + { + "epoch": 2.823639097744361, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 51943 + }, + { + "epoch": 2.8236992481203007, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 51944 + }, + { + "epoch": 2.8237593984962404, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 51945 + }, + { + "epoch": 2.8238195488721805, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 51946 + }, + { + "epoch": 2.8238796992481205, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 51947 + }, + { + "epoch": 2.82393984962406, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 51948 + }, + { + "epoch": 2.824, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.2968, + "step": 51949 + }, + { + "epoch": 2.82406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 51950 + }, + { + "epoch": 2.8241203007518796, + "grad_norm": 1.484375, + "learning_rate": 3e-05, + "loss": 1.1901, + "step": 51951 + }, + { + "epoch": 2.8241804511278197, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 51952 + }, + { + "epoch": 2.8242406015037593, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 51953 + }, + { + "epoch": 2.824300751879699, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 51954 + }, + { + "epoch": 2.824360902255639, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 51955 + }, + { + "epoch": 2.824421052631579, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 51956 + }, + { + "epoch": 2.824481203007519, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 51957 + }, + { + "epoch": 2.8245413533834585, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 51958 + }, + { + "epoch": 2.8246015037593986, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 51959 + }, + { + "epoch": 2.824661654135338, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 51960 + }, + { + "epoch": 2.8247218045112783, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 51961 + }, + { + "epoch": 2.824781954887218, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 51962 + }, + { + "epoch": 2.824842105263158, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 51963 + }, + { + "epoch": 2.8249022556390977, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 51964 + }, + { + "epoch": 2.824962406015038, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 51965 + }, + { + "epoch": 2.8250225563909774, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 51966 + }, + { + "epoch": 2.825082706766917, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 51967 + }, + { + "epoch": 2.825142857142857, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9392, + "step": 51968 + }, + { + "epoch": 2.825203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 51969 + }, + { + "epoch": 2.825263157894737, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 51970 + }, + { + "epoch": 2.8253233082706766, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 51971 + }, + { + "epoch": 2.8253834586466167, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9275, + "step": 51972 + }, + { + "epoch": 2.8254436090225563, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 51973 + }, + { + "epoch": 2.8255037593984964, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1161, + "step": 51974 + }, + { + "epoch": 2.825563909774436, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 51975 + }, + { + "epoch": 2.8256240601503757, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 51976 + }, + { + "epoch": 2.825684210526316, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 51977 + }, + { + "epoch": 2.825744360902256, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 51978 + }, + { + "epoch": 2.8258045112781955, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 51979 + }, + { + "epoch": 2.825864661654135, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 51980 + }, + { + "epoch": 2.8259248120300753, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 51981 + }, + { + "epoch": 2.825984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.933, + "step": 51982 + }, + { + "epoch": 2.826045112781955, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 51983 + }, + { + "epoch": 2.8261052631578947, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 51984 + }, + { + "epoch": 2.8261654135338343, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 51985 + }, + { + "epoch": 2.8262255639097744, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 51986 + }, + { + "epoch": 2.8262857142857145, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 51987 + }, + { + "epoch": 2.826345864661654, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 51988 + }, + { + "epoch": 2.826406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 51989 + }, + { + "epoch": 2.826466165413534, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 51990 + }, + { + "epoch": 2.8265263157894736, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 51991 + }, + { + "epoch": 2.8265864661654136, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 51992 + }, + { + "epoch": 2.8266466165413533, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 51993 + }, + { + "epoch": 2.8267067669172934, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 51994 + }, + { + "epoch": 2.826766917293233, + "grad_norm": 2.03125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 51995 + }, + { + "epoch": 2.826827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 51996 + }, + { + "epoch": 2.826887218045113, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 51997 + }, + { + "epoch": 2.8269473684210524, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 51998 + }, + { + "epoch": 2.8270075187969925, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 51999 + }, + { + "epoch": 2.827067669172932, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 52000 + }, + { + "epoch": 2.8271278195488723, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 52001 + }, + { + "epoch": 2.827187969924812, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 52002 + }, + { + "epoch": 2.827248120300752, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 52003 + }, + { + "epoch": 2.8273082706766917, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 52004 + }, + { + "epoch": 2.8273684210526318, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 52005 + }, + { + "epoch": 2.8274285714285714, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9586, + "step": 52006 + }, + { + "epoch": 2.827488721804511, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 52007 + }, + { + "epoch": 2.827548872180451, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 52008 + }, + { + "epoch": 2.8276090225563912, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 52009 + }, + { + "epoch": 2.827669172932331, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 52010 + }, + { + "epoch": 2.8277293233082705, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 52011 + }, + { + "epoch": 2.8277894736842106, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9283, + "step": 52012 + }, + { + "epoch": 2.8278496240601503, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 52013 + }, + { + "epoch": 2.8279097744360904, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.8813, + "step": 52014 + }, + { + "epoch": 2.82796992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 52015 + }, + { + "epoch": 2.8280300751879697, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 52016 + }, + { + "epoch": 2.8280902255639098, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1757, + "step": 52017 + }, + { + "epoch": 2.82815037593985, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 52018 + }, + { + "epoch": 2.8282105263157895, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 52019 + }, + { + "epoch": 2.828270676691729, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 52020 + }, + { + "epoch": 2.8283308270676693, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 52021 + }, + { + "epoch": 2.828390977443609, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 52022 + }, + { + "epoch": 2.828451127819549, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 52023 + }, + { + "epoch": 2.8285112781954886, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 52024 + }, + { + "epoch": 2.8285714285714287, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 52025 + }, + { + "epoch": 2.8286315789473684, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 52026 + }, + { + "epoch": 2.8286917293233085, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 52027 + }, + { + "epoch": 2.828751879699248, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 52028 + }, + { + "epoch": 2.8288120300751878, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 52029 + }, + { + "epoch": 2.828872180451128, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 52030 + }, + { + "epoch": 2.8289323308270675, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 52031 + }, + { + "epoch": 2.8289924812030076, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 52032 + }, + { + "epoch": 2.8290526315789473, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 52033 + }, + { + "epoch": 2.8291127819548874, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 52034 + }, + { + "epoch": 2.829172932330827, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 52035 + }, + { + "epoch": 2.829233082706767, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 52036 + }, + { + "epoch": 2.8292932330827067, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 52037 + }, + { + "epoch": 2.8293533834586464, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 52038 + }, + { + "epoch": 2.8294135338345865, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 52039 + }, + { + "epoch": 2.8294736842105266, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 52040 + }, + { + "epoch": 2.8295338345864662, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 52041 + }, + { + "epoch": 2.829593984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.8533, + "step": 52042 + }, + { + "epoch": 2.829654135338346, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 52043 + }, + { + "epoch": 2.8297142857142856, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 52044 + }, + { + "epoch": 2.8297744360902257, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 52045 + }, + { + "epoch": 2.8298345864661654, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 52046 + }, + { + "epoch": 2.829894736842105, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 52047 + }, + { + "epoch": 2.829954887218045, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 52048 + }, + { + "epoch": 2.830015037593985, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2132, + "step": 52049 + }, + { + "epoch": 2.830075187969925, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 52050 + }, + { + "epoch": 2.8301353383458645, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.957, + "step": 52051 + }, + { + "epoch": 2.8301954887218046, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9527, + "step": 52052 + }, + { + "epoch": 2.8302556390977442, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 52053 + }, + { + "epoch": 2.8303157894736843, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9586, + "step": 52054 + }, + { + "epoch": 2.830375939849624, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 52055 + }, + { + "epoch": 2.8304360902255636, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 52056 + }, + { + "epoch": 2.8304962406015037, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 52057 + }, + { + "epoch": 2.830556390977444, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 52058 + }, + { + "epoch": 2.8306165413533835, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9317, + "step": 52059 + }, + { + "epoch": 2.830676691729323, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 52060 + }, + { + "epoch": 2.830736842105263, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 52061 + }, + { + "epoch": 2.830796992481203, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.872, + "step": 52062 + }, + { + "epoch": 2.830857142857143, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 52063 + }, + { + "epoch": 2.8309172932330826, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 52064 + }, + { + "epoch": 2.8309774436090227, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 52065 + }, + { + "epoch": 2.8310375939849624, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 52066 + }, + { + "epoch": 2.8310977443609024, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9497, + "step": 52067 + }, + { + "epoch": 2.831157894736842, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 52068 + }, + { + "epoch": 2.8312180451127817, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 52069 + }, + { + "epoch": 2.831278195488722, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 52070 + }, + { + "epoch": 2.8313383458646615, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 52071 + }, + { + "epoch": 2.8313984962406016, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 52072 + }, + { + "epoch": 2.8314586466165412, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 52073 + }, + { + "epoch": 2.8315187969924813, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 52074 + }, + { + "epoch": 2.831578947368421, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 52075 + }, + { + "epoch": 2.831639097744361, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 52076 + }, + { + "epoch": 2.8316992481203007, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 52077 + }, + { + "epoch": 2.8317593984962404, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 52078 + }, + { + "epoch": 2.8318195488721805, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 52079 + }, + { + "epoch": 2.8318796992481206, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 52080 + }, + { + "epoch": 2.83193984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 52081 + }, + { + "epoch": 2.832, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 52082 + }, + { + "epoch": 2.83206015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 52083 + }, + { + "epoch": 2.8321203007518796, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9321, + "step": 52084 + }, + { + "epoch": 2.8321804511278197, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 52085 + }, + { + "epoch": 2.8322406015037593, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 52086 + }, + { + "epoch": 2.832300751879699, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 52087 + }, + { + "epoch": 2.832360902255639, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 52088 + }, + { + "epoch": 2.832421052631579, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 52089 + }, + { + "epoch": 2.832481203007519, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 52090 + }, + { + "epoch": 2.8325413533834585, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 52091 + }, + { + "epoch": 2.8326015037593986, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1606, + "step": 52092 + }, + { + "epoch": 2.832661654135338, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 52093 + }, + { + "epoch": 2.8327218045112783, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 52094 + }, + { + "epoch": 2.832781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 52095 + }, + { + "epoch": 2.832842105263158, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 52096 + }, + { + "epoch": 2.8329022556390977, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 52097 + }, + { + "epoch": 2.832962406015038, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 52098 + }, + { + "epoch": 2.8330225563909774, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 52099 + }, + { + "epoch": 2.833082706766917, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 52100 + }, + { + "epoch": 2.833142857142857, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 52101 + }, + { + "epoch": 2.833203007518797, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 52102 + }, + { + "epoch": 2.833263157894737, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 52103 + }, + { + "epoch": 2.8333233082706766, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 52104 + }, + { + "epoch": 2.8333834586466167, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 52105 + }, + { + "epoch": 2.8334436090225563, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 52106 + }, + { + "epoch": 2.8335037593984964, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9027, + "step": 52107 + }, + { + "epoch": 2.833563909774436, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 52108 + }, + { + "epoch": 2.8336240601503757, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 52109 + }, + { + "epoch": 2.833684210526316, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 52110 + }, + { + "epoch": 2.833744360902256, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 52111 + }, + { + "epoch": 2.8338045112781955, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 52112 + }, + { + "epoch": 2.833864661654135, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 52113 + }, + { + "epoch": 2.8339248120300753, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 52114 + }, + { + "epoch": 2.833984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 52115 + }, + { + "epoch": 2.834045112781955, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 52116 + }, + { + "epoch": 2.8341052631578947, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 52117 + }, + { + "epoch": 2.8341654135338343, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 52118 + }, + { + "epoch": 2.8342255639097744, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 52119 + }, + { + "epoch": 2.8342857142857145, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 52120 + }, + { + "epoch": 2.834345864661654, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 52121 + }, + { + "epoch": 2.834406015037594, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 52122 + }, + { + "epoch": 2.834466165413534, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 52123 + }, + { + "epoch": 2.8345263157894736, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9234, + "step": 52124 + }, + { + "epoch": 2.8345864661654137, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9084, + "step": 52125 + }, + { + "epoch": 2.8346466165413533, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 52126 + }, + { + "epoch": 2.8347067669172934, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 52127 + }, + { + "epoch": 2.834766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 52128 + }, + { + "epoch": 2.834827067669173, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 52129 + }, + { + "epoch": 2.834887218045113, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 52130 + }, + { + "epoch": 2.8349473684210524, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 52131 + }, + { + "epoch": 2.8350075187969925, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 52132 + }, + { + "epoch": 2.835067669172932, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.2429, + "step": 52133 + }, + { + "epoch": 2.8351278195488723, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9113, + "step": 52134 + }, + { + "epoch": 2.835187969924812, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 52135 + }, + { + "epoch": 2.835248120300752, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 52136 + }, + { + "epoch": 2.8353082706766917, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 52137 + }, + { + "epoch": 2.8353684210526318, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 52138 + }, + { + "epoch": 2.8354285714285714, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 52139 + }, + { + "epoch": 2.835488721804511, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 52140 + }, + { + "epoch": 2.835548872180451, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 52141 + }, + { + "epoch": 2.8356090225563912, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 52142 + }, + { + "epoch": 2.835669172932331, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 52143 + }, + { + "epoch": 2.8357293233082705, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9065, + "step": 52144 + }, + { + "epoch": 2.8357894736842106, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 52145 + }, + { + "epoch": 2.8358496240601503, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9306, + "step": 52146 + }, + { + "epoch": 2.8359097744360904, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 52147 + }, + { + "epoch": 2.83596992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 52148 + }, + { + "epoch": 2.8360300751879697, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 52149 + }, + { + "epoch": 2.8360902255639098, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 52150 + }, + { + "epoch": 2.83615037593985, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 52151 + }, + { + "epoch": 2.8362105263157895, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9595, + "step": 52152 + }, + { + "epoch": 2.836270676691729, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 52153 + }, + { + "epoch": 2.8363308270676693, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 52154 + }, + { + "epoch": 2.836390977443609, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 52155 + }, + { + "epoch": 2.836451127819549, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 52156 + }, + { + "epoch": 2.8365112781954887, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 52157 + }, + { + "epoch": 2.8365714285714283, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 52158 + }, + { + "epoch": 2.8366315789473684, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 52159 + }, + { + "epoch": 2.8366917293233085, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 52160 + }, + { + "epoch": 2.836751879699248, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 52161 + }, + { + "epoch": 2.836812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 52162 + }, + { + "epoch": 2.836872180451128, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 52163 + }, + { + "epoch": 2.8369323308270675, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 52164 + }, + { + "epoch": 2.8369924812030076, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 52165 + }, + { + "epoch": 2.8370526315789473, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9097, + "step": 52166 + }, + { + "epoch": 2.8371127819548874, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 52167 + }, + { + "epoch": 2.837172932330827, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 52168 + }, + { + "epoch": 2.837233082706767, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9288, + "step": 52169 + }, + { + "epoch": 2.8372932330827068, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 52170 + }, + { + "epoch": 2.8373533834586464, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 52171 + }, + { + "epoch": 2.8374135338345865, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 52172 + }, + { + "epoch": 2.8374736842105266, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 52173 + }, + { + "epoch": 2.8375338345864662, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 52174 + }, + { + "epoch": 2.837593984962406, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 52175 + }, + { + "epoch": 2.837654135338346, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 52176 + }, + { + "epoch": 2.8377142857142856, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 52177 + }, + { + "epoch": 2.8377744360902257, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1311, + "step": 52178 + }, + { + "epoch": 2.8378345864661654, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1112, + "step": 52179 + }, + { + "epoch": 2.837894736842105, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 52180 + }, + { + "epoch": 2.837954887218045, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 52181 + }, + { + "epoch": 2.838015037593985, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 52182 + }, + { + "epoch": 2.838075187969925, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 52183 + }, + { + "epoch": 2.8381353383458645, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 52184 + }, + { + "epoch": 2.8381954887218046, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 52185 + }, + { + "epoch": 2.8382556390977443, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 52186 + }, + { + "epoch": 2.8383157894736843, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 52187 + }, + { + "epoch": 2.838375939849624, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 52188 + }, + { + "epoch": 2.8384360902255636, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 52189 + }, + { + "epoch": 2.8384962406015037, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9307, + "step": 52190 + }, + { + "epoch": 2.838556390977444, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 52191 + }, + { + "epoch": 2.8386165413533835, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 52192 + }, + { + "epoch": 2.838676691729323, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 52193 + }, + { + "epoch": 2.8387368421052632, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 52194 + }, + { + "epoch": 2.838796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9511, + "step": 52195 + }, + { + "epoch": 2.838857142857143, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9344, + "step": 52196 + }, + { + "epoch": 2.8389172932330826, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 52197 + }, + { + "epoch": 2.8389774436090227, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 52198 + }, + { + "epoch": 2.8390375939849624, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 52199 + }, + { + "epoch": 2.8390977443609025, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 52200 + }, + { + "epoch": 2.839157894736842, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9205, + "step": 52201 + }, + { + "epoch": 2.8392180451127818, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 52202 + }, + { + "epoch": 2.839278195488722, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 52203 + }, + { + "epoch": 2.8393383458646615, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9058, + "step": 52204 + }, + { + "epoch": 2.8393984962406016, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 52205 + }, + { + "epoch": 2.8394586466165412, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 52206 + }, + { + "epoch": 2.8395187969924813, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 52207 + }, + { + "epoch": 2.839578947368421, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 52208 + }, + { + "epoch": 2.839639097744361, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 52209 + }, + { + "epoch": 2.8396992481203007, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 52210 + }, + { + "epoch": 2.8397593984962404, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 52211 + }, + { + "epoch": 2.8398195488721805, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 52212 + }, + { + "epoch": 2.8398796992481206, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 52213 + }, + { + "epoch": 2.83993984962406, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 52214 + }, + { + "epoch": 2.84, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 52215 + }, + { + "epoch": 2.84006015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 52216 + }, + { + "epoch": 2.8401203007518796, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 52217 + }, + { + "epoch": 2.8401804511278197, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 52218 + }, + { + "epoch": 2.8402406015037593, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 52219 + }, + { + "epoch": 2.840300751879699, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 52220 + }, + { + "epoch": 2.840360902255639, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 52221 + }, + { + "epoch": 2.840421052631579, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 52222 + }, + { + "epoch": 2.840481203007519, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 52223 + }, + { + "epoch": 2.8405413533834585, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 52224 + }, + { + "epoch": 2.8406015037593986, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 52225 + }, + { + "epoch": 2.840661654135338, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 52226 + }, + { + "epoch": 2.8407218045112783, + "grad_norm": 1.84375, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 52227 + }, + { + "epoch": 2.840781954887218, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 52228 + }, + { + "epoch": 2.840842105263158, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 52229 + }, + { + "epoch": 2.8409022556390977, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 52230 + }, + { + "epoch": 2.840962406015038, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 52231 + }, + { + "epoch": 2.8410225563909774, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 52232 + }, + { + "epoch": 2.841082706766917, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 52233 + }, + { + "epoch": 2.841142857142857, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 52234 + }, + { + "epoch": 2.841203007518797, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1523, + "step": 52235 + }, + { + "epoch": 2.841263157894737, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 52236 + }, + { + "epoch": 2.8413233082706766, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0858, + "step": 52237 + }, + { + "epoch": 2.8413834586466167, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 52238 + }, + { + "epoch": 2.8414436090225563, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.1661, + "step": 52239 + }, + { + "epoch": 2.8415037593984964, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 52240 + }, + { + "epoch": 2.841563909774436, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 52241 + }, + { + "epoch": 2.8416240601503757, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 52242 + }, + { + "epoch": 2.841684210526316, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9345, + "step": 52243 + }, + { + "epoch": 2.841744360902256, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 52244 + }, + { + "epoch": 2.8418045112781956, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 52245 + }, + { + "epoch": 2.841864661654135, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1604, + "step": 52246 + }, + { + "epoch": 2.8419248120300753, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 52247 + }, + { + "epoch": 2.841984962406015, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 52248 + }, + { + "epoch": 2.842045112781955, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 52249 + }, + { + "epoch": 2.8421052631578947, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 52250 + }, + { + "epoch": 2.8421654135338343, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9219, + "step": 52251 + }, + { + "epoch": 2.8422255639097744, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 52252 + }, + { + "epoch": 2.8422857142857145, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 52253 + }, + { + "epoch": 2.842345864661654, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 52254 + }, + { + "epoch": 2.842406015037594, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 52255 + }, + { + "epoch": 2.842466165413534, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 52256 + }, + { + "epoch": 2.8425263157894736, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 52257 + }, + { + "epoch": 2.8425864661654137, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 52258 + }, + { + "epoch": 2.8426466165413533, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 52259 + }, + { + "epoch": 2.842706766917293, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 52260 + }, + { + "epoch": 2.842766917293233, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 52261 + }, + { + "epoch": 2.842827067669173, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 52262 + }, + { + "epoch": 2.842887218045113, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 52263 + }, + { + "epoch": 2.8429473684210524, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 52264 + }, + { + "epoch": 2.8430075187969925, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 52265 + }, + { + "epoch": 2.843067669172932, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 52266 + }, + { + "epoch": 2.8431278195488723, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 52267 + }, + { + "epoch": 2.843187969924812, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 52268 + }, + { + "epoch": 2.843248120300752, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 52269 + }, + { + "epoch": 2.8433082706766917, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.8877, + "step": 52270 + }, + { + "epoch": 2.8433684210526318, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 52271 + }, + { + "epoch": 2.8434285714285714, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 52272 + }, + { + "epoch": 2.843488721804511, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.939, + "step": 52273 + }, + { + "epoch": 2.843548872180451, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 52274 + }, + { + "epoch": 2.8436090225563913, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 52275 + }, + { + "epoch": 2.843669172932331, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 52276 + }, + { + "epoch": 2.8437293233082706, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 52277 + }, + { + "epoch": 2.8437894736842106, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 52278 + }, + { + "epoch": 2.8438496240601503, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 52279 + }, + { + "epoch": 2.8439097744360904, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 52280 + }, + { + "epoch": 2.84396992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9387, + "step": 52281 + }, + { + "epoch": 2.8440300751879697, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 52282 + }, + { + "epoch": 2.84409022556391, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 52283 + }, + { + "epoch": 2.84415037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.906, + "step": 52284 + }, + { + "epoch": 2.8442105263157895, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9518, + "step": 52285 + }, + { + "epoch": 2.844270676691729, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 52286 + }, + { + "epoch": 2.8443308270676693, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 52287 + }, + { + "epoch": 2.844390977443609, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 52288 + }, + { + "epoch": 2.844451127819549, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 52289 + }, + { + "epoch": 2.8445112781954887, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 52290 + }, + { + "epoch": 2.8445714285714283, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9306, + "step": 52291 + }, + { + "epoch": 2.8446315789473684, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 52292 + }, + { + "epoch": 2.8446917293233085, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 52293 + }, + { + "epoch": 2.844751879699248, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 52294 + }, + { + "epoch": 2.844812030075188, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 52295 + }, + { + "epoch": 2.844872180451128, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 52296 + }, + { + "epoch": 2.8449323308270675, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 52297 + }, + { + "epoch": 2.8449924812030076, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1039, + "step": 52298 + }, + { + "epoch": 2.8450526315789473, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 52299 + }, + { + "epoch": 2.8451127819548874, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 52300 + }, + { + "epoch": 2.845172932330827, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 52301 + }, + { + "epoch": 2.845233082706767, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 52302 + }, + { + "epoch": 2.8452932330827068, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 52303 + }, + { + "epoch": 2.8453533834586464, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 52304 + }, + { + "epoch": 2.8454135338345865, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 52305 + }, + { + "epoch": 2.845473684210526, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 52306 + }, + { + "epoch": 2.8455338345864662, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 52307 + }, + { + "epoch": 2.845593984962406, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.922, + "step": 52308 + }, + { + "epoch": 2.845654135338346, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 52309 + }, + { + "epoch": 2.8457142857142856, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 52310 + }, + { + "epoch": 2.8457744360902257, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 52311 + }, + { + "epoch": 2.8458345864661654, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 52312 + }, + { + "epoch": 2.845894736842105, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.3048, + "step": 52313 + }, + { + "epoch": 2.845954887218045, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9244, + "step": 52314 + }, + { + "epoch": 2.846015037593985, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 52315 + }, + { + "epoch": 2.846075187969925, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 52316 + }, + { + "epoch": 2.8461353383458645, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 52317 + }, + { + "epoch": 2.8461954887218046, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 52318 + }, + { + "epoch": 2.8462556390977443, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 52319 + }, + { + "epoch": 2.8463157894736844, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 52320 + }, + { + "epoch": 2.846375939849624, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 52321 + }, + { + "epoch": 2.8464360902255637, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 52322 + }, + { + "epoch": 2.8464962406015037, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 52323 + }, + { + "epoch": 2.846556390977444, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 52324 + }, + { + "epoch": 2.8466165413533835, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 52325 + }, + { + "epoch": 2.846676691729323, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 52326 + }, + { + "epoch": 2.8467368421052632, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 52327 + }, + { + "epoch": 2.846796992481203, + "grad_norm": 0.486328125, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 52328 + }, + { + "epoch": 2.846857142857143, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.8953, + "step": 52329 + }, + { + "epoch": 2.8469172932330826, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 52330 + }, + { + "epoch": 2.8469774436090227, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 52331 + }, + { + "epoch": 2.8470375939849624, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 52332 + }, + { + "epoch": 2.8470977443609025, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9213, + "step": 52333 + }, + { + "epoch": 2.847157894736842, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 52334 + }, + { + "epoch": 2.8472180451127818, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 52335 + }, + { + "epoch": 2.847278195488722, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 52336 + }, + { + "epoch": 2.8473383458646615, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 52337 + }, + { + "epoch": 2.8473984962406016, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 52338 + }, + { + "epoch": 2.8474586466165412, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 52339 + }, + { + "epoch": 2.8475187969924813, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 52340 + }, + { + "epoch": 2.847578947368421, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 52341 + }, + { + "epoch": 2.847639097744361, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.938, + "step": 52342 + }, + { + "epoch": 2.8476992481203007, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 52343 + }, + { + "epoch": 2.8477593984962404, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9423, + "step": 52344 + }, + { + "epoch": 2.8478195488721805, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9307, + "step": 52345 + }, + { + "epoch": 2.8478796992481206, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 52346 + }, + { + "epoch": 2.84793984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 52347 + }, + { + "epoch": 2.848, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 52348 + }, + { + "epoch": 2.84806015037594, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 52349 + }, + { + "epoch": 2.8481203007518796, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 52350 + }, + { + "epoch": 2.8481804511278197, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9101, + "step": 52351 + }, + { + "epoch": 2.8482406015037594, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 52352 + }, + { + "epoch": 2.848300751879699, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 52353 + }, + { + "epoch": 2.848360902255639, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 52354 + }, + { + "epoch": 2.848421052631579, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 52355 + }, + { + "epoch": 2.848481203007519, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 52356 + }, + { + "epoch": 2.8485413533834585, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 52357 + }, + { + "epoch": 2.8486015037593986, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 52358 + }, + { + "epoch": 2.8486616541353382, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9334, + "step": 52359 + }, + { + "epoch": 2.8487218045112783, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 52360 + }, + { + "epoch": 2.848781954887218, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 52361 + }, + { + "epoch": 2.8488421052631576, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 52362 + }, + { + "epoch": 2.8489022556390977, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 52363 + }, + { + "epoch": 2.848962406015038, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 52364 + }, + { + "epoch": 2.8490225563909775, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 52365 + }, + { + "epoch": 2.849082706766917, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 52366 + }, + { + "epoch": 2.849142857142857, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 52367 + }, + { + "epoch": 2.849203007518797, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 52368 + }, + { + "epoch": 2.849263157894737, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 52369 + }, + { + "epoch": 2.8493233082706766, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.91, + "step": 52370 + }, + { + "epoch": 2.8493834586466167, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.051, + "step": 52371 + }, + { + "epoch": 2.8494436090225563, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 52372 + }, + { + "epoch": 2.8495037593984964, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 52373 + }, + { + "epoch": 2.849563909774436, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 52374 + }, + { + "epoch": 2.8496240601503757, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 52375 + }, + { + "epoch": 2.849684210526316, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 52376 + }, + { + "epoch": 2.849744360902256, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 52377 + }, + { + "epoch": 2.8498045112781956, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 52378 + }, + { + "epoch": 2.849864661654135, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 52379 + }, + { + "epoch": 2.8499248120300753, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 52380 + }, + { + "epoch": 2.849984962406015, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 52381 + }, + { + "epoch": 2.850045112781955, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 52382 + }, + { + "epoch": 2.8501052631578947, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9176, + "step": 52383 + }, + { + "epoch": 2.8501654135338343, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 52384 + }, + { + "epoch": 2.8502255639097744, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 52385 + }, + { + "epoch": 2.8502857142857145, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.8954, + "step": 52386 + }, + { + "epoch": 2.850345864661654, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 52387 + }, + { + "epoch": 2.850406015037594, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 52388 + }, + { + "epoch": 2.850466165413534, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 52389 + }, + { + "epoch": 2.8505263157894736, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 52390 + }, + { + "epoch": 2.8505864661654137, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0946, + "step": 52391 + }, + { + "epoch": 2.8506466165413533, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 52392 + }, + { + "epoch": 2.850706766917293, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 52393 + }, + { + "epoch": 2.850766917293233, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 52394 + }, + { + "epoch": 2.850827067669173, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 52395 + }, + { + "epoch": 2.850887218045113, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 52396 + }, + { + "epoch": 2.8509473684210525, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 52397 + }, + { + "epoch": 2.8510075187969925, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 52398 + }, + { + "epoch": 2.851067669172932, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 52399 + }, + { + "epoch": 2.8511278195488723, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9349, + "step": 52400 + }, + { + "epoch": 2.851187969924812, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 52401 + }, + { + "epoch": 2.851248120300752, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 52402 + }, + { + "epoch": 2.8513082706766917, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 52403 + }, + { + "epoch": 2.8513684210526318, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 52404 + }, + { + "epoch": 2.8514285714285714, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 52405 + }, + { + "epoch": 2.851488721804511, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 52406 + }, + { + "epoch": 2.851548872180451, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 52407 + }, + { + "epoch": 2.851609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 52408 + }, + { + "epoch": 2.851669172932331, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 52409 + }, + { + "epoch": 2.8517293233082706, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 52410 + }, + { + "epoch": 2.8517894736842107, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 52411 + }, + { + "epoch": 2.8518496240601503, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 52412 + }, + { + "epoch": 2.8519097744360904, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9023, + "step": 52413 + }, + { + "epoch": 2.85196992481203, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 52414 + }, + { + "epoch": 2.8520300751879697, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 52415 + }, + { + "epoch": 2.85209022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 52416 + }, + { + "epoch": 2.85215037593985, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 52417 + }, + { + "epoch": 2.8522105263157895, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 52418 + }, + { + "epoch": 2.852270676691729, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 52419 + }, + { + "epoch": 2.8523308270676693, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 52420 + }, + { + "epoch": 2.852390977443609, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 52421 + }, + { + "epoch": 2.852451127819549, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 52422 + }, + { + "epoch": 2.8525112781954887, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 52423 + }, + { + "epoch": 2.8525714285714283, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 52424 + }, + { + "epoch": 2.8526315789473684, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 52425 + }, + { + "epoch": 2.8526917293233085, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 52426 + }, + { + "epoch": 2.852751879699248, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 52427 + }, + { + "epoch": 2.852812030075188, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 52428 + }, + { + "epoch": 2.852872180451128, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9193, + "step": 52429 + }, + { + "epoch": 2.8529323308270675, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 52430 + }, + { + "epoch": 2.8529924812030076, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 52431 + }, + { + "epoch": 2.8530526315789473, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9488, + "step": 52432 + }, + { + "epoch": 2.8531127819548874, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 52433 + }, + { + "epoch": 2.853172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 52434 + }, + { + "epoch": 2.853233082706767, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 52435 + }, + { + "epoch": 2.8532932330827068, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 52436 + }, + { + "epoch": 2.8533533834586464, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 52437 + }, + { + "epoch": 2.8534135338345865, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 52438 + }, + { + "epoch": 2.853473684210526, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 52439 + }, + { + "epoch": 2.8535338345864663, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9348, + "step": 52440 + }, + { + "epoch": 2.853593984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 52441 + }, + { + "epoch": 2.853654135338346, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 52442 + }, + { + "epoch": 2.8537142857142856, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 52443 + }, + { + "epoch": 2.8537744360902257, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 52444 + }, + { + "epoch": 2.8538345864661654, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 52445 + }, + { + "epoch": 2.853894736842105, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 52446 + }, + { + "epoch": 2.853954887218045, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 52447 + }, + { + "epoch": 2.8540150375939852, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 52448 + }, + { + "epoch": 2.854075187969925, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 52449 + }, + { + "epoch": 2.8541353383458645, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 52450 + }, + { + "epoch": 2.8541954887218046, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9111, + "step": 52451 + }, + { + "epoch": 2.8542556390977443, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9537, + "step": 52452 + }, + { + "epoch": 2.8543157894736844, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 52453 + }, + { + "epoch": 2.854375939849624, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 52454 + }, + { + "epoch": 2.8544360902255637, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 52455 + }, + { + "epoch": 2.8544962406015038, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 52456 + }, + { + "epoch": 2.854556390977444, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 52457 + }, + { + "epoch": 2.8546165413533835, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 52458 + }, + { + "epoch": 2.854676691729323, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1508, + "step": 52459 + }, + { + "epoch": 2.8547368421052632, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 52460 + }, + { + "epoch": 2.854796992481203, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 52461 + }, + { + "epoch": 2.854857142857143, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 52462 + }, + { + "epoch": 2.8549172932330826, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 52463 + }, + { + "epoch": 2.8549774436090223, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 52464 + }, + { + "epoch": 2.8550375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0867, + "step": 52465 + }, + { + "epoch": 2.8550977443609025, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 52466 + }, + { + "epoch": 2.855157894736842, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 52467 + }, + { + "epoch": 2.8552180451127818, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 52468 + }, + { + "epoch": 2.855278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 52469 + }, + { + "epoch": 2.8553383458646615, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 52470 + }, + { + "epoch": 2.8553984962406016, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 52471 + }, + { + "epoch": 2.8554586466165413, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 52472 + }, + { + "epoch": 2.8555187969924813, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 52473 + }, + { + "epoch": 2.855578947368421, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 52474 + }, + { + "epoch": 2.855639097744361, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 52475 + }, + { + "epoch": 2.8556992481203007, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 52476 + }, + { + "epoch": 2.8557593984962404, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 52477 + }, + { + "epoch": 2.8558195488721805, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 52478 + }, + { + "epoch": 2.8558796992481206, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 52479 + }, + { + "epoch": 2.8559398496240602, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 52480 + }, + { + "epoch": 2.856, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 52481 + }, + { + "epoch": 2.85606015037594, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 52482 + }, + { + "epoch": 2.8561203007518796, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 52483 + }, + { + "epoch": 2.8561804511278197, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2629, + "step": 52484 + }, + { + "epoch": 2.8562406015037594, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.2292, + "step": 52485 + }, + { + "epoch": 2.856300751879699, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 52486 + }, + { + "epoch": 2.856360902255639, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 52487 + }, + { + "epoch": 2.856421052631579, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 52488 + }, + { + "epoch": 2.856481203007519, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 52489 + }, + { + "epoch": 2.8565413533834585, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 52490 + }, + { + "epoch": 2.8566015037593986, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 52491 + }, + { + "epoch": 2.8566616541353382, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 52492 + }, + { + "epoch": 2.8567218045112783, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 52493 + }, + { + "epoch": 2.856781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.94, + "step": 52494 + }, + { + "epoch": 2.8568421052631576, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 52495 + }, + { + "epoch": 2.8569022556390977, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9285, + "step": 52496 + }, + { + "epoch": 2.856962406015038, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 52497 + }, + { + "epoch": 2.8570225563909775, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 52498 + }, + { + "epoch": 2.857082706766917, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 52499 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 52500 + }, + { + "epoch": 2.857203007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 52501 + }, + { + "epoch": 2.857263157894737, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.136, + "step": 52502 + }, + { + "epoch": 2.8573233082706766, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 52503 + }, + { + "epoch": 2.8573834586466167, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 52504 + }, + { + "epoch": 2.8574436090225563, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 52505 + }, + { + "epoch": 2.8575037593984964, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 52506 + }, + { + "epoch": 2.857563909774436, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9433, + "step": 52507 + }, + { + "epoch": 2.8576240601503757, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9307, + "step": 52508 + }, + { + "epoch": 2.857684210526316, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 52509 + }, + { + "epoch": 2.8577443609022555, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 52510 + }, + { + "epoch": 2.8578045112781956, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9388, + "step": 52511 + }, + { + "epoch": 2.857864661654135, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 52512 + }, + { + "epoch": 2.8579248120300753, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 52513 + }, + { + "epoch": 2.857984962406015, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.935, + "step": 52514 + }, + { + "epoch": 2.858045112781955, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 52515 + }, + { + "epoch": 2.8581052631578947, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 52516 + }, + { + "epoch": 2.8581654135338344, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 52517 + }, + { + "epoch": 2.8582255639097744, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 52518 + }, + { + "epoch": 2.8582857142857145, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 52519 + }, + { + "epoch": 2.858345864661654, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 52520 + }, + { + "epoch": 2.858406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 52521 + }, + { + "epoch": 2.858466165413534, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 52522 + }, + { + "epoch": 2.8585263157894736, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 52523 + }, + { + "epoch": 2.8585864661654137, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 52524 + }, + { + "epoch": 2.8586466165413533, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 52525 + }, + { + "epoch": 2.858706766917293, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 52526 + }, + { + "epoch": 2.858766917293233, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 52527 + }, + { + "epoch": 2.858827067669173, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 52528 + }, + { + "epoch": 2.858887218045113, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 52529 + }, + { + "epoch": 2.8589473684210525, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 52530 + }, + { + "epoch": 2.8590075187969926, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 52531 + }, + { + "epoch": 2.859067669172932, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 52532 + }, + { + "epoch": 2.8591278195488723, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 52533 + }, + { + "epoch": 2.859187969924812, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 52534 + }, + { + "epoch": 2.859248120300752, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 52535 + }, + { + "epoch": 2.8593082706766917, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 52536 + }, + { + "epoch": 2.859368421052632, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9325, + "step": 52537 + }, + { + "epoch": 2.8594285714285714, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 52538 + }, + { + "epoch": 2.859488721804511, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 52539 + }, + { + "epoch": 2.859548872180451, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 52540 + }, + { + "epoch": 2.859609022556391, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 52541 + }, + { + "epoch": 2.859669172932331, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9537, + "step": 52542 + }, + { + "epoch": 2.8597293233082706, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 52543 + }, + { + "epoch": 2.8597894736842107, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 52544 + }, + { + "epoch": 2.8598496240601503, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 52545 + }, + { + "epoch": 2.8599097744360904, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 52546 + }, + { + "epoch": 2.85996992481203, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9202, + "step": 52547 + }, + { + "epoch": 2.8600300751879697, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 52548 + }, + { + "epoch": 2.86009022556391, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 52549 + }, + { + "epoch": 2.86015037593985, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 52550 + }, + { + "epoch": 2.8602105263157895, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 52551 + }, + { + "epoch": 2.860270676691729, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 52552 + }, + { + "epoch": 2.8603308270676693, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 52553 + }, + { + "epoch": 2.860390977443609, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9514, + "step": 52554 + }, + { + "epoch": 2.860451127819549, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 52555 + }, + { + "epoch": 2.8605112781954887, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 52556 + }, + { + "epoch": 2.8605714285714283, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 52557 + }, + { + "epoch": 2.8606315789473684, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 52558 + }, + { + "epoch": 2.8606917293233085, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 52559 + }, + { + "epoch": 2.860751879699248, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 52560 + }, + { + "epoch": 2.860812030075188, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9252, + "step": 52561 + }, + { + "epoch": 2.860872180451128, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 52562 + }, + { + "epoch": 2.8609323308270675, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 52563 + }, + { + "epoch": 2.8609924812030076, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 52564 + }, + { + "epoch": 2.8610526315789473, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 52565 + }, + { + "epoch": 2.861112781954887, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 52566 + }, + { + "epoch": 2.861172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 52567 + }, + { + "epoch": 2.861233082706767, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9576, + "step": 52568 + }, + { + "epoch": 2.8612932330827068, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 52569 + }, + { + "epoch": 2.8613533834586464, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 52570 + }, + { + "epoch": 2.8614135338345865, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.946, + "step": 52571 + }, + { + "epoch": 2.861473684210526, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 52572 + }, + { + "epoch": 2.8615338345864663, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 52573 + }, + { + "epoch": 2.861593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 52574 + }, + { + "epoch": 2.861654135338346, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 52575 + }, + { + "epoch": 2.8617142857142857, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 52576 + }, + { + "epoch": 2.8617744360902257, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 52577 + }, + { + "epoch": 2.8618345864661654, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 52578 + }, + { + "epoch": 2.861894736842105, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 52579 + }, + { + "epoch": 2.861954887218045, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 52580 + }, + { + "epoch": 2.8620150375939852, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 52581 + }, + { + "epoch": 2.862075187969925, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 52582 + }, + { + "epoch": 2.8621353383458645, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9499, + "step": 52583 + }, + { + "epoch": 2.8621954887218046, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 52584 + }, + { + "epoch": 2.8622556390977443, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 52585 + }, + { + "epoch": 2.8623157894736844, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 52586 + }, + { + "epoch": 2.862375939849624, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 52587 + }, + { + "epoch": 2.8624360902255637, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 52588 + }, + { + "epoch": 2.8624962406015038, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 52589 + }, + { + "epoch": 2.862556390977444, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 52590 + }, + { + "epoch": 2.8626165413533835, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 52591 + }, + { + "epoch": 2.862676691729323, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 52592 + }, + { + "epoch": 2.8627368421052632, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 52593 + }, + { + "epoch": 2.862796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 52594 + }, + { + "epoch": 2.862857142857143, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 52595 + }, + { + "epoch": 2.8629172932330826, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 52596 + }, + { + "epoch": 2.8629774436090223, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 52597 + }, + { + "epoch": 2.8630375939849624, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 52598 + }, + { + "epoch": 2.8630977443609025, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 52599 + }, + { + "epoch": 2.863157894736842, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 52600 + }, + { + "epoch": 2.8632180451127818, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9429, + "step": 52601 + }, + { + "epoch": 2.863278195488722, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 52602 + }, + { + "epoch": 2.8633383458646615, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 52603 + }, + { + "epoch": 2.8633984962406016, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 52604 + }, + { + "epoch": 2.8634586466165413, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 52605 + }, + { + "epoch": 2.8635187969924814, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 52606 + }, + { + "epoch": 2.863578947368421, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 52607 + }, + { + "epoch": 2.863639097744361, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 52608 + }, + { + "epoch": 2.8636992481203007, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 52609 + }, + { + "epoch": 2.8637593984962404, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1743, + "step": 52610 + }, + { + "epoch": 2.8638195488721805, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 52611 + }, + { + "epoch": 2.86387969924812, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 52612 + }, + { + "epoch": 2.8639398496240602, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 52613 + }, + { + "epoch": 2.864, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 52614 + }, + { + "epoch": 2.86406015037594, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.2005, + "step": 52615 + }, + { + "epoch": 2.8641203007518796, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 52616 + }, + { + "epoch": 2.8641804511278197, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9065, + "step": 52617 + }, + { + "epoch": 2.8642406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 52618 + }, + { + "epoch": 2.864300751879699, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9324, + "step": 52619 + }, + { + "epoch": 2.864360902255639, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 52620 + }, + { + "epoch": 2.864421052631579, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 52621 + }, + { + "epoch": 2.864481203007519, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 52622 + }, + { + "epoch": 2.8645413533834585, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 52623 + }, + { + "epoch": 2.8646015037593986, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 52624 + }, + { + "epoch": 2.8646616541353382, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 52625 + }, + { + "epoch": 2.8647218045112783, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 52626 + }, + { + "epoch": 2.864781954887218, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 52627 + }, + { + "epoch": 2.8648421052631576, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 52628 + }, + { + "epoch": 2.8649022556390977, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 52629 + }, + { + "epoch": 2.864962406015038, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 52630 + }, + { + "epoch": 2.8650225563909775, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 52631 + }, + { + "epoch": 2.865082706766917, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 52632 + }, + { + "epoch": 2.865142857142857, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 52633 + }, + { + "epoch": 2.865203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 52634 + }, + { + "epoch": 2.865263157894737, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 52635 + }, + { + "epoch": 2.8653233082706766, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 52636 + }, + { + "epoch": 2.8653834586466167, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 52637 + }, + { + "epoch": 2.8654436090225563, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 52638 + }, + { + "epoch": 2.8655037593984964, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 52639 + }, + { + "epoch": 2.865563909774436, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 52640 + }, + { + "epoch": 2.8656240601503757, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 52641 + }, + { + "epoch": 2.865684210526316, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9356, + "step": 52642 + }, + { + "epoch": 2.8657443609022555, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 52643 + }, + { + "epoch": 2.8658045112781956, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 52644 + }, + { + "epoch": 2.8658646616541352, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 52645 + }, + { + "epoch": 2.8659248120300753, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9257, + "step": 52646 + }, + { + "epoch": 2.865984962406015, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 52647 + }, + { + "epoch": 2.866045112781955, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 52648 + }, + { + "epoch": 2.8661052631578947, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 52649 + }, + { + "epoch": 2.8661654135338344, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 52650 + }, + { + "epoch": 2.8662255639097745, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9193, + "step": 52651 + }, + { + "epoch": 2.8662857142857145, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 52652 + }, + { + "epoch": 2.866345864661654, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 52653 + }, + { + "epoch": 2.866406015037594, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 52654 + }, + { + "epoch": 2.866466165413534, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 52655 + }, + { + "epoch": 2.8665263157894736, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 52656 + }, + { + "epoch": 2.8665864661654137, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 52657 + }, + { + "epoch": 2.8666466165413533, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 52658 + }, + { + "epoch": 2.866706766917293, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 52659 + }, + { + "epoch": 2.866766917293233, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 52660 + }, + { + "epoch": 2.866827067669173, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0656, + "step": 52661 + }, + { + "epoch": 2.866887218045113, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9061, + "step": 52662 + }, + { + "epoch": 2.8669473684210525, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 52663 + }, + { + "epoch": 2.8670075187969926, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 52664 + }, + { + "epoch": 2.867067669172932, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 52665 + }, + { + "epoch": 2.8671278195488723, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.938, + "step": 52666 + }, + { + "epoch": 2.867187969924812, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 52667 + }, + { + "epoch": 2.867248120300752, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 52668 + }, + { + "epoch": 2.8673082706766917, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 52669 + }, + { + "epoch": 2.867368421052632, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1052, + "step": 52670 + }, + { + "epoch": 2.8674285714285714, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 52671 + }, + { + "epoch": 2.867488721804511, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 52672 + }, + { + "epoch": 2.867548872180451, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9014, + "step": 52673 + }, + { + "epoch": 2.867609022556391, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 52674 + }, + { + "epoch": 2.867669172932331, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 52675 + }, + { + "epoch": 2.8677293233082706, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 52676 + }, + { + "epoch": 2.8677894736842107, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9192, + "step": 52677 + }, + { + "epoch": 2.8678496240601503, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9495, + "step": 52678 + }, + { + "epoch": 2.8679097744360904, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 52679 + }, + { + "epoch": 2.86796992481203, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9186, + "step": 52680 + }, + { + "epoch": 2.8680300751879697, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9138, + "step": 52681 + }, + { + "epoch": 2.86809022556391, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 52682 + }, + { + "epoch": 2.86815037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 52683 + }, + { + "epoch": 2.8682105263157895, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 52684 + }, + { + "epoch": 2.868270676691729, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 52685 + }, + { + "epoch": 2.8683308270676693, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 52686 + }, + { + "epoch": 2.868390977443609, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 52687 + }, + { + "epoch": 2.868451127819549, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 52688 + }, + { + "epoch": 2.8685112781954887, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 52689 + }, + { + "epoch": 2.8685714285714283, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 52690 + }, + { + "epoch": 2.8686315789473684, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 52691 + }, + { + "epoch": 2.8686917293233085, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 52692 + }, + { + "epoch": 2.868751879699248, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 52693 + }, + { + "epoch": 2.868812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 52694 + }, + { + "epoch": 2.868872180451128, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 52695 + }, + { + "epoch": 2.8689323308270676, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 52696 + }, + { + "epoch": 2.8689924812030076, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 52697 + }, + { + "epoch": 2.8690526315789473, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 52698 + }, + { + "epoch": 2.869112781954887, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 52699 + }, + { + "epoch": 2.869172932330827, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 52700 + }, + { + "epoch": 2.869233082706767, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 52701 + }, + { + "epoch": 2.869293233082707, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 52702 + }, + { + "epoch": 2.8693533834586464, + "grad_norm": 1.4765625, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 52703 + }, + { + "epoch": 2.8694135338345865, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 52704 + }, + { + "epoch": 2.869473684210526, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.946, + "step": 52705 + }, + { + "epoch": 2.8695338345864663, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 52706 + }, + { + "epoch": 2.869593984962406, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 52707 + }, + { + "epoch": 2.869654135338346, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 52708 + }, + { + "epoch": 2.8697142857142857, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 52709 + }, + { + "epoch": 2.8697744360902258, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 52710 + }, + { + "epoch": 2.8698345864661654, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 52711 + }, + { + "epoch": 2.869894736842105, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 52712 + }, + { + "epoch": 2.869954887218045, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 52713 + }, + { + "epoch": 2.870015037593985, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 52714 + }, + { + "epoch": 2.870075187969925, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2528, + "step": 52715 + }, + { + "epoch": 2.8701353383458645, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 52716 + }, + { + "epoch": 2.8701954887218046, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 52717 + }, + { + "epoch": 2.8702556390977443, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 52718 + }, + { + "epoch": 2.8703157894736844, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 52719 + }, + { + "epoch": 2.870375939849624, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 52720 + }, + { + "epoch": 2.8704360902255637, + "grad_norm": 1.4140625, + "learning_rate": 3e-05, + "loss": 0.9489, + "step": 52721 + }, + { + "epoch": 2.8704962406015038, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 52722 + }, + { + "epoch": 2.870556390977444, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 52723 + }, + { + "epoch": 2.8706165413533835, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 52724 + }, + { + "epoch": 2.870676691729323, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 52725 + }, + { + "epoch": 2.8707368421052633, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 52726 + }, + { + "epoch": 2.870796992481203, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 52727 + }, + { + "epoch": 2.870857142857143, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 52728 + }, + { + "epoch": 2.8709172932330826, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 52729 + }, + { + "epoch": 2.8709774436090223, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 52730 + }, + { + "epoch": 2.8710375939849624, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 52731 + }, + { + "epoch": 2.8710977443609025, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 52732 + }, + { + "epoch": 2.871157894736842, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 52733 + }, + { + "epoch": 2.871218045112782, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 52734 + }, + { + "epoch": 2.871278195488722, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 52735 + }, + { + "epoch": 2.8713383458646615, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 52736 + }, + { + "epoch": 2.8713984962406016, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 52737 + }, + { + "epoch": 2.8714586466165413, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 52738 + }, + { + "epoch": 2.8715187969924814, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 52739 + }, + { + "epoch": 2.871578947368421, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 52740 + }, + { + "epoch": 2.871639097744361, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 52741 + }, + { + "epoch": 2.8716992481203008, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 52742 + }, + { + "epoch": 2.8717593984962404, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 52743 + }, + { + "epoch": 2.8718195488721805, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9495, + "step": 52744 + }, + { + "epoch": 2.87187969924812, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2928, + "step": 52745 + }, + { + "epoch": 2.8719398496240602, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 52746 + }, + { + "epoch": 2.872, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 52747 + }, + { + "epoch": 2.87206015037594, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 52748 + }, + { + "epoch": 2.8721203007518796, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 52749 + }, + { + "epoch": 2.8721804511278197, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 52750 + }, + { + "epoch": 2.8722406015037594, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 52751 + }, + { + "epoch": 2.872300751879699, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 52752 + }, + { + "epoch": 2.872360902255639, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 52753 + }, + { + "epoch": 2.872421052631579, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9423, + "step": 52754 + }, + { + "epoch": 2.872481203007519, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 52755 + }, + { + "epoch": 2.8725413533834585, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.161, + "step": 52756 + }, + { + "epoch": 2.8726015037593986, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 52757 + }, + { + "epoch": 2.8726616541353382, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 52758 + }, + { + "epoch": 2.8727218045112783, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9443, + "step": 52759 + }, + { + "epoch": 2.872781954887218, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 52760 + }, + { + "epoch": 2.8728421052631576, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 52761 + }, + { + "epoch": 2.8729022556390977, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9377, + "step": 52762 + }, + { + "epoch": 2.872962406015038, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9413, + "step": 52763 + }, + { + "epoch": 2.8730225563909775, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 52764 + }, + { + "epoch": 2.873082706766917, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 52765 + }, + { + "epoch": 2.873142857142857, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 52766 + }, + { + "epoch": 2.873203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 52767 + }, + { + "epoch": 2.873263157894737, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 52768 + }, + { + "epoch": 2.8733233082706766, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 52769 + }, + { + "epoch": 2.8733834586466167, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 52770 + }, + { + "epoch": 2.8734436090225564, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9123, + "step": 52771 + }, + { + "epoch": 2.8735037593984964, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 52772 + }, + { + "epoch": 2.873563909774436, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 52773 + }, + { + "epoch": 2.8736240601503757, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 52774 + }, + { + "epoch": 2.873684210526316, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 52775 + }, + { + "epoch": 2.8737443609022555, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 52776 + }, + { + "epoch": 2.8738045112781956, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 52777 + }, + { + "epoch": 2.8738646616541352, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 52778 + }, + { + "epoch": 2.8739248120300753, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 52779 + }, + { + "epoch": 2.873984962406015, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 52780 + }, + { + "epoch": 2.874045112781955, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 52781 + }, + { + "epoch": 2.8741052631578947, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 52782 + }, + { + "epoch": 2.8741654135338344, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 52783 + }, + { + "epoch": 2.8742255639097745, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 52784 + }, + { + "epoch": 2.8742857142857146, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0899, + "step": 52785 + }, + { + "epoch": 2.874345864661654, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 52786 + }, + { + "epoch": 2.874406015037594, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 52787 + }, + { + "epoch": 2.874466165413534, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 52788 + }, + { + "epoch": 2.8745263157894736, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 52789 + }, + { + "epoch": 2.8745864661654137, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 52790 + }, + { + "epoch": 2.8746466165413533, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 52791 + }, + { + "epoch": 2.874706766917293, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 52792 + }, + { + "epoch": 2.874766917293233, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 52793 + }, + { + "epoch": 2.874827067669173, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 52794 + }, + { + "epoch": 2.874887218045113, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 52795 + }, + { + "epoch": 2.8749473684210525, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 52796 + }, + { + "epoch": 2.8750075187969926, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 52797 + }, + { + "epoch": 2.875067669172932, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 52798 + }, + { + "epoch": 2.8751278195488723, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 52799 + }, + { + "epoch": 2.875187969924812, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 52800 + }, + { + "epoch": 2.8752481203007516, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 52801 + }, + { + "epoch": 2.8753082706766917, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 52802 + }, + { + "epoch": 2.875368421052632, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 52803 + }, + { + "epoch": 2.8754285714285714, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9332, + "step": 52804 + }, + { + "epoch": 2.875488721804511, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 52805 + }, + { + "epoch": 2.875548872180451, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 52806 + }, + { + "epoch": 2.875609022556391, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 52807 + }, + { + "epoch": 2.875669172932331, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 52808 + }, + { + "epoch": 2.8757293233082706, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 52809 + }, + { + "epoch": 2.8757894736842107, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 52810 + }, + { + "epoch": 2.8758496240601503, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9347, + "step": 52811 + }, + { + "epoch": 2.8759097744360904, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 52812 + }, + { + "epoch": 2.87596992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 52813 + }, + { + "epoch": 2.8760300751879697, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 52814 + }, + { + "epoch": 2.87609022556391, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 52815 + }, + { + "epoch": 2.87615037593985, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 52816 + }, + { + "epoch": 2.8762105263157896, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 52817 + }, + { + "epoch": 2.876270676691729, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 52818 + }, + { + "epoch": 2.8763308270676693, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 52819 + }, + { + "epoch": 2.876390977443609, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 52820 + }, + { + "epoch": 2.876451127819549, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 52821 + }, + { + "epoch": 2.8765112781954887, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 52822 + }, + { + "epoch": 2.8765714285714283, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 52823 + }, + { + "epoch": 2.8766315789473684, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 52824 + }, + { + "epoch": 2.8766917293233085, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 52825 + }, + { + "epoch": 2.876751879699248, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.867, + "step": 52826 + }, + { + "epoch": 2.876812030075188, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 52827 + }, + { + "epoch": 2.876872180451128, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1317, + "step": 52828 + }, + { + "epoch": 2.8769323308270676, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 52829 + }, + { + "epoch": 2.8769924812030077, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 52830 + }, + { + "epoch": 2.8770526315789473, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 52831 + }, + { + "epoch": 2.877112781954887, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 52832 + }, + { + "epoch": 2.877172932330827, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 52833 + }, + { + "epoch": 2.877233082706767, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 52834 + }, + { + "epoch": 2.877293233082707, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 52835 + }, + { + "epoch": 2.8773533834586464, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 52836 + }, + { + "epoch": 2.8774135338345865, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 52837 + }, + { + "epoch": 2.877473684210526, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 52838 + }, + { + "epoch": 2.8775338345864663, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 52839 + }, + { + "epoch": 2.877593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 52840 + }, + { + "epoch": 2.877654135338346, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 52841 + }, + { + "epoch": 2.8777142857142857, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1185, + "step": 52842 + }, + { + "epoch": 2.8777744360902258, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 52843 + }, + { + "epoch": 2.8778345864661654, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9322, + "step": 52844 + }, + { + "epoch": 2.877894736842105, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 52845 + }, + { + "epoch": 2.877954887218045, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9236, + "step": 52846 + }, + { + "epoch": 2.878015037593985, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 52847 + }, + { + "epoch": 2.878075187969925, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 52848 + }, + { + "epoch": 2.8781353383458645, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 52849 + }, + { + "epoch": 2.8781954887218046, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 52850 + }, + { + "epoch": 2.8782556390977443, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 52851 + }, + { + "epoch": 2.8783157894736844, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 52852 + }, + { + "epoch": 2.878375939849624, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 52853 + }, + { + "epoch": 2.8784360902255637, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 52854 + }, + { + "epoch": 2.8784962406015038, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 52855 + }, + { + "epoch": 2.878556390977444, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 52856 + }, + { + "epoch": 2.8786165413533835, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 52857 + }, + { + "epoch": 2.878676691729323, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 52858 + }, + { + "epoch": 2.8787368421052633, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 52859 + }, + { + "epoch": 2.878796992481203, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 52860 + }, + { + "epoch": 2.878857142857143, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 52861 + }, + { + "epoch": 2.8789172932330827, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 52862 + }, + { + "epoch": 2.8789774436090223, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 52863 + }, + { + "epoch": 2.8790375939849624, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 52864 + }, + { + "epoch": 2.8790977443609025, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 52865 + }, + { + "epoch": 2.879157894736842, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 52866 + }, + { + "epoch": 2.879218045112782, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 52867 + }, + { + "epoch": 2.879278195488722, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 52868 + }, + { + "epoch": 2.8793383458646615, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 52869 + }, + { + "epoch": 2.8793984962406016, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.2253, + "step": 52870 + }, + { + "epoch": 2.8794586466165413, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 52871 + }, + { + "epoch": 2.8795187969924814, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 52872 + }, + { + "epoch": 2.879578947368421, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 52873 + }, + { + "epoch": 2.879639097744361, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.117, + "step": 52874 + }, + { + "epoch": 2.8796992481203008, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 52875 + }, + { + "epoch": 2.8797593984962404, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 52876 + }, + { + "epoch": 2.8798195488721805, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 52877 + }, + { + "epoch": 2.87987969924812, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 52878 + }, + { + "epoch": 2.8799398496240602, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 52879 + }, + { + "epoch": 2.88, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 52880 + }, + { + "epoch": 2.88006015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 52881 + }, + { + "epoch": 2.8801203007518796, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 52882 + }, + { + "epoch": 2.8801804511278197, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 52883 + }, + { + "epoch": 2.8802406015037594, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 52884 + }, + { + "epoch": 2.880300751879699, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 52885 + }, + { + "epoch": 2.880360902255639, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 52886 + }, + { + "epoch": 2.880421052631579, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 52887 + }, + { + "epoch": 2.880481203007519, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 52888 + }, + { + "epoch": 2.8805413533834585, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 52889 + }, + { + "epoch": 2.8806015037593986, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 52890 + }, + { + "epoch": 2.8806616541353383, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 52891 + }, + { + "epoch": 2.8807218045112783, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 52892 + }, + { + "epoch": 2.880781954887218, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 52893 + }, + { + "epoch": 2.8808421052631576, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 52894 + }, + { + "epoch": 2.8809022556390977, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 52895 + }, + { + "epoch": 2.880962406015038, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 52896 + }, + { + "epoch": 2.8810225563909775, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 52897 + }, + { + "epoch": 2.881082706766917, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 52898 + }, + { + "epoch": 2.8811428571428572, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 52899 + }, + { + "epoch": 2.881203007518797, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 52900 + }, + { + "epoch": 2.881263157894737, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 52901 + }, + { + "epoch": 2.8813233082706766, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 52902 + }, + { + "epoch": 2.8813834586466163, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 52903 + }, + { + "epoch": 2.8814436090225564, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 52904 + }, + { + "epoch": 2.8815037593984965, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9298, + "step": 52905 + }, + { + "epoch": 2.881563909774436, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9013, + "step": 52906 + }, + { + "epoch": 2.8816240601503758, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 52907 + }, + { + "epoch": 2.881684210526316, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 52908 + }, + { + "epoch": 2.8817443609022555, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 52909 + }, + { + "epoch": 2.8818045112781956, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 52910 + }, + { + "epoch": 2.8818646616541352, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 52911 + }, + { + "epoch": 2.8819248120300753, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 52912 + }, + { + "epoch": 2.881984962406015, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 52913 + }, + { + "epoch": 2.882045112781955, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.8948, + "step": 52914 + }, + { + "epoch": 2.8821052631578947, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 52915 + }, + { + "epoch": 2.8821654135338344, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 52916 + }, + { + "epoch": 2.8822255639097745, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 52917 + }, + { + "epoch": 2.8822857142857146, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 52918 + }, + { + "epoch": 2.882345864661654, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 52919 + }, + { + "epoch": 2.882406015037594, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 52920 + }, + { + "epoch": 2.882466165413534, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 52921 + }, + { + "epoch": 2.8825263157894736, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0811, + "step": 52922 + }, + { + "epoch": 2.8825864661654137, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 52923 + }, + { + "epoch": 2.8826466165413533, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 52924 + }, + { + "epoch": 2.882706766917293, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 52925 + }, + { + "epoch": 2.882766917293233, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 52926 + }, + { + "epoch": 2.882827067669173, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9227, + "step": 52927 + }, + { + "epoch": 2.882887218045113, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 52928 + }, + { + "epoch": 2.8829473684210525, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 52929 + }, + { + "epoch": 2.8830075187969926, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 52930 + }, + { + "epoch": 2.8830676691729322, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9249, + "step": 52931 + }, + { + "epoch": 2.8831278195488723, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 52932 + }, + { + "epoch": 2.883187969924812, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 52933 + }, + { + "epoch": 2.8832481203007516, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 52934 + }, + { + "epoch": 2.8833082706766917, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9063, + "step": 52935 + }, + { + "epoch": 2.883368421052632, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 52936 + }, + { + "epoch": 2.8834285714285715, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 52937 + }, + { + "epoch": 2.883488721804511, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 52938 + }, + { + "epoch": 2.883548872180451, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 52939 + }, + { + "epoch": 2.883609022556391, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 52940 + }, + { + "epoch": 2.883669172932331, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9396, + "step": 52941 + }, + { + "epoch": 2.8837293233082706, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 52942 + }, + { + "epoch": 2.8837894736842107, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1426, + "step": 52943 + }, + { + "epoch": 2.8838496240601503, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 52944 + }, + { + "epoch": 2.8839097744360904, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 52945 + }, + { + "epoch": 2.88396992481203, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 52946 + }, + { + "epoch": 2.8840300751879697, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 52947 + }, + { + "epoch": 2.88409022556391, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 52948 + }, + { + "epoch": 2.8841503759398495, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 52949 + }, + { + "epoch": 2.8842105263157896, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 52950 + }, + { + "epoch": 2.884270676691729, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 52951 + }, + { + "epoch": 2.8843308270676693, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1569, + "step": 52952 + }, + { + "epoch": 2.884390977443609, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9286, + "step": 52953 + }, + { + "epoch": 2.884451127819549, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 52954 + }, + { + "epoch": 2.8845112781954887, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 52955 + }, + { + "epoch": 2.8845714285714283, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 52956 + }, + { + "epoch": 2.8846315789473684, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9313, + "step": 52957 + }, + { + "epoch": 2.8846917293233085, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 52958 + }, + { + "epoch": 2.884751879699248, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 52959 + }, + { + "epoch": 2.884812030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 52960 + }, + { + "epoch": 2.884872180451128, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 52961 + }, + { + "epoch": 2.8849323308270676, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 52962 + }, + { + "epoch": 2.8849924812030077, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 52963 + }, + { + "epoch": 2.8850526315789473, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 52964 + }, + { + "epoch": 2.885112781954887, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 52965 + }, + { + "epoch": 2.885172932330827, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 52966 + }, + { + "epoch": 2.885233082706767, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 52967 + }, + { + "epoch": 2.885293233082707, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 52968 + }, + { + "epoch": 2.8853533834586464, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9371, + "step": 52969 + }, + { + "epoch": 2.8854135338345865, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 52970 + }, + { + "epoch": 2.885473684210526, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 52971 + }, + { + "epoch": 2.8855338345864663, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 52972 + }, + { + "epoch": 2.885593984962406, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 52973 + }, + { + "epoch": 2.885654135338346, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 52974 + }, + { + "epoch": 2.8857142857142857, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 52975 + }, + { + "epoch": 2.8857744360902258, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 52976 + }, + { + "epoch": 2.8858345864661654, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 52977 + }, + { + "epoch": 2.885894736842105, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 52978 + }, + { + "epoch": 2.885954887218045, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 52979 + }, + { + "epoch": 2.886015037593985, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 52980 + }, + { + "epoch": 2.886075187969925, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.941, + "step": 52981 + }, + { + "epoch": 2.8861353383458646, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9251, + "step": 52982 + }, + { + "epoch": 2.8861954887218046, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 52983 + }, + { + "epoch": 2.8862556390977443, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 52984 + }, + { + "epoch": 2.8863157894736844, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1288, + "step": 52985 + }, + { + "epoch": 2.886375939849624, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 52986 + }, + { + "epoch": 2.8864360902255637, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 52987 + }, + { + "epoch": 2.886496240601504, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 52988 + }, + { + "epoch": 2.886556390977444, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 52989 + }, + { + "epoch": 2.8866165413533835, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 52990 + }, + { + "epoch": 2.886676691729323, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1373, + "step": 52991 + }, + { + "epoch": 2.8867368421052633, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 52992 + }, + { + "epoch": 2.886796992481203, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 0.8361, + "step": 52993 + }, + { + "epoch": 2.886857142857143, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 52994 + }, + { + "epoch": 2.8869172932330827, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 52995 + }, + { + "epoch": 2.8869774436090223, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 52996 + }, + { + "epoch": 2.8870375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 52997 + }, + { + "epoch": 2.8870977443609025, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 52998 + }, + { + "epoch": 2.887157894736842, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 52999 + }, + { + "epoch": 2.887218045112782, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 53000 + }, + { + "epoch": 2.887278195488722, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 53001 + }, + { + "epoch": 2.8873383458646615, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9438, + "step": 53002 + }, + { + "epoch": 2.8873984962406016, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 53003 + }, + { + "epoch": 2.8874586466165413, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 53004 + }, + { + "epoch": 2.887518796992481, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 53005 + }, + { + "epoch": 2.887578947368421, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 53006 + }, + { + "epoch": 2.887639097744361, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 53007 + }, + { + "epoch": 2.8876992481203008, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 53008 + }, + { + "epoch": 2.8877593984962404, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1644, + "step": 53009 + }, + { + "epoch": 2.8878195488721805, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 53010 + }, + { + "epoch": 2.88787969924812, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 53011 + }, + { + "epoch": 2.8879398496240603, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 53012 + }, + { + "epoch": 2.888, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.223, + "step": 53013 + }, + { + "epoch": 2.88806015037594, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 53014 + }, + { + "epoch": 2.8881203007518796, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 53015 + }, + { + "epoch": 2.8881804511278197, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 53016 + }, + { + "epoch": 2.8882406015037594, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 53017 + }, + { + "epoch": 2.888300751879699, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 53018 + }, + { + "epoch": 2.888360902255639, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 53019 + }, + { + "epoch": 2.8884210526315792, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 53020 + }, + { + "epoch": 2.888481203007519, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 53021 + }, + { + "epoch": 2.8885413533834585, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9488, + "step": 53022 + }, + { + "epoch": 2.8886015037593986, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 53023 + }, + { + "epoch": 2.8886616541353383, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 53024 + }, + { + "epoch": 2.8887218045112784, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 53025 + }, + { + "epoch": 2.888781954887218, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 53026 + }, + { + "epoch": 2.8888421052631577, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 53027 + }, + { + "epoch": 2.8889022556390977, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9322, + "step": 53028 + }, + { + "epoch": 2.888962406015038, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 53029 + }, + { + "epoch": 2.8890225563909775, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 53030 + }, + { + "epoch": 2.889082706766917, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 53031 + }, + { + "epoch": 2.8891428571428572, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 53032 + }, + { + "epoch": 2.889203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 53033 + }, + { + "epoch": 2.889263157894737, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 53034 + }, + { + "epoch": 2.8893233082706766, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 53035 + }, + { + "epoch": 2.8893834586466163, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 53036 + }, + { + "epoch": 2.8894436090225564, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 53037 + }, + { + "epoch": 2.8895037593984965, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 53038 + }, + { + "epoch": 2.889563909774436, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 53039 + }, + { + "epoch": 2.8896240601503758, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 53040 + }, + { + "epoch": 2.889684210526316, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 53041 + }, + { + "epoch": 2.8897443609022555, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 53042 + }, + { + "epoch": 2.8898045112781956, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 53043 + }, + { + "epoch": 2.8898646616541352, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 53044 + }, + { + "epoch": 2.8899248120300753, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 53045 + }, + { + "epoch": 2.889984962406015, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 53046 + }, + { + "epoch": 2.890045112781955, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 53047 + }, + { + "epoch": 2.8901052631578947, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 53048 + }, + { + "epoch": 2.8901654135338344, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 53049 + }, + { + "epoch": 2.8902255639097745, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 53050 + }, + { + "epoch": 2.890285714285714, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 53051 + }, + { + "epoch": 2.890345864661654, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1583, + "step": 53052 + }, + { + "epoch": 2.890406015037594, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 53053 + }, + { + "epoch": 2.890466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9344, + "step": 53054 + }, + { + "epoch": 2.8905263157894736, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 53055 + }, + { + "epoch": 2.8905864661654137, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9527, + "step": 53056 + }, + { + "epoch": 2.8906466165413534, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1249, + "step": 53057 + }, + { + "epoch": 2.890706766917293, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 53058 + }, + { + "epoch": 2.890766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 53059 + }, + { + "epoch": 2.890827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1177, + "step": 53060 + }, + { + "epoch": 2.890887218045113, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.933, + "step": 53061 + }, + { + "epoch": 2.8909473684210525, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 53062 + }, + { + "epoch": 2.8910075187969926, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.92, + "step": 53063 + }, + { + "epoch": 2.8910676691729322, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 53064 + }, + { + "epoch": 2.8911278195488723, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 53065 + }, + { + "epoch": 2.891187969924812, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 53066 + }, + { + "epoch": 2.8912481203007516, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 53067 + }, + { + "epoch": 2.8913082706766917, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 53068 + }, + { + "epoch": 2.891368421052632, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 53069 + }, + { + "epoch": 2.8914285714285715, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 53070 + }, + { + "epoch": 2.891488721804511, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 53071 + }, + { + "epoch": 2.891548872180451, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 53072 + }, + { + "epoch": 2.891609022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 53073 + }, + { + "epoch": 2.891669172932331, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9396, + "step": 53074 + }, + { + "epoch": 2.8917293233082706, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9479, + "step": 53075 + }, + { + "epoch": 2.8917894736842107, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 53076 + }, + { + "epoch": 2.8918496240601503, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 53077 + }, + { + "epoch": 2.8919097744360904, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 53078 + }, + { + "epoch": 2.89196992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 53079 + }, + { + "epoch": 2.8920300751879697, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 53080 + }, + { + "epoch": 2.89209022556391, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 53081 + }, + { + "epoch": 2.8921503759398495, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9431, + "step": 53082 + }, + { + "epoch": 2.8922105263157896, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 53083 + }, + { + "epoch": 2.892270676691729, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 53084 + }, + { + "epoch": 2.8923308270676693, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9253, + "step": 53085 + }, + { + "epoch": 2.892390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 53086 + }, + { + "epoch": 2.892451127819549, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 53087 + }, + { + "epoch": 2.8925112781954887, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 53088 + }, + { + "epoch": 2.8925714285714283, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 53089 + }, + { + "epoch": 2.8926315789473684, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 53090 + }, + { + "epoch": 2.8926917293233085, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 53091 + }, + { + "epoch": 2.892751879699248, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 53092 + }, + { + "epoch": 2.892812030075188, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 53093 + }, + { + "epoch": 2.892872180451128, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 53094 + }, + { + "epoch": 2.8929323308270676, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 53095 + }, + { + "epoch": 2.8929924812030077, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 53096 + }, + { + "epoch": 2.8930526315789473, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 53097 + }, + { + "epoch": 2.893112781954887, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 53098 + }, + { + "epoch": 2.893172932330827, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 53099 + }, + { + "epoch": 2.893233082706767, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9307, + "step": 53100 + }, + { + "epoch": 2.893293233082707, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 53101 + }, + { + "epoch": 2.8933533834586465, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 53102 + }, + { + "epoch": 2.8934135338345865, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 53103 + }, + { + "epoch": 2.893473684210526, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 53104 + }, + { + "epoch": 2.8935338345864663, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 53105 + }, + { + "epoch": 2.893593984962406, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 53106 + }, + { + "epoch": 2.8936541353383456, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 53107 + }, + { + "epoch": 2.8937142857142857, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 53108 + }, + { + "epoch": 2.8937744360902258, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 53109 + }, + { + "epoch": 2.8938345864661654, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 53110 + }, + { + "epoch": 2.893894736842105, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 53111 + }, + { + "epoch": 2.893954887218045, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 53112 + }, + { + "epoch": 2.894015037593985, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 53113 + }, + { + "epoch": 2.894075187969925, + "grad_norm": 1.4765625, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 53114 + }, + { + "epoch": 2.8941353383458646, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 53115 + }, + { + "epoch": 2.8941954887218047, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 53116 + }, + { + "epoch": 2.8942556390977443, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 53117 + }, + { + "epoch": 2.8943157894736844, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 53118 + }, + { + "epoch": 2.894375939849624, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 53119 + }, + { + "epoch": 2.8944360902255637, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 53120 + }, + { + "epoch": 2.894496240601504, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 53121 + }, + { + "epoch": 2.894556390977444, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 53122 + }, + { + "epoch": 2.8946165413533835, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 53123 + }, + { + "epoch": 2.894676691729323, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 53124 + }, + { + "epoch": 2.8947368421052633, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 53125 + }, + { + "epoch": 2.894796992481203, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 53126 + }, + { + "epoch": 2.894857142857143, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 53127 + }, + { + "epoch": 2.8949172932330827, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9403, + "step": 53128 + }, + { + "epoch": 2.8949774436090223, + "grad_norm": 0.474609375, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 53129 + }, + { + "epoch": 2.8950375939849624, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 53130 + }, + { + "epoch": 2.8950977443609025, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 53131 + }, + { + "epoch": 2.895157894736842, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.155, + "step": 53132 + }, + { + "epoch": 2.895218045112782, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 53133 + }, + { + "epoch": 2.895278195488722, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 53134 + }, + { + "epoch": 2.8953383458646615, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 53135 + }, + { + "epoch": 2.8953984962406016, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 53136 + }, + { + "epoch": 2.8954586466165413, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 53137 + }, + { + "epoch": 2.895518796992481, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 53138 + }, + { + "epoch": 2.895578947368421, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 53139 + }, + { + "epoch": 2.895639097744361, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 53140 + }, + { + "epoch": 2.8956992481203008, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 53141 + }, + { + "epoch": 2.8957593984962404, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 53142 + }, + { + "epoch": 2.8958195488721805, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 53143 + }, + { + "epoch": 2.89587969924812, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 53144 + }, + { + "epoch": 2.8959398496240603, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 53145 + }, + { + "epoch": 2.896, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 53146 + }, + { + "epoch": 2.89606015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 53147 + }, + { + "epoch": 2.8961203007518797, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 53148 + }, + { + "epoch": 2.8961804511278197, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 53149 + }, + { + "epoch": 2.8962406015037594, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 53150 + }, + { + "epoch": 2.896300751879699, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 53151 + }, + { + "epoch": 2.896360902255639, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 53152 + }, + { + "epoch": 2.896421052631579, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 53153 + }, + { + "epoch": 2.896481203007519, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 53154 + }, + { + "epoch": 2.8965413533834585, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 53155 + }, + { + "epoch": 2.8966015037593986, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 53156 + }, + { + "epoch": 2.8966616541353383, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9153, + "step": 53157 + }, + { + "epoch": 2.8967218045112784, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 53158 + }, + { + "epoch": 2.896781954887218, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 53159 + }, + { + "epoch": 2.8968421052631577, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 53160 + }, + { + "epoch": 2.8969022556390978, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 53161 + }, + { + "epoch": 2.896962406015038, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 53162 + }, + { + "epoch": 2.8970225563909775, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.8993, + "step": 53163 + }, + { + "epoch": 2.897082706766917, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 53164 + }, + { + "epoch": 2.8971428571428572, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 53165 + }, + { + "epoch": 2.897203007518797, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 53166 + }, + { + "epoch": 2.897263157894737, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 53167 + }, + { + "epoch": 2.8973233082706766, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 53168 + }, + { + "epoch": 2.8973834586466163, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 53169 + }, + { + "epoch": 2.8974436090225564, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 53170 + }, + { + "epoch": 2.8975037593984965, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 53171 + }, + { + "epoch": 2.897563909774436, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 53172 + }, + { + "epoch": 2.8976240601503758, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 53173 + }, + { + "epoch": 2.897684210526316, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 53174 + }, + { + "epoch": 2.8977443609022555, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9376, + "step": 53175 + }, + { + "epoch": 2.8978045112781956, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 53176 + }, + { + "epoch": 2.8978646616541353, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 53177 + }, + { + "epoch": 2.8979248120300753, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 53178 + }, + { + "epoch": 2.897984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 53179 + }, + { + "epoch": 2.898045112781955, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9482, + "step": 53180 + }, + { + "epoch": 2.8981052631578947, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 53181 + }, + { + "epoch": 2.8981654135338344, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 53182 + }, + { + "epoch": 2.8982255639097745, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 53183 + }, + { + "epoch": 2.898285714285714, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 53184 + }, + { + "epoch": 2.8983458646616542, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 53185 + }, + { + "epoch": 2.898406015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 53186 + }, + { + "epoch": 2.898466165413534, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9647, + "step": 53187 + }, + { + "epoch": 2.8985263157894736, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 53188 + }, + { + "epoch": 2.8985864661654137, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 53189 + }, + { + "epoch": 2.8986466165413534, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 53190 + }, + { + "epoch": 2.898706766917293, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 53191 + }, + { + "epoch": 2.898766917293233, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 53192 + }, + { + "epoch": 2.898827067669173, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 53193 + }, + { + "epoch": 2.898887218045113, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9283, + "step": 53194 + }, + { + "epoch": 2.8989473684210525, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 53195 + }, + { + "epoch": 2.8990075187969926, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 53196 + }, + { + "epoch": 2.8990676691729322, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 53197 + }, + { + "epoch": 2.8991278195488723, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9358, + "step": 53198 + }, + { + "epoch": 2.899187969924812, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 53199 + }, + { + "epoch": 2.8992481203007516, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 53200 + }, + { + "epoch": 2.8993082706766917, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 53201 + }, + { + "epoch": 2.899368421052632, + "grad_norm": 1.546875, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 53202 + }, + { + "epoch": 2.8994285714285715, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 53203 + }, + { + "epoch": 2.899488721804511, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 53204 + }, + { + "epoch": 2.899548872180451, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 53205 + }, + { + "epoch": 2.899609022556391, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9133, + "step": 53206 + }, + { + "epoch": 2.899669172932331, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 53207 + }, + { + "epoch": 2.8997293233082706, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 53208 + }, + { + "epoch": 2.8997894736842103, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 53209 + }, + { + "epoch": 2.8998496240601503, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 53210 + }, + { + "epoch": 2.8999097744360904, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 53211 + }, + { + "epoch": 2.89996992481203, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 53212 + }, + { + "epoch": 2.9000300751879697, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 53213 + }, + { + "epoch": 2.90009022556391, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 53214 + }, + { + "epoch": 2.9001503759398495, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 53215 + }, + { + "epoch": 2.9002105263157896, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 53216 + }, + { + "epoch": 2.900270676691729, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 53217 + }, + { + "epoch": 2.9003308270676693, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 53218 + }, + { + "epoch": 2.900390977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9518, + "step": 53219 + }, + { + "epoch": 2.900451127819549, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 53220 + }, + { + "epoch": 2.9005112781954887, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 53221 + }, + { + "epoch": 2.9005714285714284, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 53222 + }, + { + "epoch": 2.9006315789473684, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9488, + "step": 53223 + }, + { + "epoch": 2.9006917293233085, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 53224 + }, + { + "epoch": 2.900751879699248, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 53225 + }, + { + "epoch": 2.900812030075188, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 53226 + }, + { + "epoch": 2.900872180451128, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 53227 + }, + { + "epoch": 2.9009323308270676, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 53228 + }, + { + "epoch": 2.9009924812030077, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 53229 + }, + { + "epoch": 2.9010526315789473, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9647, + "step": 53230 + }, + { + "epoch": 2.901112781954887, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 53231 + }, + { + "epoch": 2.901172932330827, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 53232 + }, + { + "epoch": 2.901233082706767, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 53233 + }, + { + "epoch": 2.901293233082707, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 53234 + }, + { + "epoch": 2.9013533834586465, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9158, + "step": 53235 + }, + { + "epoch": 2.9014135338345866, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 53236 + }, + { + "epoch": 2.901473684210526, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 53237 + }, + { + "epoch": 2.9015338345864663, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 53238 + }, + { + "epoch": 2.901593984962406, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1059, + "step": 53239 + }, + { + "epoch": 2.9016541353383456, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 53240 + }, + { + "epoch": 2.9017142857142857, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 53241 + }, + { + "epoch": 2.901774436090226, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 53242 + }, + { + "epoch": 2.9018345864661654, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 53243 + }, + { + "epoch": 2.901894736842105, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 53244 + }, + { + "epoch": 2.901954887218045, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 53245 + }, + { + "epoch": 2.902015037593985, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 53246 + }, + { + "epoch": 2.902075187969925, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 53247 + }, + { + "epoch": 2.9021353383458646, + "grad_norm": 1.40625, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 53248 + }, + { + "epoch": 2.9021954887218047, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 53249 + }, + { + "epoch": 2.9022556390977443, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 53250 + }, + { + "epoch": 2.9023157894736844, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 53251 + }, + { + "epoch": 2.902375939849624, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 53252 + }, + { + "epoch": 2.9024360902255637, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 53253 + }, + { + "epoch": 2.902496240601504, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9389, + "step": 53254 + }, + { + "epoch": 2.9025563909774434, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 53255 + }, + { + "epoch": 2.9026165413533835, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 0.8307, + "step": 53256 + }, + { + "epoch": 2.902676691729323, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 53257 + }, + { + "epoch": 2.9027368421052633, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 53258 + }, + { + "epoch": 2.902796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9196, + "step": 53259 + }, + { + "epoch": 2.902857142857143, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 53260 + }, + { + "epoch": 2.9029172932330827, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 53261 + }, + { + "epoch": 2.9029774436090223, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 53262 + }, + { + "epoch": 2.9030375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 53263 + }, + { + "epoch": 2.9030977443609025, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.1406, + "step": 53264 + }, + { + "epoch": 2.903157894736842, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 53265 + }, + { + "epoch": 2.903218045112782, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 53266 + }, + { + "epoch": 2.903278195488722, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 53267 + }, + { + "epoch": 2.9033383458646616, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 53268 + }, + { + "epoch": 2.9033984962406016, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 53269 + }, + { + "epoch": 2.9034586466165413, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.8448, + "step": 53270 + }, + { + "epoch": 2.903518796992481, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 53271 + }, + { + "epoch": 2.903578947368421, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 53272 + }, + { + "epoch": 2.903639097744361, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 53273 + }, + { + "epoch": 2.903699248120301, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 53274 + }, + { + "epoch": 2.9037593984962404, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 53275 + }, + { + "epoch": 2.9038195488721805, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 53276 + }, + { + "epoch": 2.90387969924812, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 53277 + }, + { + "epoch": 2.9039398496240603, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 53278 + }, + { + "epoch": 2.904, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 53279 + }, + { + "epoch": 2.90406015037594, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 53280 + }, + { + "epoch": 2.9041203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 53281 + }, + { + "epoch": 2.9041804511278198, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 53282 + }, + { + "epoch": 2.9042406015037594, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9309, + "step": 53283 + }, + { + "epoch": 2.904300751879699, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 53284 + }, + { + "epoch": 2.904360902255639, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 53285 + }, + { + "epoch": 2.904421052631579, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 53286 + }, + { + "epoch": 2.904481203007519, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 53287 + }, + { + "epoch": 2.9045413533834585, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 53288 + }, + { + "epoch": 2.9046015037593986, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1507, + "step": 53289 + }, + { + "epoch": 2.9046616541353383, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.1171, + "step": 53290 + }, + { + "epoch": 2.9047218045112784, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 53291 + }, + { + "epoch": 2.904781954887218, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 53292 + }, + { + "epoch": 2.9048421052631577, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 53293 + }, + { + "epoch": 2.9049022556390978, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 53294 + }, + { + "epoch": 2.904962406015038, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 53295 + }, + { + "epoch": 2.9050225563909775, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 53296 + }, + { + "epoch": 2.905082706766917, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 53297 + }, + { + "epoch": 2.9051428571428572, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 53298 + }, + { + "epoch": 2.905203007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1118, + "step": 53299 + }, + { + "epoch": 2.905263157894737, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.922, + "step": 53300 + }, + { + "epoch": 2.9053233082706766, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.237, + "step": 53301 + }, + { + "epoch": 2.9053834586466163, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 53302 + }, + { + "epoch": 2.9054436090225564, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 53303 + }, + { + "epoch": 2.9055037593984965, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.8839, + "step": 53304 + }, + { + "epoch": 2.905563909774436, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 53305 + }, + { + "epoch": 2.9056240601503758, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 53306 + }, + { + "epoch": 2.905684210526316, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 53307 + }, + { + "epoch": 2.9057443609022555, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 53308 + }, + { + "epoch": 2.9058045112781956, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 53309 + }, + { + "epoch": 2.9058646616541353, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9586, + "step": 53310 + }, + { + "epoch": 2.905924812030075, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 53311 + }, + { + "epoch": 2.905984962406015, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 53312 + }, + { + "epoch": 2.906045112781955, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 53313 + }, + { + "epoch": 2.9061052631578947, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 53314 + }, + { + "epoch": 2.9061654135338344, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 53315 + }, + { + "epoch": 2.9062255639097745, + "grad_norm": 1.7734375, + "learning_rate": 3e-05, + "loss": 0.8946, + "step": 53316 + }, + { + "epoch": 2.906285714285714, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 53317 + }, + { + "epoch": 2.9063458646616542, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 53318 + }, + { + "epoch": 2.906406015037594, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 53319 + }, + { + "epoch": 2.906466165413534, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 53320 + }, + { + "epoch": 2.9065263157894736, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 53321 + }, + { + "epoch": 2.9065864661654137, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.167, + "step": 53322 + }, + { + "epoch": 2.9066466165413534, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 53323 + }, + { + "epoch": 2.906706766917293, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 53324 + }, + { + "epoch": 2.906766917293233, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 53325 + }, + { + "epoch": 2.906827067669173, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.925, + "step": 53326 + }, + { + "epoch": 2.906887218045113, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 53327 + }, + { + "epoch": 2.9069473684210525, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 53328 + }, + { + "epoch": 2.9070075187969926, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 53329 + }, + { + "epoch": 2.9070676691729322, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 53330 + }, + { + "epoch": 2.9071278195488723, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 53331 + }, + { + "epoch": 2.907187969924812, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 53332 + }, + { + "epoch": 2.9072481203007516, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 53333 + }, + { + "epoch": 2.9073082706766917, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 53334 + }, + { + "epoch": 2.907368421052632, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 53335 + }, + { + "epoch": 2.9074285714285715, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 53336 + }, + { + "epoch": 2.907488721804511, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9472, + "step": 53337 + }, + { + "epoch": 2.907548872180451, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 53338 + }, + { + "epoch": 2.907609022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 53339 + }, + { + "epoch": 2.907669172932331, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 53340 + }, + { + "epoch": 2.9077293233082706, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9216, + "step": 53341 + }, + { + "epoch": 2.9077894736842103, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 53342 + }, + { + "epoch": 2.9078496240601504, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 53343 + }, + { + "epoch": 2.9079097744360904, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 53344 + }, + { + "epoch": 2.90796992481203, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 53345 + }, + { + "epoch": 2.9080300751879697, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 53346 + }, + { + "epoch": 2.90809022556391, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 53347 + }, + { + "epoch": 2.9081503759398495, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 53348 + }, + { + "epoch": 2.9082105263157896, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 53349 + }, + { + "epoch": 2.9082706766917292, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9299, + "step": 53350 + }, + { + "epoch": 2.9083308270676693, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9132, + "step": 53351 + }, + { + "epoch": 2.908390977443609, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 53352 + }, + { + "epoch": 2.908451127819549, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 53353 + }, + { + "epoch": 2.9085112781954887, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.8978, + "step": 53354 + }, + { + "epoch": 2.9085714285714284, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 53355 + }, + { + "epoch": 2.9086315789473685, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 53356 + }, + { + "epoch": 2.908691729323308, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9328, + "step": 53357 + }, + { + "epoch": 2.908751879699248, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 53358 + }, + { + "epoch": 2.908812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 53359 + }, + { + "epoch": 2.908872180451128, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 53360 + }, + { + "epoch": 2.9089323308270676, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.905, + "step": 53361 + }, + { + "epoch": 2.9089924812030077, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 53362 + }, + { + "epoch": 2.9090526315789473, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 53363 + }, + { + "epoch": 2.909112781954887, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 53364 + }, + { + "epoch": 2.909172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 53365 + }, + { + "epoch": 2.909233082706767, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 53366 + }, + { + "epoch": 2.909293233082707, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 53367 + }, + { + "epoch": 2.9093533834586465, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 53368 + }, + { + "epoch": 2.9094135338345866, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 53369 + }, + { + "epoch": 2.909473684210526, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 53370 + }, + { + "epoch": 2.9095338345864663, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 53371 + }, + { + "epoch": 2.909593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 53372 + }, + { + "epoch": 2.9096541353383456, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 53373 + }, + { + "epoch": 2.9097142857142857, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 53374 + }, + { + "epoch": 2.909774436090226, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 53375 + }, + { + "epoch": 2.9098345864661654, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 53376 + }, + { + "epoch": 2.909894736842105, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 53377 + }, + { + "epoch": 2.909954887218045, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 53378 + }, + { + "epoch": 2.910015037593985, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 53379 + }, + { + "epoch": 2.910075187969925, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 53380 + }, + { + "epoch": 2.9101353383458646, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9017, + "step": 53381 + }, + { + "epoch": 2.9101954887218047, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 53382 + }, + { + "epoch": 2.9102556390977443, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 53383 + }, + { + "epoch": 2.9103157894736844, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9203, + "step": 53384 + }, + { + "epoch": 2.910375939849624, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 53385 + }, + { + "epoch": 2.9104360902255637, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 53386 + }, + { + "epoch": 2.910496240601504, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 53387 + }, + { + "epoch": 2.9105563909774435, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 53388 + }, + { + "epoch": 2.9106165413533835, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 53389 + }, + { + "epoch": 2.910676691729323, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 53390 + }, + { + "epoch": 2.9107368421052633, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0939, + "step": 53391 + }, + { + "epoch": 2.910796992481203, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 53392 + }, + { + "epoch": 2.910857142857143, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 53393 + }, + { + "epoch": 2.9109172932330827, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 53394 + }, + { + "epoch": 2.9109774436090223, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 53395 + }, + { + "epoch": 2.9110375939849624, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 53396 + }, + { + "epoch": 2.9110977443609025, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 53397 + }, + { + "epoch": 2.911157894736842, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 53398 + }, + { + "epoch": 2.911218045112782, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 53399 + }, + { + "epoch": 2.911278195488722, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9037, + "step": 53400 + }, + { + "epoch": 2.9113383458646616, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 53401 + }, + { + "epoch": 2.9113984962406017, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 53402 + }, + { + "epoch": 2.9114586466165413, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 53403 + }, + { + "epoch": 2.911518796992481, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 53404 + }, + { + "epoch": 2.911578947368421, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 53405 + }, + { + "epoch": 2.911639097744361, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1073, + "step": 53406 + }, + { + "epoch": 2.911699248120301, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 53407 + }, + { + "epoch": 2.9117593984962404, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 53408 + }, + { + "epoch": 2.9118195488721805, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 53409 + }, + { + "epoch": 2.91187969924812, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 53410 + }, + { + "epoch": 2.9119398496240603, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 53411 + }, + { + "epoch": 2.912, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 53412 + }, + { + "epoch": 2.91206015037594, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 53413 + }, + { + "epoch": 2.9121203007518797, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1238, + "step": 53414 + }, + { + "epoch": 2.9121804511278198, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 53415 + }, + { + "epoch": 2.9122406015037594, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 53416 + }, + { + "epoch": 2.912300751879699, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.1081, + "step": 53417 + }, + { + "epoch": 2.912360902255639, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 53418 + }, + { + "epoch": 2.912421052631579, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 53419 + }, + { + "epoch": 2.912481203007519, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 53420 + }, + { + "epoch": 2.9125413533834585, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 53421 + }, + { + "epoch": 2.9126015037593986, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 53422 + }, + { + "epoch": 2.9126616541353383, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 53423 + }, + { + "epoch": 2.9127218045112784, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 53424 + }, + { + "epoch": 2.912781954887218, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 53425 + }, + { + "epoch": 2.9128421052631577, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9082, + "step": 53426 + }, + { + "epoch": 2.9129022556390978, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 53427 + }, + { + "epoch": 2.912962406015038, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9199, + "step": 53428 + }, + { + "epoch": 2.9130225563909775, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 53429 + }, + { + "epoch": 2.913082706766917, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 53430 + }, + { + "epoch": 2.9131428571428573, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 53431 + }, + { + "epoch": 2.913203007518797, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 53432 + }, + { + "epoch": 2.913263157894737, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 53433 + }, + { + "epoch": 2.9133233082706766, + "grad_norm": 1.421875, + "learning_rate": 3e-05, + "loss": 1.147, + "step": 53434 + }, + { + "epoch": 2.9133834586466163, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 53435 + }, + { + "epoch": 2.9134436090225564, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 53436 + }, + { + "epoch": 2.9135037593984965, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 53437 + }, + { + "epoch": 2.913563909774436, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 53438 + }, + { + "epoch": 2.913624060150376, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 53439 + }, + { + "epoch": 2.913684210526316, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 53440 + }, + { + "epoch": 2.9137443609022555, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9112, + "step": 53441 + }, + { + "epoch": 2.9138045112781956, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 53442 + }, + { + "epoch": 2.9138646616541353, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 53443 + }, + { + "epoch": 2.913924812030075, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 53444 + }, + { + "epoch": 2.913984962406015, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 53445 + }, + { + "epoch": 2.914045112781955, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 53446 + }, + { + "epoch": 2.9141052631578948, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 53447 + }, + { + "epoch": 2.9141654135338344, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 53448 + }, + { + "epoch": 2.9142255639097745, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 53449 + }, + { + "epoch": 2.914285714285714, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 53450 + }, + { + "epoch": 2.9143458646616542, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 53451 + }, + { + "epoch": 2.914406015037594, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 53452 + }, + { + "epoch": 2.914466165413534, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 53453 + }, + { + "epoch": 2.9145263157894736, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 53454 + }, + { + "epoch": 2.9145864661654137, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 53455 + }, + { + "epoch": 2.9146466165413534, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9347, + "step": 53456 + }, + { + "epoch": 2.914706766917293, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 53457 + }, + { + "epoch": 2.914766917293233, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 53458 + }, + { + "epoch": 2.9148270676691728, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 53459 + }, + { + "epoch": 2.914887218045113, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 53460 + }, + { + "epoch": 2.9149473684210525, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9148, + "step": 53461 + }, + { + "epoch": 2.9150075187969926, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 53462 + }, + { + "epoch": 2.9150676691729323, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 53463 + }, + { + "epoch": 2.9151278195488723, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 53464 + }, + { + "epoch": 2.915187969924812, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 53465 + }, + { + "epoch": 2.9152481203007516, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 53466 + }, + { + "epoch": 2.9153082706766917, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 53467 + }, + { + "epoch": 2.915368421052632, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 53468 + }, + { + "epoch": 2.9154285714285715, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 53469 + }, + { + "epoch": 2.915488721804511, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 53470 + }, + { + "epoch": 2.9155488721804512, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 53471 + }, + { + "epoch": 2.915609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 53472 + }, + { + "epoch": 2.915669172932331, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 53473 + }, + { + "epoch": 2.9157293233082706, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9312, + "step": 53474 + }, + { + "epoch": 2.9157894736842103, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 53475 + }, + { + "epoch": 2.9158496240601504, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 53476 + }, + { + "epoch": 2.9159097744360905, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 53477 + }, + { + "epoch": 2.91596992481203, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 53478 + }, + { + "epoch": 2.9160300751879698, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 53479 + }, + { + "epoch": 2.91609022556391, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 53480 + }, + { + "epoch": 2.9161503759398495, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 53481 + }, + { + "epoch": 2.9162105263157896, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 53482 + }, + { + "epoch": 2.9162706766917292, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9235, + "step": 53483 + }, + { + "epoch": 2.9163308270676693, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 53484 + }, + { + "epoch": 2.916390977443609, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.8757, + "step": 53485 + }, + { + "epoch": 2.916451127819549, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 53486 + }, + { + "epoch": 2.9165112781954887, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 53487 + }, + { + "epoch": 2.9165714285714284, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 53488 + }, + { + "epoch": 2.9166315789473685, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 53489 + }, + { + "epoch": 2.916691729323308, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 53490 + }, + { + "epoch": 2.916751879699248, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 53491 + }, + { + "epoch": 2.916812030075188, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 53492 + }, + { + "epoch": 2.916872180451128, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 53493 + }, + { + "epoch": 2.9169323308270676, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 53494 + }, + { + "epoch": 2.9169924812030077, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 53495 + }, + { + "epoch": 2.9170526315789473, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 53496 + }, + { + "epoch": 2.917112781954887, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 53497 + }, + { + "epoch": 2.917172932330827, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 53498 + }, + { + "epoch": 2.917233082706767, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 53499 + }, + { + "epoch": 2.917293233082707, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 53500 + }, + { + "epoch": 2.9173533834586465, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 53501 + }, + { + "epoch": 2.9174135338345866, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 53502 + }, + { + "epoch": 2.917473684210526, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 53503 + }, + { + "epoch": 2.9175338345864663, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 53504 + }, + { + "epoch": 2.917593984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 53505 + }, + { + "epoch": 2.9176541353383456, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 53506 + }, + { + "epoch": 2.9177142857142857, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 53507 + }, + { + "epoch": 2.917774436090226, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 53508 + }, + { + "epoch": 2.9178345864661654, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 53509 + }, + { + "epoch": 2.917894736842105, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9529, + "step": 53510 + }, + { + "epoch": 2.917954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 53511 + }, + { + "epoch": 2.918015037593985, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 53512 + }, + { + "epoch": 2.918075187969925, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.2023, + "step": 53513 + }, + { + "epoch": 2.9181353383458646, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1038, + "step": 53514 + }, + { + "epoch": 2.9181954887218047, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9635, + "step": 53515 + }, + { + "epoch": 2.9182556390977443, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 53516 + }, + { + "epoch": 2.9183157894736844, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 53517 + }, + { + "epoch": 2.918375939849624, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 53518 + }, + { + "epoch": 2.9184360902255637, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 53519 + }, + { + "epoch": 2.918496240601504, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 53520 + }, + { + "epoch": 2.9185563909774435, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 53521 + }, + { + "epoch": 2.9186165413533836, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9375, + "step": 53522 + }, + { + "epoch": 2.918676691729323, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9377, + "step": 53523 + }, + { + "epoch": 2.9187368421052633, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 53524 + }, + { + "epoch": 2.918796992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9293, + "step": 53525 + }, + { + "epoch": 2.918857142857143, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 53526 + }, + { + "epoch": 2.9189172932330827, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 53527 + }, + { + "epoch": 2.9189774436090223, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 53528 + }, + { + "epoch": 2.9190375939849624, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.936, + "step": 53529 + }, + { + "epoch": 2.9190977443609025, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 53530 + }, + { + "epoch": 2.919157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 53531 + }, + { + "epoch": 2.919218045112782, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 53532 + }, + { + "epoch": 2.919278195488722, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 53533 + }, + { + "epoch": 2.9193383458646616, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.8465, + "step": 53534 + }, + { + "epoch": 2.9193984962406017, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.912, + "step": 53535 + }, + { + "epoch": 2.9194586466165413, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 53536 + }, + { + "epoch": 2.919518796992481, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2145, + "step": 53537 + }, + { + "epoch": 2.919578947368421, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 53538 + }, + { + "epoch": 2.919639097744361, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1578, + "step": 53539 + }, + { + "epoch": 2.919699248120301, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 53540 + }, + { + "epoch": 2.9197593984962404, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 53541 + }, + { + "epoch": 2.9198195488721805, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 53542 + }, + { + "epoch": 2.91987969924812, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 53543 + }, + { + "epoch": 2.9199398496240603, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 53544 + }, + { + "epoch": 2.92, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 53545 + }, + { + "epoch": 2.9200601503759396, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 53546 + }, + { + "epoch": 2.9201203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 53547 + }, + { + "epoch": 2.9201804511278198, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 53548 + }, + { + "epoch": 2.9202406015037594, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 53549 + }, + { + "epoch": 2.920300751879699, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 53550 + }, + { + "epoch": 2.920360902255639, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9499, + "step": 53551 + }, + { + "epoch": 2.920421052631579, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 53552 + }, + { + "epoch": 2.920481203007519, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 53553 + }, + { + "epoch": 2.9205413533834585, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1394, + "step": 53554 + }, + { + "epoch": 2.9206015037593986, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 53555 + }, + { + "epoch": 2.9206616541353383, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 0.932, + "step": 53556 + }, + { + "epoch": 2.9207218045112784, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 53557 + }, + { + "epoch": 2.920781954887218, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 53558 + }, + { + "epoch": 2.9208421052631577, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 53559 + }, + { + "epoch": 2.9209022556390978, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 53560 + }, + { + "epoch": 2.920962406015038, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1192, + "step": 53561 + }, + { + "epoch": 2.9210225563909775, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 53562 + }, + { + "epoch": 2.921082706766917, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 53563 + }, + { + "epoch": 2.9211428571428573, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 53564 + }, + { + "epoch": 2.921203007518797, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 53565 + }, + { + "epoch": 2.921263157894737, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 53566 + }, + { + "epoch": 2.9213233082706767, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 53567 + }, + { + "epoch": 2.9213834586466163, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 53568 + }, + { + "epoch": 2.9214436090225564, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 53569 + }, + { + "epoch": 2.9215037593984965, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 53570 + }, + { + "epoch": 2.921563909774436, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 53571 + }, + { + "epoch": 2.921624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 53572 + }, + { + "epoch": 2.921684210526316, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 53573 + }, + { + "epoch": 2.9217443609022555, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.8947, + "step": 53574 + }, + { + "epoch": 2.9218045112781956, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 53575 + }, + { + "epoch": 2.9218646616541353, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 53576 + }, + { + "epoch": 2.921924812030075, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 53577 + }, + { + "epoch": 2.921984962406015, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9421, + "step": 53578 + }, + { + "epoch": 2.922045112781955, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 53579 + }, + { + "epoch": 2.9221052631578948, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 53580 + }, + { + "epoch": 2.9221654135338344, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 53581 + }, + { + "epoch": 2.9222255639097745, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 53582 + }, + { + "epoch": 2.922285714285714, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 53583 + }, + { + "epoch": 2.9223458646616542, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 53584 + }, + { + "epoch": 2.922406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 53585 + }, + { + "epoch": 2.922466165413534, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 53586 + }, + { + "epoch": 2.9225263157894736, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 53587 + }, + { + "epoch": 2.9225864661654137, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0831, + "step": 53588 + }, + { + "epoch": 2.9226466165413534, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 53589 + }, + { + "epoch": 2.922706766917293, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 53590 + }, + { + "epoch": 2.922766917293233, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9191, + "step": 53591 + }, + { + "epoch": 2.9228270676691728, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 53592 + }, + { + "epoch": 2.922887218045113, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 53593 + }, + { + "epoch": 2.9229473684210525, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 53594 + }, + { + "epoch": 2.9230075187969926, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 53595 + }, + { + "epoch": 2.9230676691729323, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 53596 + }, + { + "epoch": 2.9231278195488724, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 53597 + }, + { + "epoch": 2.923187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 53598 + }, + { + "epoch": 2.9232481203007517, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 53599 + }, + { + "epoch": 2.9233082706766917, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 53600 + }, + { + "epoch": 2.923368421052632, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 53601 + }, + { + "epoch": 2.9234285714285715, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 53602 + }, + { + "epoch": 2.923488721804511, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 53603 + }, + { + "epoch": 2.9235488721804512, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 53604 + }, + { + "epoch": 2.923609022556391, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 53605 + }, + { + "epoch": 2.923669172932331, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 53606 + }, + { + "epoch": 2.9237293233082706, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 53607 + }, + { + "epoch": 2.9237894736842103, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 53608 + }, + { + "epoch": 2.9238496240601504, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1989, + "step": 53609 + }, + { + "epoch": 2.9239097744360905, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 53610 + }, + { + "epoch": 2.92396992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 53611 + }, + { + "epoch": 2.9240300751879698, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 53612 + }, + { + "epoch": 2.92409022556391, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 53613 + }, + { + "epoch": 2.9241503759398495, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 53614 + }, + { + "epoch": 2.9242105263157896, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 53615 + }, + { + "epoch": 2.9242706766917292, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 53616 + }, + { + "epoch": 2.9243308270676693, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 53617 + }, + { + "epoch": 2.924390977443609, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 53618 + }, + { + "epoch": 2.924451127819549, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 53619 + }, + { + "epoch": 2.9245112781954887, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 53620 + }, + { + "epoch": 2.9245714285714284, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 53621 + }, + { + "epoch": 2.9246315789473685, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 53622 + }, + { + "epoch": 2.924691729323308, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 53623 + }, + { + "epoch": 2.924751879699248, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 53624 + }, + { + "epoch": 2.924812030075188, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 53625 + }, + { + "epoch": 2.924872180451128, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 53626 + }, + { + "epoch": 2.9249323308270676, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 53627 + }, + { + "epoch": 2.9249924812030077, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 53628 + }, + { + "epoch": 2.9250526315789473, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 53629 + }, + { + "epoch": 2.925112781954887, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9459, + "step": 53630 + }, + { + "epoch": 2.925172932330827, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 53631 + }, + { + "epoch": 2.925233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 53632 + }, + { + "epoch": 2.925293233082707, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 53633 + }, + { + "epoch": 2.9253533834586465, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 53634 + }, + { + "epoch": 2.9254135338345866, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9293, + "step": 53635 + }, + { + "epoch": 2.9254736842105262, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 53636 + }, + { + "epoch": 2.9255338345864663, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 53637 + }, + { + "epoch": 2.925593984962406, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 53638 + }, + { + "epoch": 2.9256541353383456, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 53639 + }, + { + "epoch": 2.9257142857142857, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 53640 + }, + { + "epoch": 2.925774436090226, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 53641 + }, + { + "epoch": 2.9258345864661655, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 53642 + }, + { + "epoch": 2.925894736842105, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 53643 + }, + { + "epoch": 2.925954887218045, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9527, + "step": 53644 + }, + { + "epoch": 2.926015037593985, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 53645 + }, + { + "epoch": 2.926075187969925, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 53646 + }, + { + "epoch": 2.9261353383458646, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.905, + "step": 53647 + }, + { + "epoch": 2.9261954887218042, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9037, + "step": 53648 + }, + { + "epoch": 2.9262556390977443, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9348, + "step": 53649 + }, + { + "epoch": 2.9263157894736844, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 53650 + }, + { + "epoch": 2.926375939849624, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 53651 + }, + { + "epoch": 2.9264360902255637, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 53652 + }, + { + "epoch": 2.926496240601504, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9288, + "step": 53653 + }, + { + "epoch": 2.9265563909774435, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 53654 + }, + { + "epoch": 2.9266165413533836, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 53655 + }, + { + "epoch": 2.926676691729323, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 53656 + }, + { + "epoch": 2.9267368421052633, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 53657 + }, + { + "epoch": 2.926796992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 53658 + }, + { + "epoch": 2.926857142857143, + "grad_norm": 1.453125, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 53659 + }, + { + "epoch": 2.9269172932330827, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 53660 + }, + { + "epoch": 2.9269774436090223, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 53661 + }, + { + "epoch": 2.9270375939849624, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 53662 + }, + { + "epoch": 2.9270977443609025, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 53663 + }, + { + "epoch": 2.927157894736842, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 53664 + }, + { + "epoch": 2.927218045112782, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9356, + "step": 53665 + }, + { + "epoch": 2.927278195488722, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 53666 + }, + { + "epoch": 2.9273383458646616, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 53667 + }, + { + "epoch": 2.9273984962406017, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 53668 + }, + { + "epoch": 2.9274586466165413, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 53669 + }, + { + "epoch": 2.927518796992481, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 53670 + }, + { + "epoch": 2.927578947368421, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 53671 + }, + { + "epoch": 2.927639097744361, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 53672 + }, + { + "epoch": 2.927699248120301, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 53673 + }, + { + "epoch": 2.9277593984962405, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 53674 + }, + { + "epoch": 2.9278195488721805, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.2599, + "step": 53675 + }, + { + "epoch": 2.92787969924812, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 53676 + }, + { + "epoch": 2.9279398496240603, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 53677 + }, + { + "epoch": 2.928, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 53678 + }, + { + "epoch": 2.9280601503759396, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 53679 + }, + { + "epoch": 2.9281203007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 53680 + }, + { + "epoch": 2.9281804511278198, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 53681 + }, + { + "epoch": 2.9282406015037594, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 53682 + }, + { + "epoch": 2.928300751879699, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 53683 + }, + { + "epoch": 2.928360902255639, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.2255, + "step": 53684 + }, + { + "epoch": 2.928421052631579, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 53685 + }, + { + "epoch": 2.928481203007519, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 53686 + }, + { + "epoch": 2.9285413533834586, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 53687 + }, + { + "epoch": 2.9286015037593986, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 53688 + }, + { + "epoch": 2.9286616541353383, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 53689 + }, + { + "epoch": 2.9287218045112784, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 53690 + }, + { + "epoch": 2.928781954887218, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 53691 + }, + { + "epoch": 2.9288421052631577, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 53692 + }, + { + "epoch": 2.928902255639098, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 53693 + }, + { + "epoch": 2.9289624060150374, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 53694 + }, + { + "epoch": 2.9290225563909775, + "grad_norm": 1.6328125, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 53695 + }, + { + "epoch": 2.929082706766917, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 53696 + }, + { + "epoch": 2.9291428571428573, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 53697 + }, + { + "epoch": 2.929203007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 53698 + }, + { + "epoch": 2.929263157894737, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 53699 + }, + { + "epoch": 2.9293233082706767, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 53700 + }, + { + "epoch": 2.9293834586466163, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 53701 + }, + { + "epoch": 2.9294436090225564, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 53702 + }, + { + "epoch": 2.9295037593984965, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 53703 + }, + { + "epoch": 2.929563909774436, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 53704 + }, + { + "epoch": 2.929624060150376, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1774, + "step": 53705 + }, + { + "epoch": 2.929684210526316, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9202, + "step": 53706 + }, + { + "epoch": 2.9297443609022555, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 53707 + }, + { + "epoch": 2.9298045112781956, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 53708 + }, + { + "epoch": 2.9298646616541353, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 53709 + }, + { + "epoch": 2.929924812030075, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 53710 + }, + { + "epoch": 2.929984962406015, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 53711 + }, + { + "epoch": 2.930045112781955, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 53712 + }, + { + "epoch": 2.9301052631578948, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 53713 + }, + { + "epoch": 2.9301654135338344, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 53714 + }, + { + "epoch": 2.9302255639097745, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 53715 + }, + { + "epoch": 2.930285714285714, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 53716 + }, + { + "epoch": 2.9303458646616543, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 53717 + }, + { + "epoch": 2.930406015037594, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 53718 + }, + { + "epoch": 2.930466165413534, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 53719 + }, + { + "epoch": 2.9305263157894736, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 53720 + }, + { + "epoch": 2.9305864661654137, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 53721 + }, + { + "epoch": 2.9306466165413534, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 53722 + }, + { + "epoch": 2.930706766917293, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 53723 + }, + { + "epoch": 2.930766917293233, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9159, + "step": 53724 + }, + { + "epoch": 2.930827067669173, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 53725 + }, + { + "epoch": 2.930887218045113, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 53726 + }, + { + "epoch": 2.9309473684210525, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 53727 + }, + { + "epoch": 2.9310075187969926, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 53728 + }, + { + "epoch": 2.9310676691729323, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 53729 + }, + { + "epoch": 2.9311278195488724, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 53730 + }, + { + "epoch": 2.931187969924812, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9537, + "step": 53731 + }, + { + "epoch": 2.9312481203007517, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 53732 + }, + { + "epoch": 2.9313082706766918, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 53733 + }, + { + "epoch": 2.931368421052632, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1129, + "step": 53734 + }, + { + "epoch": 2.9314285714285715, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 53735 + }, + { + "epoch": 2.931488721804511, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 53736 + }, + { + "epoch": 2.9315488721804512, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 53737 + }, + { + "epoch": 2.931609022556391, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 53738 + }, + { + "epoch": 2.931669172932331, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 53739 + }, + { + "epoch": 2.9317293233082706, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 53740 + }, + { + "epoch": 2.9317894736842103, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 53741 + }, + { + "epoch": 2.9318496240601504, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 53742 + }, + { + "epoch": 2.9319097744360905, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 53743 + }, + { + "epoch": 2.93196992481203, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 53744 + }, + { + "epoch": 2.9320300751879698, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 53745 + }, + { + "epoch": 2.93209022556391, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 53746 + }, + { + "epoch": 2.9321503759398495, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 53747 + }, + { + "epoch": 2.9322105263157896, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9297, + "step": 53748 + }, + { + "epoch": 2.9322706766917292, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 53749 + }, + { + "epoch": 2.932330827067669, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 53750 + }, + { + "epoch": 2.932390977443609, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 53751 + }, + { + "epoch": 2.932451127819549, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 53752 + }, + { + "epoch": 2.9325112781954887, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 53753 + }, + { + "epoch": 2.9325714285714284, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 53754 + }, + { + "epoch": 2.9326315789473685, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9303, + "step": 53755 + }, + { + "epoch": 2.932691729323308, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 53756 + }, + { + "epoch": 2.932751879699248, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 53757 + }, + { + "epoch": 2.932812030075188, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 53758 + }, + { + "epoch": 2.932872180451128, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 53759 + }, + { + "epoch": 2.9329323308270676, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 53760 + }, + { + "epoch": 2.9329924812030077, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9143, + "step": 53761 + }, + { + "epoch": 2.9330526315789474, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 53762 + }, + { + "epoch": 2.933112781954887, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 53763 + }, + { + "epoch": 2.933172932330827, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 53764 + }, + { + "epoch": 2.933233082706767, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 53765 + }, + { + "epoch": 2.933293233082707, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 53766 + }, + { + "epoch": 2.9333533834586465, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9357, + "step": 53767 + }, + { + "epoch": 2.9334135338345866, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 53768 + }, + { + "epoch": 2.9334736842105262, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 53769 + }, + { + "epoch": 2.9335338345864663, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 53770 + }, + { + "epoch": 2.933593984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 53771 + }, + { + "epoch": 2.9336541353383456, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 53772 + }, + { + "epoch": 2.9337142857142857, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 53773 + }, + { + "epoch": 2.933774436090226, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 53774 + }, + { + "epoch": 2.9338345864661655, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 53775 + }, + { + "epoch": 2.933894736842105, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 53776 + }, + { + "epoch": 2.933954887218045, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 53777 + }, + { + "epoch": 2.934015037593985, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 53778 + }, + { + "epoch": 2.934075187969925, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 53779 + }, + { + "epoch": 2.9341353383458646, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 53780 + }, + { + "epoch": 2.9341954887218042, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 53781 + }, + { + "epoch": 2.9342556390977443, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 53782 + }, + { + "epoch": 2.9343157894736844, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9278, + "step": 53783 + }, + { + "epoch": 2.934375939849624, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 53784 + }, + { + "epoch": 2.9344360902255637, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 53785 + }, + { + "epoch": 2.934496240601504, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 53786 + }, + { + "epoch": 2.9345563909774435, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 53787 + }, + { + "epoch": 2.9346165413533836, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 53788 + }, + { + "epoch": 2.934676691729323, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 53789 + }, + { + "epoch": 2.9347368421052633, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 53790 + }, + { + "epoch": 2.934796992481203, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 53791 + }, + { + "epoch": 2.934857142857143, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 53792 + }, + { + "epoch": 2.9349172932330827, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 53793 + }, + { + "epoch": 2.9349774436090224, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 53794 + }, + { + "epoch": 2.9350375939849624, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 53795 + }, + { + "epoch": 2.935097744360902, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 53796 + }, + { + "epoch": 2.935157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 53797 + }, + { + "epoch": 2.935218045112782, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 53798 + }, + { + "epoch": 2.935278195488722, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 53799 + }, + { + "epoch": 2.9353383458646616, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 53800 + }, + { + "epoch": 2.9353984962406017, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 53801 + }, + { + "epoch": 2.9354586466165413, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 53802 + }, + { + "epoch": 2.935518796992481, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 53803 + }, + { + "epoch": 2.935578947368421, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 53804 + }, + { + "epoch": 2.935639097744361, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 53805 + }, + { + "epoch": 2.935699248120301, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9433, + "step": 53806 + }, + { + "epoch": 2.9357593984962405, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 53807 + }, + { + "epoch": 2.9358195488721806, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 53808 + }, + { + "epoch": 2.93587969924812, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 1.1659, + "step": 53809 + }, + { + "epoch": 2.9359398496240603, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 53810 + }, + { + "epoch": 2.936, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 53811 + }, + { + "epoch": 2.9360601503759396, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9093, + "step": 53812 + }, + { + "epoch": 2.9361203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 53813 + }, + { + "epoch": 2.93618045112782, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 53814 + }, + { + "epoch": 2.9362406015037594, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 53815 + }, + { + "epoch": 2.936300751879699, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 53816 + }, + { + "epoch": 2.936360902255639, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 53817 + }, + { + "epoch": 2.936421052631579, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 53818 + }, + { + "epoch": 2.936481203007519, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 53819 + }, + { + "epoch": 2.9365413533834586, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 53820 + }, + { + "epoch": 2.9366015037593987, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 53821 + }, + { + "epoch": 2.9366616541353383, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 53822 + }, + { + "epoch": 2.9367218045112784, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 53823 + }, + { + "epoch": 2.936781954887218, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 53824 + }, + { + "epoch": 2.9368421052631577, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 53825 + }, + { + "epoch": 2.936902255639098, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 53826 + }, + { + "epoch": 2.9369624060150374, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 53827 + }, + { + "epoch": 2.9370225563909775, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 53828 + }, + { + "epoch": 2.937082706766917, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 53829 + }, + { + "epoch": 2.9371428571428573, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 53830 + }, + { + "epoch": 2.937203007518797, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 53831 + }, + { + "epoch": 2.937263157894737, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 53832 + }, + { + "epoch": 2.9373233082706767, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 53833 + }, + { + "epoch": 2.9373834586466163, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 53834 + }, + { + "epoch": 2.9374436090225564, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 53835 + }, + { + "epoch": 2.9375037593984965, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 53836 + }, + { + "epoch": 2.937563909774436, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9487, + "step": 53837 + }, + { + "epoch": 2.937624060150376, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 53838 + }, + { + "epoch": 2.937684210526316, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 53839 + }, + { + "epoch": 2.9377443609022555, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 53840 + }, + { + "epoch": 2.9378045112781956, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.116, + "step": 53841 + }, + { + "epoch": 2.9378646616541353, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 53842 + }, + { + "epoch": 2.937924812030075, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 53843 + }, + { + "epoch": 2.937984962406015, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 53844 + }, + { + "epoch": 2.938045112781955, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 53845 + }, + { + "epoch": 2.9381052631578948, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 53846 + }, + { + "epoch": 2.9381654135338344, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 53847 + }, + { + "epoch": 2.9382255639097745, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 53848 + }, + { + "epoch": 2.938285714285714, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 53849 + }, + { + "epoch": 2.9383458646616543, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 53850 + }, + { + "epoch": 2.938406015037594, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 53851 + }, + { + "epoch": 2.9384661654135336, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 53852 + }, + { + "epoch": 2.9385263157894737, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 53853 + }, + { + "epoch": 2.9385864661654137, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 53854 + }, + { + "epoch": 2.9386466165413534, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 53855 + }, + { + "epoch": 2.938706766917293, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 53856 + }, + { + "epoch": 2.938766917293233, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 53857 + }, + { + "epoch": 2.938827067669173, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9256, + "step": 53858 + }, + { + "epoch": 2.938887218045113, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 53859 + }, + { + "epoch": 2.9389473684210525, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 53860 + }, + { + "epoch": 2.9390075187969926, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 53861 + }, + { + "epoch": 2.9390676691729323, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 53862 + }, + { + "epoch": 2.9391278195488724, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 53863 + }, + { + "epoch": 2.939187969924812, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 53864 + }, + { + "epoch": 2.9392481203007517, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 53865 + }, + { + "epoch": 2.9393082706766918, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 53866 + }, + { + "epoch": 2.939368421052632, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 53867 + }, + { + "epoch": 2.9394285714285715, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 53868 + }, + { + "epoch": 2.939488721804511, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 53869 + }, + { + "epoch": 2.9395488721804512, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 53870 + }, + { + "epoch": 2.939609022556391, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 53871 + }, + { + "epoch": 2.939669172932331, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 53872 + }, + { + "epoch": 2.9397293233082706, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 53873 + }, + { + "epoch": 2.9397894736842103, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 53874 + }, + { + "epoch": 2.9398496240601504, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 53875 + }, + { + "epoch": 2.9399097744360905, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 53876 + }, + { + "epoch": 2.93996992481203, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 53877 + }, + { + "epoch": 2.9400300751879698, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 53878 + }, + { + "epoch": 2.94009022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 53879 + }, + { + "epoch": 2.9401503759398495, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 53880 + }, + { + "epoch": 2.9402105263157896, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 53881 + }, + { + "epoch": 2.9402706766917293, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 53882 + }, + { + "epoch": 2.940330827067669, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 53883 + }, + { + "epoch": 2.940390977443609, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9331, + "step": 53884 + }, + { + "epoch": 2.940451127819549, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 53885 + }, + { + "epoch": 2.9405112781954887, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 53886 + }, + { + "epoch": 2.9405714285714284, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 53887 + }, + { + "epoch": 2.9406315789473685, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 53888 + }, + { + "epoch": 2.940691729323308, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 53889 + }, + { + "epoch": 2.9407518796992482, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 53890 + }, + { + "epoch": 2.940812030075188, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1121, + "step": 53891 + }, + { + "epoch": 2.940872180451128, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 53892 + }, + { + "epoch": 2.9409323308270676, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 53893 + }, + { + "epoch": 2.9409924812030077, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 53894 + }, + { + "epoch": 2.9410526315789474, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 53895 + }, + { + "epoch": 2.941112781954887, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 53896 + }, + { + "epoch": 2.941172932330827, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 53897 + }, + { + "epoch": 2.9412330827067668, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 53898 + }, + { + "epoch": 2.941293233082707, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 53899 + }, + { + "epoch": 2.9413533834586465, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 53900 + }, + { + "epoch": 2.9414135338345866, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 53901 + }, + { + "epoch": 2.9414736842105262, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 53902 + }, + { + "epoch": 2.9415338345864663, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 53903 + }, + { + "epoch": 2.941593984962406, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 53904 + }, + { + "epoch": 2.9416541353383456, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 53905 + }, + { + "epoch": 2.9417142857142857, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 53906 + }, + { + "epoch": 2.941774436090226, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 53907 + }, + { + "epoch": 2.9418345864661655, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 53908 + }, + { + "epoch": 2.941894736842105, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.2231, + "step": 53909 + }, + { + "epoch": 2.941954887218045, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 53910 + }, + { + "epoch": 2.942015037593985, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 53911 + }, + { + "epoch": 2.942075187969925, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 53912 + }, + { + "epoch": 2.9421353383458646, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 53913 + }, + { + "epoch": 2.9421954887218043, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 53914 + }, + { + "epoch": 2.9422556390977443, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9314, + "step": 53915 + }, + { + "epoch": 2.9423157894736844, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 53916 + }, + { + "epoch": 2.942375939849624, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 53917 + }, + { + "epoch": 2.9424360902255637, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 53918 + }, + { + "epoch": 2.942496240601504, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 53919 + }, + { + "epoch": 2.9425563909774435, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 53920 + }, + { + "epoch": 2.9426165413533836, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 53921 + }, + { + "epoch": 2.9426766917293232, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 53922 + }, + { + "epoch": 2.9427368421052633, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0958, + "step": 53923 + }, + { + "epoch": 2.942796992481203, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 53924 + }, + { + "epoch": 2.942857142857143, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 53925 + }, + { + "epoch": 2.9429172932330827, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.929, + "step": 53926 + }, + { + "epoch": 2.9429774436090224, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 53927 + }, + { + "epoch": 2.9430375939849625, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 53928 + }, + { + "epoch": 2.943097744360902, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 53929 + }, + { + "epoch": 2.943157894736842, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 53930 + }, + { + "epoch": 2.943218045112782, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 53931 + }, + { + "epoch": 2.943278195488722, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 53932 + }, + { + "epoch": 2.9433383458646616, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 53933 + }, + { + "epoch": 2.9433984962406017, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 53934 + }, + { + "epoch": 2.9434586466165413, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 53935 + }, + { + "epoch": 2.943518796992481, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 53936 + }, + { + "epoch": 2.943578947368421, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 53937 + }, + { + "epoch": 2.943639097744361, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 53938 + }, + { + "epoch": 2.943699248120301, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 53939 + }, + { + "epoch": 2.9437593984962405, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9133, + "step": 53940 + }, + { + "epoch": 2.9438195488721806, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9175, + "step": 53941 + }, + { + "epoch": 2.94387969924812, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.917, + "step": 53942 + }, + { + "epoch": 2.9439398496240603, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9031, + "step": 53943 + }, + { + "epoch": 2.944, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 53944 + }, + { + "epoch": 2.9440601503759396, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 53945 + }, + { + "epoch": 2.9441203007518797, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 53946 + }, + { + "epoch": 2.94418045112782, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 53947 + }, + { + "epoch": 2.9442406015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 53948 + }, + { + "epoch": 2.944300751879699, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 53949 + }, + { + "epoch": 2.944360902255639, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 53950 + }, + { + "epoch": 2.944421052631579, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 53951 + }, + { + "epoch": 2.944481203007519, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 53952 + }, + { + "epoch": 2.9445413533834586, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 53953 + }, + { + "epoch": 2.944601503759398, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 53954 + }, + { + "epoch": 2.9446616541353383, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 53955 + }, + { + "epoch": 2.9447218045112784, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 53956 + }, + { + "epoch": 2.944781954887218, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 53957 + }, + { + "epoch": 2.9448421052631577, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 53958 + }, + { + "epoch": 2.944902255639098, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 53959 + }, + { + "epoch": 2.9449624060150374, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 53960 + }, + { + "epoch": 2.9450225563909775, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 53961 + }, + { + "epoch": 2.945082706766917, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 53962 + }, + { + "epoch": 2.9451428571428573, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 53963 + }, + { + "epoch": 2.945203007518797, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 53964 + }, + { + "epoch": 2.945263157894737, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 53965 + }, + { + "epoch": 2.9453233082706767, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 53966 + }, + { + "epoch": 2.9453834586466163, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 53967 + }, + { + "epoch": 2.9454436090225564, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 53968 + }, + { + "epoch": 2.9455037593984965, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 53969 + }, + { + "epoch": 2.945563909774436, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 53970 + }, + { + "epoch": 2.945624060150376, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 53971 + }, + { + "epoch": 2.945684210526316, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1218, + "step": 53972 + }, + { + "epoch": 2.9457443609022556, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9234, + "step": 53973 + }, + { + "epoch": 2.9458045112781956, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 53974 + }, + { + "epoch": 2.9458646616541353, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 53975 + }, + { + "epoch": 2.945924812030075, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 53976 + }, + { + "epoch": 2.945984962406015, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 53977 + }, + { + "epoch": 2.946045112781955, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.9161, + "step": 53978 + }, + { + "epoch": 2.946105263157895, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 53979 + }, + { + "epoch": 2.9461654135338344, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 53980 + }, + { + "epoch": 2.9462255639097745, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 53981 + }, + { + "epoch": 2.946285714285714, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 53982 + }, + { + "epoch": 2.9463458646616543, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 53983 + }, + { + "epoch": 2.946406015037594, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1464, + "step": 53984 + }, + { + "epoch": 2.9464661654135336, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 53985 + }, + { + "epoch": 2.9465263157894737, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 53986 + }, + { + "epoch": 2.9465864661654138, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 53987 + }, + { + "epoch": 2.9466466165413534, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 53988 + }, + { + "epoch": 2.946706766917293, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 53989 + }, + { + "epoch": 2.946766917293233, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 53990 + }, + { + "epoch": 2.946827067669173, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 53991 + }, + { + "epoch": 2.946887218045113, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 53992 + }, + { + "epoch": 2.9469473684210525, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 53993 + }, + { + "epoch": 2.9470075187969926, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 53994 + }, + { + "epoch": 2.9470676691729323, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 53995 + }, + { + "epoch": 2.9471278195488724, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 53996 + }, + { + "epoch": 2.947187969924812, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 53997 + }, + { + "epoch": 2.9472481203007517, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9447, + "step": 53998 + }, + { + "epoch": 2.9473082706766918, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 53999 + }, + { + "epoch": 2.9473684210526314, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 54000 + }, + { + "epoch": 2.9474285714285715, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 54001 + }, + { + "epoch": 2.947488721804511, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 54002 + }, + { + "epoch": 2.9475488721804513, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 54003 + }, + { + "epoch": 2.947609022556391, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 54004 + }, + { + "epoch": 2.947669172932331, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 54005 + }, + { + "epoch": 2.9477293233082706, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 54006 + }, + { + "epoch": 2.9477894736842103, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 54007 + }, + { + "epoch": 2.9478496240601504, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.8969, + "step": 54008 + }, + { + "epoch": 2.9479097744360905, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 54009 + }, + { + "epoch": 2.94796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 54010 + }, + { + "epoch": 2.94803007518797, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 54011 + }, + { + "epoch": 2.94809022556391, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 54012 + }, + { + "epoch": 2.9481503759398495, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 54013 + }, + { + "epoch": 2.9482105263157896, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 54014 + }, + { + "epoch": 2.9482706766917293, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 54015 + }, + { + "epoch": 2.948330827067669, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 54016 + }, + { + "epoch": 2.948390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 54017 + }, + { + "epoch": 2.948451127819549, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 54018 + }, + { + "epoch": 2.9485112781954887, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 54019 + }, + { + "epoch": 2.9485714285714284, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 54020 + }, + { + "epoch": 2.9486315789473685, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 54021 + }, + { + "epoch": 2.948691729323308, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 54022 + }, + { + "epoch": 2.9487518796992482, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 54023 + }, + { + "epoch": 2.948812030075188, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9414, + "step": 54024 + }, + { + "epoch": 2.948872180451128, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 54025 + }, + { + "epoch": 2.9489323308270676, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 54026 + }, + { + "epoch": 2.9489924812030077, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 54027 + }, + { + "epoch": 2.9490526315789474, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 54028 + }, + { + "epoch": 2.949112781954887, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1125, + "step": 54029 + }, + { + "epoch": 2.949172932330827, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 54030 + }, + { + "epoch": 2.9492330827067668, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 54031 + }, + { + "epoch": 2.949293233082707, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 54032 + }, + { + "epoch": 2.9493533834586465, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9255, + "step": 54033 + }, + { + "epoch": 2.9494135338345866, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 54034 + }, + { + "epoch": 2.9494736842105262, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 54035 + }, + { + "epoch": 2.9495338345864663, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 54036 + }, + { + "epoch": 2.949593984962406, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 54037 + }, + { + "epoch": 2.9496541353383456, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 54038 + }, + { + "epoch": 2.9497142857142857, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 54039 + }, + { + "epoch": 2.949774436090226, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9258, + "step": 54040 + }, + { + "epoch": 2.9498345864661655, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 54041 + }, + { + "epoch": 2.949894736842105, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 54042 + }, + { + "epoch": 2.949954887218045, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 54043 + }, + { + "epoch": 2.950015037593985, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 54044 + }, + { + "epoch": 2.950075187969925, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 54045 + }, + { + "epoch": 2.9501353383458646, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 54046 + }, + { + "epoch": 2.9501954887218043, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 54047 + }, + { + "epoch": 2.9502556390977444, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 54048 + }, + { + "epoch": 2.9503157894736844, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 54049 + }, + { + "epoch": 2.950375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9214, + "step": 54050 + }, + { + "epoch": 2.9504360902255637, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 54051 + }, + { + "epoch": 2.950496240601504, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 54052 + }, + { + "epoch": 2.9505563909774435, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 54053 + }, + { + "epoch": 2.9506165413533836, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.929, + "step": 54054 + }, + { + "epoch": 2.9506766917293232, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 54055 + }, + { + "epoch": 2.950736842105263, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9148, + "step": 54056 + }, + { + "epoch": 2.950796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 54057 + }, + { + "epoch": 2.950857142857143, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 54058 + }, + { + "epoch": 2.9509172932330827, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 54059 + }, + { + "epoch": 2.9509774436090224, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 54060 + }, + { + "epoch": 2.9510375939849625, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 54061 + }, + { + "epoch": 2.951097744360902, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 54062 + }, + { + "epoch": 2.951157894736842, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 54063 + }, + { + "epoch": 2.951218045112782, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 54064 + }, + { + "epoch": 2.951278195488722, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 54065 + }, + { + "epoch": 2.9513383458646616, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 54066 + }, + { + "epoch": 2.9513984962406017, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 54067 + }, + { + "epoch": 2.9514586466165413, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 54068 + }, + { + "epoch": 2.951518796992481, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 54069 + }, + { + "epoch": 2.951578947368421, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 54070 + }, + { + "epoch": 2.951639097744361, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 54071 + }, + { + "epoch": 2.951699248120301, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 54072 + }, + { + "epoch": 2.9517593984962405, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 54073 + }, + { + "epoch": 2.9518195488721806, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 54074 + }, + { + "epoch": 2.95187969924812, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 54075 + }, + { + "epoch": 2.9519398496240603, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 54076 + }, + { + "epoch": 2.952, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 54077 + }, + { + "epoch": 2.9520601503759396, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 54078 + }, + { + "epoch": 2.9521203007518797, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 54079 + }, + { + "epoch": 2.95218045112782, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 54080 + }, + { + "epoch": 2.9522406015037594, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 54081 + }, + { + "epoch": 2.952300751879699, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 54082 + }, + { + "epoch": 2.952360902255639, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 54083 + }, + { + "epoch": 2.952421052631579, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 54084 + }, + { + "epoch": 2.952481203007519, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 54085 + }, + { + "epoch": 2.9525413533834586, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 54086 + }, + { + "epoch": 2.9526015037593982, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 54087 + }, + { + "epoch": 2.9526616541353383, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 54088 + }, + { + "epoch": 2.9527218045112784, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 54089 + }, + { + "epoch": 2.952781954887218, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 54090 + }, + { + "epoch": 2.9528421052631577, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.2026, + "step": 54091 + }, + { + "epoch": 2.952902255639098, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 54092 + }, + { + "epoch": 2.9529624060150375, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 54093 + }, + { + "epoch": 2.9530225563909775, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 54094 + }, + { + "epoch": 2.953082706766917, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 54095 + }, + { + "epoch": 2.9531428571428573, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 54096 + }, + { + "epoch": 2.953203007518797, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 54097 + }, + { + "epoch": 2.953263157894737, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 54098 + }, + { + "epoch": 2.9533233082706767, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 54099 + }, + { + "epoch": 2.9533834586466163, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 54100 + }, + { + "epoch": 2.9534436090225564, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2046, + "step": 54101 + }, + { + "epoch": 2.953503759398496, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9505, + "step": 54102 + }, + { + "epoch": 2.953563909774436, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9277, + "step": 54103 + }, + { + "epoch": 2.953624060150376, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 54104 + }, + { + "epoch": 2.953684210526316, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 54105 + }, + { + "epoch": 2.9537443609022556, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 54106 + }, + { + "epoch": 2.9538045112781957, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 54107 + }, + { + "epoch": 2.9538646616541353, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 54108 + }, + { + "epoch": 2.953924812030075, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 54109 + }, + { + "epoch": 2.953984962406015, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 54110 + }, + { + "epoch": 2.954045112781955, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 54111 + }, + { + "epoch": 2.954105263157895, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 54112 + }, + { + "epoch": 2.9541654135338344, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 54113 + }, + { + "epoch": 2.9542255639097745, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 54114 + }, + { + "epoch": 2.954285714285714, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 54115 + }, + { + "epoch": 2.9543458646616543, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 54116 + }, + { + "epoch": 2.954406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 54117 + }, + { + "epoch": 2.9544661654135336, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 54118 + }, + { + "epoch": 2.9545263157894737, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.1078, + "step": 54119 + }, + { + "epoch": 2.9545864661654138, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 54120 + }, + { + "epoch": 2.9546466165413534, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 54121 + }, + { + "epoch": 2.954706766917293, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 54122 + }, + { + "epoch": 2.954766917293233, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 54123 + }, + { + "epoch": 2.954827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 54124 + }, + { + "epoch": 2.954887218045113, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 54125 + }, + { + "epoch": 2.9549473684210525, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 54126 + }, + { + "epoch": 2.9550075187969926, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9504, + "step": 54127 + }, + { + "epoch": 2.9550676691729323, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 54128 + }, + { + "epoch": 2.9551278195488724, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 54129 + }, + { + "epoch": 2.955187969924812, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 54130 + }, + { + "epoch": 2.9552481203007517, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 54131 + }, + { + "epoch": 2.9553082706766918, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9514, + "step": 54132 + }, + { + "epoch": 2.9553684210526314, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 54133 + }, + { + "epoch": 2.9554285714285715, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 54134 + }, + { + "epoch": 2.955488721804511, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 54135 + }, + { + "epoch": 2.9555488721804513, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1119, + "step": 54136 + }, + { + "epoch": 2.955609022556391, + "grad_norm": 1.4140625, + "learning_rate": 3e-05, + "loss": 1.1976, + "step": 54137 + }, + { + "epoch": 2.955669172932331, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 54138 + }, + { + "epoch": 2.9557293233082707, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 54139 + }, + { + "epoch": 2.9557894736842103, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 54140 + }, + { + "epoch": 2.9558496240601504, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 54141 + }, + { + "epoch": 2.9559097744360905, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 54142 + }, + { + "epoch": 2.95596992481203, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 54143 + }, + { + "epoch": 2.95603007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 54144 + }, + { + "epoch": 2.95609022556391, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 54145 + }, + { + "epoch": 2.9561503759398495, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 54146 + }, + { + "epoch": 2.9562105263157896, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 54147 + }, + { + "epoch": 2.9562706766917293, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 54148 + }, + { + "epoch": 2.956330827067669, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 54149 + }, + { + "epoch": 2.956390977443609, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 54150 + }, + { + "epoch": 2.956451127819549, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9027, + "step": 54151 + }, + { + "epoch": 2.9565112781954888, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 54152 + }, + { + "epoch": 2.9565714285714284, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 54153 + }, + { + "epoch": 2.9566315789473685, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 54154 + }, + { + "epoch": 2.956691729323308, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 54155 + }, + { + "epoch": 2.9567518796992482, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 54156 + }, + { + "epoch": 2.956812030075188, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 54157 + }, + { + "epoch": 2.956872180451128, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 54158 + }, + { + "epoch": 2.9569323308270676, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 54159 + }, + { + "epoch": 2.9569924812030077, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 54160 + }, + { + "epoch": 2.9570526315789474, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 54161 + }, + { + "epoch": 2.957112781954887, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9407, + "step": 54162 + }, + { + "epoch": 2.957172932330827, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 54163 + }, + { + "epoch": 2.9572330827067668, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 54164 + }, + { + "epoch": 2.957293233082707, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 54165 + }, + { + "epoch": 2.9573533834586465, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 54166 + }, + { + "epoch": 2.9574135338345866, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 54167 + }, + { + "epoch": 2.9574736842105263, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 54168 + }, + { + "epoch": 2.9575338345864663, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 54169 + }, + { + "epoch": 2.957593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 54170 + }, + { + "epoch": 2.9576541353383456, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 54171 + }, + { + "epoch": 2.9577142857142857, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 54172 + }, + { + "epoch": 2.957774436090226, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 54173 + }, + { + "epoch": 2.9578345864661655, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 54174 + }, + { + "epoch": 2.957894736842105, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 54175 + }, + { + "epoch": 2.9579548872180452, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 54176 + }, + { + "epoch": 2.958015037593985, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 54177 + }, + { + "epoch": 2.958075187969925, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9169, + "step": 54178 + }, + { + "epoch": 2.9581353383458646, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 54179 + }, + { + "epoch": 2.9581954887218043, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 54180 + }, + { + "epoch": 2.9582556390977444, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 54181 + }, + { + "epoch": 2.9583157894736845, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 54182 + }, + { + "epoch": 2.958375939849624, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 54183 + }, + { + "epoch": 2.9584360902255638, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 54184 + }, + { + "epoch": 2.958496240601504, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 54185 + }, + { + "epoch": 2.9585563909774435, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 54186 + }, + { + "epoch": 2.9586165413533836, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 54187 + }, + { + "epoch": 2.9586766917293232, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 54188 + }, + { + "epoch": 2.958736842105263, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 54189 + }, + { + "epoch": 2.958796992481203, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 54190 + }, + { + "epoch": 2.958857142857143, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 54191 + }, + { + "epoch": 2.9589172932330827, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 54192 + }, + { + "epoch": 2.9589774436090224, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9389, + "step": 54193 + }, + { + "epoch": 2.9590375939849625, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 54194 + }, + { + "epoch": 2.959097744360902, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 54195 + }, + { + "epoch": 2.959157894736842, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 54196 + }, + { + "epoch": 2.959218045112782, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0851, + "step": 54197 + }, + { + "epoch": 2.959278195488722, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 54198 + }, + { + "epoch": 2.9593383458646616, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 54199 + }, + { + "epoch": 2.9593984962406017, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9533, + "step": 54200 + }, + { + "epoch": 2.9594586466165413, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 54201 + }, + { + "epoch": 2.959518796992481, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 54202 + }, + { + "epoch": 2.959578947368421, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 54203 + }, + { + "epoch": 2.9596390977443607, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 54204 + }, + { + "epoch": 2.959699248120301, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 54205 + }, + { + "epoch": 2.9597593984962405, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 54206 + }, + { + "epoch": 2.9598195488721806, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 54207 + }, + { + "epoch": 2.95987969924812, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 54208 + }, + { + "epoch": 2.9599398496240603, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 54209 + }, + { + "epoch": 2.96, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 54210 + }, + { + "epoch": 2.9600601503759396, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 54211 + }, + { + "epoch": 2.9601203007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 54212 + }, + { + "epoch": 2.96018045112782, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 54213 + }, + { + "epoch": 2.9602406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 54214 + }, + { + "epoch": 2.960300751879699, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9176, + "step": 54215 + }, + { + "epoch": 2.960360902255639, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.8762, + "step": 54216 + }, + { + "epoch": 2.960421052631579, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 54217 + }, + { + "epoch": 2.960481203007519, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.929, + "step": 54218 + }, + { + "epoch": 2.9605413533834586, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 54219 + }, + { + "epoch": 2.9606015037593982, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 54220 + }, + { + "epoch": 2.9606616541353383, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 54221 + }, + { + "epoch": 2.9607218045112784, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 54222 + }, + { + "epoch": 2.960781954887218, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 54223 + }, + { + "epoch": 2.9608421052631577, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 54224 + }, + { + "epoch": 2.960902255639098, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 54225 + }, + { + "epoch": 2.9609624060150375, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 54226 + }, + { + "epoch": 2.9610225563909776, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 54227 + }, + { + "epoch": 2.961082706766917, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 54228 + }, + { + "epoch": 2.9611428571428573, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 54229 + }, + { + "epoch": 2.961203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 54230 + }, + { + "epoch": 2.961263157894737, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 54231 + }, + { + "epoch": 2.9613233082706767, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 54232 + }, + { + "epoch": 2.9613834586466163, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 54233 + }, + { + "epoch": 2.9614436090225564, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 54234 + }, + { + "epoch": 2.961503759398496, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 54235 + }, + { + "epoch": 2.961563909774436, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 54236 + }, + { + "epoch": 2.961624060150376, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 54237 + }, + { + "epoch": 2.961684210526316, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 54238 + }, + { + "epoch": 2.9617443609022556, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 54239 + }, + { + "epoch": 2.9618045112781957, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 54240 + }, + { + "epoch": 2.9618646616541353, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 54241 + }, + { + "epoch": 2.961924812030075, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 54242 + }, + { + "epoch": 2.961984962406015, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 54243 + }, + { + "epoch": 2.962045112781955, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 54244 + }, + { + "epoch": 2.962105263157895, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 54245 + }, + { + "epoch": 2.9621654135338344, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 54246 + }, + { + "epoch": 2.9622255639097745, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 54247 + }, + { + "epoch": 2.962285714285714, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 54248 + }, + { + "epoch": 2.9623458646616543, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9443, + "step": 54249 + }, + { + "epoch": 2.962406015037594, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 54250 + }, + { + "epoch": 2.9624661654135336, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1168, + "step": 54251 + }, + { + "epoch": 2.9625263157894737, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 54252 + }, + { + "epoch": 2.9625864661654138, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 54253 + }, + { + "epoch": 2.9626466165413534, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 54254 + }, + { + "epoch": 2.962706766917293, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 54255 + }, + { + "epoch": 2.962766917293233, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 54256 + }, + { + "epoch": 2.962827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1221, + "step": 54257 + }, + { + "epoch": 2.962887218045113, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 54258 + }, + { + "epoch": 2.9629473684210526, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 54259 + }, + { + "epoch": 2.9630075187969926, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 54260 + }, + { + "epoch": 2.9630676691729323, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 54261 + }, + { + "epoch": 2.9631278195488724, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 54262 + }, + { + "epoch": 2.963187969924812, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 54263 + }, + { + "epoch": 2.9632481203007517, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 54264 + }, + { + "epoch": 2.963308270676692, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 54265 + }, + { + "epoch": 2.9633684210526314, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 54266 + }, + { + "epoch": 2.9634285714285715, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 54267 + }, + { + "epoch": 2.963488721804511, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 54268 + }, + { + "epoch": 2.9635488721804513, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 54269 + }, + { + "epoch": 2.963609022556391, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.8925, + "step": 54270 + }, + { + "epoch": 2.963669172932331, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 54271 + }, + { + "epoch": 2.9637293233082707, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 54272 + }, + { + "epoch": 2.9637894736842103, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 54273 + }, + { + "epoch": 2.9638496240601504, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 54274 + }, + { + "epoch": 2.9639097744360905, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 54275 + }, + { + "epoch": 2.96396992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 54276 + }, + { + "epoch": 2.96403007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 54277 + }, + { + "epoch": 2.96409022556391, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 54278 + }, + { + "epoch": 2.9641503759398495, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0947, + "step": 54279 + }, + { + "epoch": 2.9642105263157896, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 54280 + }, + { + "epoch": 2.9642706766917293, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1369, + "step": 54281 + }, + { + "epoch": 2.964330827067669, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 54282 + }, + { + "epoch": 2.964390977443609, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 54283 + }, + { + "epoch": 2.964451127819549, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 54284 + }, + { + "epoch": 2.9645112781954888, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 54285 + }, + { + "epoch": 2.9645714285714284, + "grad_norm": 1.953125, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 54286 + }, + { + "epoch": 2.9646315789473685, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 54287 + }, + { + "epoch": 2.964691729323308, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 54288 + }, + { + "epoch": 2.9647518796992482, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 54289 + }, + { + "epoch": 2.964812030075188, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 54290 + }, + { + "epoch": 2.9648721804511275, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 54291 + }, + { + "epoch": 2.9649323308270676, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 54292 + }, + { + "epoch": 2.9649924812030077, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 54293 + }, + { + "epoch": 2.9650526315789474, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 54294 + }, + { + "epoch": 2.965112781954887, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 54295 + }, + { + "epoch": 2.965172932330827, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 54296 + }, + { + "epoch": 2.9652330827067668, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 54297 + }, + { + "epoch": 2.965293233082707, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 54298 + }, + { + "epoch": 2.9653533834586465, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 54299 + }, + { + "epoch": 2.9654135338345866, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 54300 + }, + { + "epoch": 2.9654736842105263, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 54301 + }, + { + "epoch": 2.9655338345864664, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 54302 + }, + { + "epoch": 2.965593984962406, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9386, + "step": 54303 + }, + { + "epoch": 2.9656541353383457, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 54304 + }, + { + "epoch": 2.9657142857142857, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 54305 + }, + { + "epoch": 2.965774436090226, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 54306 + }, + { + "epoch": 2.9658345864661655, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 54307 + }, + { + "epoch": 2.965894736842105, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 54308 + }, + { + "epoch": 2.9659548872180452, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 54309 + }, + { + "epoch": 2.966015037593985, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 54310 + }, + { + "epoch": 2.966075187969925, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 54311 + }, + { + "epoch": 2.9661353383458646, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 54312 + }, + { + "epoch": 2.9661954887218043, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 54313 + }, + { + "epoch": 2.9662556390977444, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 54314 + }, + { + "epoch": 2.9663157894736845, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 54315 + }, + { + "epoch": 2.966375939849624, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 54316 + }, + { + "epoch": 2.9664360902255638, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 54317 + }, + { + "epoch": 2.966496240601504, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9393, + "step": 54318 + }, + { + "epoch": 2.9665563909774435, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 54319 + }, + { + "epoch": 2.9666165413533836, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 54320 + }, + { + "epoch": 2.9666766917293232, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 54321 + }, + { + "epoch": 2.966736842105263, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 54322 + }, + { + "epoch": 2.966796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 54323 + }, + { + "epoch": 2.966857142857143, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 54324 + }, + { + "epoch": 2.9669172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 54325 + }, + { + "epoch": 2.9669774436090224, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9228, + "step": 54326 + }, + { + "epoch": 2.9670375939849625, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 54327 + }, + { + "epoch": 2.967097744360902, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 54328 + }, + { + "epoch": 2.967157894736842, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 54329 + }, + { + "epoch": 2.967218045112782, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 54330 + }, + { + "epoch": 2.967278195488722, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 54331 + }, + { + "epoch": 2.9673383458646616, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 54332 + }, + { + "epoch": 2.9673984962406017, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 54333 + }, + { + "epoch": 2.9674586466165414, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 54334 + }, + { + "epoch": 2.967518796992481, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 54335 + }, + { + "epoch": 2.967578947368421, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 54336 + }, + { + "epoch": 2.9676390977443607, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 54337 + }, + { + "epoch": 2.967699248120301, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 54338 + }, + { + "epoch": 2.9677593984962405, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 54339 + }, + { + "epoch": 2.9678195488721806, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 54340 + }, + { + "epoch": 2.9678796992481202, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 54341 + }, + { + "epoch": 2.9679398496240603, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 54342 + }, + { + "epoch": 2.968, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 54343 + }, + { + "epoch": 2.9680601503759396, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 54344 + }, + { + "epoch": 2.9681203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.927, + "step": 54345 + }, + { + "epoch": 2.96818045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 54346 + }, + { + "epoch": 2.9682406015037595, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 54347 + }, + { + "epoch": 2.968300751879699, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 54348 + }, + { + "epoch": 2.968360902255639, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2158, + "step": 54349 + }, + { + "epoch": 2.968421052631579, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 54350 + }, + { + "epoch": 2.968481203007519, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 54351 + }, + { + "epoch": 2.9685413533834586, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 54352 + }, + { + "epoch": 2.9686015037593982, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 54353 + }, + { + "epoch": 2.9686616541353383, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 54354 + }, + { + "epoch": 2.9687218045112784, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 54355 + }, + { + "epoch": 2.968781954887218, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 54356 + }, + { + "epoch": 2.9688421052631577, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 54357 + }, + { + "epoch": 2.968902255639098, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 54358 + }, + { + "epoch": 2.9689624060150375, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1635, + "step": 54359 + }, + { + "epoch": 2.9690225563909776, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 54360 + }, + { + "epoch": 2.969082706766917, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.139, + "step": 54361 + }, + { + "epoch": 2.9691428571428573, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 54362 + }, + { + "epoch": 2.969203007518797, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.94, + "step": 54363 + }, + { + "epoch": 2.969263157894737, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 54364 + }, + { + "epoch": 2.9693233082706767, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1268, + "step": 54365 + }, + { + "epoch": 2.9693834586466163, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 54366 + }, + { + "epoch": 2.9694436090225564, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 54367 + }, + { + "epoch": 2.969503759398496, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 54368 + }, + { + "epoch": 2.969563909774436, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 54369 + }, + { + "epoch": 2.969624060150376, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 54370 + }, + { + "epoch": 2.969684210526316, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 54371 + }, + { + "epoch": 2.9697443609022556, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 54372 + }, + { + "epoch": 2.9698045112781957, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 54373 + }, + { + "epoch": 2.9698646616541353, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 54374 + }, + { + "epoch": 2.969924812030075, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 54375 + }, + { + "epoch": 2.969984962406015, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 54376 + }, + { + "epoch": 2.970045112781955, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 54377 + }, + { + "epoch": 2.970105263157895, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 54378 + }, + { + "epoch": 2.9701654135338345, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 54379 + }, + { + "epoch": 2.9702255639097745, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 54380 + }, + { + "epoch": 2.970285714285714, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 54381 + }, + { + "epoch": 2.9703458646616543, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 54382 + }, + { + "epoch": 2.970406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 54383 + }, + { + "epoch": 2.9704661654135336, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 54384 + }, + { + "epoch": 2.9705263157894737, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.3392, + "step": 54385 + }, + { + "epoch": 2.9705864661654138, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 54386 + }, + { + "epoch": 2.9706466165413534, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 54387 + }, + { + "epoch": 2.970706766917293, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 54388 + }, + { + "epoch": 2.970766917293233, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 54389 + }, + { + "epoch": 2.970827067669173, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 54390 + }, + { + "epoch": 2.970887218045113, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 54391 + }, + { + "epoch": 2.9709473684210526, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 54392 + }, + { + "epoch": 2.971007518796992, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 54393 + }, + { + "epoch": 2.9710676691729323, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9505, + "step": 54394 + }, + { + "epoch": 2.9711278195488724, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 54395 + }, + { + "epoch": 2.971187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1817, + "step": 54396 + }, + { + "epoch": 2.9712481203007517, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9253, + "step": 54397 + }, + { + "epoch": 2.971308270676692, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 54398 + }, + { + "epoch": 2.9713684210526314, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 54399 + }, + { + "epoch": 2.9714285714285715, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 54400 + }, + { + "epoch": 2.971488721804511, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 54401 + }, + { + "epoch": 2.9715488721804513, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 54402 + }, + { + "epoch": 2.971609022556391, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 54403 + }, + { + "epoch": 2.971669172932331, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 54404 + }, + { + "epoch": 2.9717293233082707, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 54405 + }, + { + "epoch": 2.9717894736842103, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 54406 + }, + { + "epoch": 2.9718496240601504, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 54407 + }, + { + "epoch": 2.9719097744360905, + "grad_norm": 2.171875, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 54408 + }, + { + "epoch": 2.97196992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 54409 + }, + { + "epoch": 2.97203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 54410 + }, + { + "epoch": 2.97209022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 54411 + }, + { + "epoch": 2.9721503759398495, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 54412 + }, + { + "epoch": 2.9722105263157896, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 54413 + }, + { + "epoch": 2.9722706766917293, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 54414 + }, + { + "epoch": 2.972330827067669, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 54415 + }, + { + "epoch": 2.972390977443609, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.936, + "step": 54416 + }, + { + "epoch": 2.972451127819549, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 54417 + }, + { + "epoch": 2.9725112781954888, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 54418 + }, + { + "epoch": 2.9725714285714284, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9279, + "step": 54419 + }, + { + "epoch": 2.9726315789473685, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 54420 + }, + { + "epoch": 2.972691729323308, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9179, + "step": 54421 + }, + { + "epoch": 2.9727518796992483, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 54422 + }, + { + "epoch": 2.972812030075188, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 54423 + }, + { + "epoch": 2.9728721804511276, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 54424 + }, + { + "epoch": 2.9729323308270676, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 54425 + }, + { + "epoch": 2.9729924812030077, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9184, + "step": 54426 + }, + { + "epoch": 2.9730526315789474, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 54427 + }, + { + "epoch": 2.973112781954887, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 54428 + }, + { + "epoch": 2.973172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 54429 + }, + { + "epoch": 2.973233082706767, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 54430 + }, + { + "epoch": 2.973293233082707, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9337, + "step": 54431 + }, + { + "epoch": 2.9733533834586465, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 54432 + }, + { + "epoch": 2.9734135338345866, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 54433 + }, + { + "epoch": 2.9734736842105263, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 54434 + }, + { + "epoch": 2.9735338345864664, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 54435 + }, + { + "epoch": 2.973593984962406, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 54436 + }, + { + "epoch": 2.9736541353383457, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 54437 + }, + { + "epoch": 2.9737142857142858, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 54438 + }, + { + "epoch": 2.9737744360902254, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 54439 + }, + { + "epoch": 2.9738345864661655, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 54440 + }, + { + "epoch": 2.973894736842105, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 54441 + }, + { + "epoch": 2.9739548872180452, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 54442 + }, + { + "epoch": 2.974015037593985, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 54443 + }, + { + "epoch": 2.974075187969925, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 54444 + }, + { + "epoch": 2.9741353383458646, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 54445 + }, + { + "epoch": 2.9741954887218043, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 54446 + }, + { + "epoch": 2.9742556390977444, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 54447 + }, + { + "epoch": 2.9743157894736845, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 54448 + }, + { + "epoch": 2.974375939849624, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 54449 + }, + { + "epoch": 2.9744360902255638, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9213, + "step": 54450 + }, + { + "epoch": 2.974496240601504, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 54451 + }, + { + "epoch": 2.9745563909774435, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 54452 + }, + { + "epoch": 2.9746165413533836, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 54453 + }, + { + "epoch": 2.9746766917293233, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1884, + "step": 54454 + }, + { + "epoch": 2.974736842105263, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 54455 + }, + { + "epoch": 2.974796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 54456 + }, + { + "epoch": 2.974857142857143, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 54457 + }, + { + "epoch": 2.9749172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 54458 + }, + { + "epoch": 2.9749774436090224, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 54459 + }, + { + "epoch": 2.9750375939849625, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.084, + "step": 54460 + }, + { + "epoch": 2.975097744360902, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 54461 + }, + { + "epoch": 2.9751578947368422, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 54462 + }, + { + "epoch": 2.975218045112782, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 54463 + }, + { + "epoch": 2.975278195488722, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 54464 + }, + { + "epoch": 2.9753383458646616, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 54465 + }, + { + "epoch": 2.9753984962406017, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 54466 + }, + { + "epoch": 2.9754586466165414, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9301, + "step": 54467 + }, + { + "epoch": 2.975518796992481, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1228, + "step": 54468 + }, + { + "epoch": 2.975578947368421, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 54469 + }, + { + "epoch": 2.9756390977443608, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 54470 + }, + { + "epoch": 2.975699248120301, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 54471 + }, + { + "epoch": 2.9757593984962405, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 54472 + }, + { + "epoch": 2.9758195488721806, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 54473 + }, + { + "epoch": 2.9758796992481202, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 54474 + }, + { + "epoch": 2.9759398496240603, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 54475 + }, + { + "epoch": 2.976, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 54476 + }, + { + "epoch": 2.9760601503759396, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 54477 + }, + { + "epoch": 2.9761203007518797, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 54478 + }, + { + "epoch": 2.97618045112782, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 54479 + }, + { + "epoch": 2.9762406015037595, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 54480 + }, + { + "epoch": 2.976300751879699, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 54481 + }, + { + "epoch": 2.976360902255639, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 54482 + }, + { + "epoch": 2.976421052631579, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 54483 + }, + { + "epoch": 2.976481203007519, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 54484 + }, + { + "epoch": 2.9765413533834586, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 54485 + }, + { + "epoch": 2.9766015037593982, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 54486 + }, + { + "epoch": 2.9766616541353383, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 54487 + }, + { + "epoch": 2.9767218045112784, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 54488 + }, + { + "epoch": 2.976781954887218, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 54489 + }, + { + "epoch": 2.9768421052631577, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 54490 + }, + { + "epoch": 2.976902255639098, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 54491 + }, + { + "epoch": 2.9769624060150375, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 54492 + }, + { + "epoch": 2.9770225563909776, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 54493 + }, + { + "epoch": 2.977082706766917, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9121, + "step": 54494 + }, + { + "epoch": 2.977142857142857, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 54495 + }, + { + "epoch": 2.977203007518797, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9444, + "step": 54496 + }, + { + "epoch": 2.977263157894737, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9595, + "step": 54497 + }, + { + "epoch": 2.9773233082706767, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 54498 + }, + { + "epoch": 2.9773834586466164, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 54499 + }, + { + "epoch": 2.9774436090225564, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 54500 + }, + { + "epoch": 2.977503759398496, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 54501 + }, + { + "epoch": 2.977563909774436, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 54502 + }, + { + "epoch": 2.977624060150376, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.2372, + "step": 54503 + }, + { + "epoch": 2.977684210526316, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 54504 + }, + { + "epoch": 2.9777443609022556, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 54505 + }, + { + "epoch": 2.9778045112781957, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 54506 + }, + { + "epoch": 2.9778646616541353, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 54507 + }, + { + "epoch": 2.977924812030075, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 54508 + }, + { + "epoch": 2.977984962406015, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 54509 + }, + { + "epoch": 2.978045112781955, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 54510 + }, + { + "epoch": 2.978105263157895, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 54511 + }, + { + "epoch": 2.9781654135338345, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 54512 + }, + { + "epoch": 2.9782255639097746, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 54513 + }, + { + "epoch": 2.978285714285714, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 54514 + }, + { + "epoch": 2.9783458646616543, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 54515 + }, + { + "epoch": 2.978406015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 54516 + }, + { + "epoch": 2.9784661654135336, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 54517 + }, + { + "epoch": 2.9785263157894737, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 54518 + }, + { + "epoch": 2.978586466165414, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 54519 + }, + { + "epoch": 2.9786466165413534, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 54520 + }, + { + "epoch": 2.978706766917293, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.8555, + "step": 54521 + }, + { + "epoch": 2.978766917293233, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 54522 + }, + { + "epoch": 2.978827067669173, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 54523 + }, + { + "epoch": 2.978887218045113, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 54524 + }, + { + "epoch": 2.9789473684210526, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9306, + "step": 54525 + }, + { + "epoch": 2.979007518796992, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 54526 + }, + { + "epoch": 2.9790676691729323, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 54527 + }, + { + "epoch": 2.9791278195488724, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 54528 + }, + { + "epoch": 2.979187969924812, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 54529 + }, + { + "epoch": 2.9792481203007517, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 54530 + }, + { + "epoch": 2.979308270676692, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 54531 + }, + { + "epoch": 2.9793684210526314, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 54532 + }, + { + "epoch": 2.9794285714285715, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 54533 + }, + { + "epoch": 2.979488721804511, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 54534 + }, + { + "epoch": 2.9795488721804513, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 54535 + }, + { + "epoch": 2.979609022556391, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 54536 + }, + { + "epoch": 2.979669172932331, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 54537 + }, + { + "epoch": 2.9797293233082707, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 54538 + }, + { + "epoch": 2.9797894736842103, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 54539 + }, + { + "epoch": 2.9798496240601504, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1396, + "step": 54540 + }, + { + "epoch": 2.97990977443609, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 54541 + }, + { + "epoch": 2.97996992481203, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 54542 + }, + { + "epoch": 2.98003007518797, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 54543 + }, + { + "epoch": 2.98009022556391, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 54544 + }, + { + "epoch": 2.9801503759398495, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 54545 + }, + { + "epoch": 2.9802105263157896, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 54546 + }, + { + "epoch": 2.9802706766917293, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 54547 + }, + { + "epoch": 2.980330827067669, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 54548 + }, + { + "epoch": 2.980390977443609, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 54549 + }, + { + "epoch": 2.980451127819549, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 54550 + }, + { + "epoch": 2.9805112781954888, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.093, + "step": 54551 + }, + { + "epoch": 2.9805714285714284, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 54552 + }, + { + "epoch": 2.9806315789473685, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 54553 + }, + { + "epoch": 2.980691729323308, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 54554 + }, + { + "epoch": 2.9807518796992483, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9374, + "step": 54555 + }, + { + "epoch": 2.980812030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 54556 + }, + { + "epoch": 2.9808721804511276, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 54557 + }, + { + "epoch": 2.9809323308270677, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 54558 + }, + { + "epoch": 2.9809924812030077, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 54559 + }, + { + "epoch": 2.9810526315789474, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 54560 + }, + { + "epoch": 2.981112781954887, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 54561 + }, + { + "epoch": 2.981172932330827, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 54562 + }, + { + "epoch": 2.981233082706767, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 54563 + }, + { + "epoch": 2.981293233082707, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 54564 + }, + { + "epoch": 2.9813533834586465, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 54565 + }, + { + "epoch": 2.9814135338345866, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 54566 + }, + { + "epoch": 2.9814736842105263, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 54567 + }, + { + "epoch": 2.9815338345864664, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 54568 + }, + { + "epoch": 2.981593984962406, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 54569 + }, + { + "epoch": 2.9816541353383457, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 54570 + }, + { + "epoch": 2.9817142857142858, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 54571 + }, + { + "epoch": 2.9817744360902254, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 54572 + }, + { + "epoch": 2.9818345864661655, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 54573 + }, + { + "epoch": 2.981894736842105, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9007, + "step": 54574 + }, + { + "epoch": 2.9819548872180452, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 54575 + }, + { + "epoch": 2.982015037593985, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 54576 + }, + { + "epoch": 2.982075187969925, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 54577 + }, + { + "epoch": 2.9821353383458646, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 54578 + }, + { + "epoch": 2.9821954887218043, + "grad_norm": 1.6953125, + "learning_rate": 3e-05, + "loss": 1.1172, + "step": 54579 + }, + { + "epoch": 2.9822556390977444, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 54580 + }, + { + "epoch": 2.9823157894736845, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 54581 + }, + { + "epoch": 2.982375939849624, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 54582 + }, + { + "epoch": 2.9824360902255638, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1064, + "step": 54583 + }, + { + "epoch": 2.982496240601504, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 54584 + }, + { + "epoch": 2.9825563909774435, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 54585 + }, + { + "epoch": 2.9826165413533836, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 54586 + }, + { + "epoch": 2.9826766917293233, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 54587 + }, + { + "epoch": 2.982736842105263, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 54588 + }, + { + "epoch": 2.982796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9309, + "step": 54589 + }, + { + "epoch": 2.982857142857143, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 54590 + }, + { + "epoch": 2.9829172932330827, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.119, + "step": 54591 + }, + { + "epoch": 2.9829774436090224, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 54592 + }, + { + "epoch": 2.9830375939849625, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 54593 + }, + { + "epoch": 2.983097744360902, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 54594 + }, + { + "epoch": 2.9831578947368422, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 54595 + }, + { + "epoch": 2.983218045112782, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 54596 + }, + { + "epoch": 2.9832781954887215, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 54597 + }, + { + "epoch": 2.9833383458646616, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.894, + "step": 54598 + }, + { + "epoch": 2.9833984962406017, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 54599 + }, + { + "epoch": 2.9834586466165414, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 54600 + }, + { + "epoch": 2.983518796992481, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 54601 + }, + { + "epoch": 2.983578947368421, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 54602 + }, + { + "epoch": 2.9836390977443608, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 54603 + }, + { + "epoch": 2.983699248120301, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9249, + "step": 54604 + }, + { + "epoch": 2.9837593984962405, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 54605 + }, + { + "epoch": 2.9838195488721806, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 54606 + }, + { + "epoch": 2.9838796992481202, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 54607 + }, + { + "epoch": 2.9839398496240603, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 54608 + }, + { + "epoch": 2.984, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 54609 + }, + { + "epoch": 2.9840601503759396, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1294, + "step": 54610 + }, + { + "epoch": 2.9841203007518797, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 54611 + }, + { + "epoch": 2.98418045112782, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 54612 + }, + { + "epoch": 2.9842406015037595, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 54613 + }, + { + "epoch": 2.984300751879699, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.8769, + "step": 54614 + }, + { + "epoch": 2.984360902255639, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 54615 + }, + { + "epoch": 2.984421052631579, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 54616 + }, + { + "epoch": 2.984481203007519, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 54617 + }, + { + "epoch": 2.9845413533834586, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 54618 + }, + { + "epoch": 2.9846015037593983, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 54619 + }, + { + "epoch": 2.9846616541353383, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1141, + "step": 54620 + }, + { + "epoch": 2.9847218045112784, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 54621 + }, + { + "epoch": 2.984781954887218, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 54622 + }, + { + "epoch": 2.9848421052631577, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 54623 + }, + { + "epoch": 2.984902255639098, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9241, + "step": 54624 + }, + { + "epoch": 2.9849624060150375, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 54625 + }, + { + "epoch": 2.9850225563909776, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 54626 + }, + { + "epoch": 2.9850827067669172, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9384, + "step": 54627 + }, + { + "epoch": 2.985142857142857, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 54628 + }, + { + "epoch": 2.985203007518797, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 54629 + }, + { + "epoch": 2.985263157894737, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 54630 + }, + { + "epoch": 2.9853233082706767, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 54631 + }, + { + "epoch": 2.9853834586466164, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 54632 + }, + { + "epoch": 2.9854436090225565, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 54633 + }, + { + "epoch": 2.985503759398496, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 54634 + }, + { + "epoch": 2.985563909774436, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 54635 + }, + { + "epoch": 2.985624060150376, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 54636 + }, + { + "epoch": 2.985684210526316, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 54637 + }, + { + "epoch": 2.9857443609022556, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 54638 + }, + { + "epoch": 2.9858045112781957, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 54639 + }, + { + "epoch": 2.9858646616541353, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 54640 + }, + { + "epoch": 2.985924812030075, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 54641 + }, + { + "epoch": 2.985984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 54642 + }, + { + "epoch": 2.9860451127819547, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 54643 + }, + { + "epoch": 2.986105263157895, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 54644 + }, + { + "epoch": 2.9861654135338345, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 54645 + }, + { + "epoch": 2.9862255639097746, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 54646 + }, + { + "epoch": 2.986285714285714, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 54647 + }, + { + "epoch": 2.9863458646616543, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 54648 + }, + { + "epoch": 2.986406015037594, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 54649 + }, + { + "epoch": 2.9864661654135336, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 54650 + }, + { + "epoch": 2.9865263157894737, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 54651 + }, + { + "epoch": 2.986586466165414, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 54652 + }, + { + "epoch": 2.9866466165413534, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 54653 + }, + { + "epoch": 2.986706766917293, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.152, + "step": 54654 + }, + { + "epoch": 2.986766917293233, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 54655 + }, + { + "epoch": 2.986827067669173, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 54656 + }, + { + "epoch": 2.986887218045113, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 54657 + }, + { + "epoch": 2.9869473684210526, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 54658 + }, + { + "epoch": 2.9870075187969922, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 54659 + }, + { + "epoch": 2.9870676691729323, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 54660 + }, + { + "epoch": 2.9871278195488724, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 54661 + }, + { + "epoch": 2.987187969924812, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 54662 + }, + { + "epoch": 2.9872481203007517, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 54663 + }, + { + "epoch": 2.987308270676692, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 54664 + }, + { + "epoch": 2.9873684210526315, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 54665 + }, + { + "epoch": 2.9874285714285715, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 54666 + }, + { + "epoch": 2.987488721804511, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 54667 + }, + { + "epoch": 2.9875488721804513, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9344, + "step": 54668 + }, + { + "epoch": 2.987609022556391, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 54669 + }, + { + "epoch": 2.987669172932331, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.947, + "step": 54670 + }, + { + "epoch": 2.9877293233082707, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 54671 + }, + { + "epoch": 2.9877894736842103, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 54672 + }, + { + "epoch": 2.9878496240601504, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 54673 + }, + { + "epoch": 2.98790977443609, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 54674 + }, + { + "epoch": 2.98796992481203, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 54675 + }, + { + "epoch": 2.98803007518797, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 54676 + }, + { + "epoch": 2.98809022556391, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 54677 + }, + { + "epoch": 2.9881503759398496, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 54678 + }, + { + "epoch": 2.9882105263157897, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 54679 + }, + { + "epoch": 2.9882706766917293, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 54680 + }, + { + "epoch": 2.988330827067669, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 54681 + }, + { + "epoch": 2.988390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 54682 + }, + { + "epoch": 2.988451127819549, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9407, + "step": 54683 + }, + { + "epoch": 2.988511278195489, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 54684 + }, + { + "epoch": 2.9885714285714284, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9364, + "step": 54685 + }, + { + "epoch": 2.9886315789473685, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 54686 + }, + { + "epoch": 2.988691729323308, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 54687 + }, + { + "epoch": 2.9887518796992483, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 54688 + }, + { + "epoch": 2.988812030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 54689 + }, + { + "epoch": 2.9888721804511276, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 54690 + }, + { + "epoch": 2.9889323308270677, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1056, + "step": 54691 + }, + { + "epoch": 2.9889924812030078, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 54692 + }, + { + "epoch": 2.9890526315789474, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9306, + "step": 54693 + }, + { + "epoch": 2.989112781954887, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 54694 + }, + { + "epoch": 2.989172932330827, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 54695 + }, + { + "epoch": 2.989233082706767, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 54696 + }, + { + "epoch": 2.989293233082707, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 54697 + }, + { + "epoch": 2.9893533834586465, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 54698 + }, + { + "epoch": 2.989413533834586, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 54699 + }, + { + "epoch": 2.9894736842105263, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 54700 + }, + { + "epoch": 2.9895338345864664, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 54701 + }, + { + "epoch": 2.989593984962406, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 54702 + }, + { + "epoch": 2.9896541353383457, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 54703 + }, + { + "epoch": 2.9897142857142858, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 54704 + }, + { + "epoch": 2.9897744360902254, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 54705 + }, + { + "epoch": 2.9898345864661655, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 54706 + }, + { + "epoch": 2.989894736842105, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1097, + "step": 54707 + }, + { + "epoch": 2.9899548872180453, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 54708 + }, + { + "epoch": 2.990015037593985, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 54709 + }, + { + "epoch": 2.990075187969925, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 54710 + }, + { + "epoch": 2.9901353383458646, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 54711 + }, + { + "epoch": 2.9901954887218043, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 54712 + }, + { + "epoch": 2.9902556390977444, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 54713 + }, + { + "epoch": 2.9903157894736845, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 54714 + }, + { + "epoch": 2.990375939849624, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 54715 + }, + { + "epoch": 2.990436090225564, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 54716 + }, + { + "epoch": 2.990496240601504, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 54717 + }, + { + "epoch": 2.9905563909774435, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 54718 + }, + { + "epoch": 2.9906165413533836, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 54719 + }, + { + "epoch": 2.9906766917293233, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 54720 + }, + { + "epoch": 2.990736842105263, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 54721 + }, + { + "epoch": 2.990796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 54722 + }, + { + "epoch": 2.990857142857143, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 54723 + }, + { + "epoch": 2.9909172932330828, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 54724 + }, + { + "epoch": 2.9909774436090224, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 54725 + }, + { + "epoch": 2.9910375939849625, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 54726 + }, + { + "epoch": 2.991097744360902, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 54727 + }, + { + "epoch": 2.9911578947368422, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1959, + "step": 54728 + }, + { + "epoch": 2.991218045112782, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 54729 + }, + { + "epoch": 2.9912781954887215, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 54730 + }, + { + "epoch": 2.9913383458646616, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 54731 + }, + { + "epoch": 2.9913984962406017, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 54732 + }, + { + "epoch": 2.9914586466165414, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 54733 + }, + { + "epoch": 2.991518796992481, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 54734 + }, + { + "epoch": 2.991578947368421, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 54735 + }, + { + "epoch": 2.9916390977443608, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 54736 + }, + { + "epoch": 2.991699248120301, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 54737 + }, + { + "epoch": 2.9917593984962405, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9388, + "step": 54738 + }, + { + "epoch": 2.9918195488721806, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 54739 + }, + { + "epoch": 2.9918796992481203, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 54740 + }, + { + "epoch": 2.9919398496240603, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 54741 + }, + { + "epoch": 2.992, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9516, + "step": 54742 + }, + { + "epoch": 2.9920601503759396, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 54743 + }, + { + "epoch": 2.9921203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 54744 + }, + { + "epoch": 2.9921804511278194, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9407, + "step": 54745 + }, + { + "epoch": 2.9922406015037595, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 54746 + }, + { + "epoch": 2.992300751879699, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 54747 + }, + { + "epoch": 2.992360902255639, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 54748 + }, + { + "epoch": 2.992421052631579, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 54749 + }, + { + "epoch": 2.992481203007519, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 54750 + }, + { + "epoch": 2.9925413533834586, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9342, + "step": 54751 + }, + { + "epoch": 2.9926015037593983, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 54752 + }, + { + "epoch": 2.9926616541353384, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1107, + "step": 54753 + }, + { + "epoch": 2.9927218045112784, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 54754 + }, + { + "epoch": 2.992781954887218, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 54755 + }, + { + "epoch": 2.9928421052631577, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 54756 + }, + { + "epoch": 2.992902255639098, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0856, + "step": 54757 + }, + { + "epoch": 2.9929624060150375, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 54758 + }, + { + "epoch": 2.9930225563909776, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 54759 + }, + { + "epoch": 2.9930827067669172, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 54760 + }, + { + "epoch": 2.993142857142857, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 54761 + }, + { + "epoch": 2.993203007518797, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 54762 + }, + { + "epoch": 2.993263157894737, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1888, + "step": 54763 + }, + { + "epoch": 2.9933233082706767, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 54764 + }, + { + "epoch": 2.9933834586466164, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 54765 + }, + { + "epoch": 2.9934436090225565, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 54766 + }, + { + "epoch": 2.993503759398496, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 54767 + }, + { + "epoch": 2.993563909774436, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 54768 + }, + { + "epoch": 2.993624060150376, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9099, + "step": 54769 + }, + { + "epoch": 2.993684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 54770 + }, + { + "epoch": 2.9937443609022556, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9256, + "step": 54771 + }, + { + "epoch": 2.9938045112781957, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 54772 + }, + { + "epoch": 2.9938646616541353, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 54773 + }, + { + "epoch": 2.993924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 54774 + }, + { + "epoch": 2.993984962406015, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 54775 + }, + { + "epoch": 2.9940451127819547, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 54776 + }, + { + "epoch": 2.994105263157895, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 54777 + }, + { + "epoch": 2.9941654135338345, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 54778 + }, + { + "epoch": 2.9942255639097746, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 54779 + }, + { + "epoch": 2.994285714285714, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 54780 + }, + { + "epoch": 2.9943458646616543, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 54781 + }, + { + "epoch": 2.994406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 54782 + }, + { + "epoch": 2.9944661654135336, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9317, + "step": 54783 + }, + { + "epoch": 2.9945263157894737, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 54784 + }, + { + "epoch": 2.994586466165414, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 54785 + }, + { + "epoch": 2.9946466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 54786 + }, + { + "epoch": 2.994706766917293, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 54787 + }, + { + "epoch": 2.994766917293233, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 54788 + }, + { + "epoch": 2.994827067669173, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 54789 + }, + { + "epoch": 2.994887218045113, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 54790 + }, + { + "epoch": 2.9949473684210526, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 54791 + }, + { + "epoch": 2.9950075187969922, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0751, + "step": 54792 + }, + { + "epoch": 2.9950676691729323, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 54793 + }, + { + "epoch": 2.9951278195488724, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 54794 + }, + { + "epoch": 2.995187969924812, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 54795 + }, + { + "epoch": 2.9952481203007517, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 54796 + }, + { + "epoch": 2.995308270676692, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9084, + "step": 54797 + }, + { + "epoch": 2.9953684210526315, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 54798 + }, + { + "epoch": 2.9954285714285716, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 54799 + }, + { + "epoch": 2.995488721804511, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9338, + "step": 54800 + }, + { + "epoch": 2.9955488721804513, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 54801 + }, + { + "epoch": 2.995609022556391, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 54802 + }, + { + "epoch": 2.995669172932331, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 54803 + }, + { + "epoch": 2.9957293233082707, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9124, + "step": 54804 + }, + { + "epoch": 2.9957894736842103, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 54805 + }, + { + "epoch": 2.9958496240601504, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 54806 + }, + { + "epoch": 2.99590977443609, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 54807 + }, + { + "epoch": 2.99596992481203, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 54808 + }, + { + "epoch": 2.99603007518797, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 54809 + }, + { + "epoch": 2.99609022556391, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 54810 + }, + { + "epoch": 2.9961503759398496, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 54811 + }, + { + "epoch": 2.9962105263157897, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 54812 + }, + { + "epoch": 2.9962706766917293, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 54813 + }, + { + "epoch": 2.996330827067669, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9161, + "step": 54814 + }, + { + "epoch": 2.996390977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 54815 + }, + { + "epoch": 2.996451127819549, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 54816 + }, + { + "epoch": 2.996511278195489, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 54817 + }, + { + "epoch": 2.9965714285714284, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 54818 + }, + { + "epoch": 2.9966315789473685, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 54819 + }, + { + "epoch": 2.996691729323308, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 54820 + }, + { + "epoch": 2.9967518796992483, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 54821 + }, + { + "epoch": 2.996812030075188, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 54822 + }, + { + "epoch": 2.9968721804511276, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 54823 + }, + { + "epoch": 2.9969323308270677, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 54824 + }, + { + "epoch": 2.9969924812030078, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 54825 + }, + { + "epoch": 2.9970526315789474, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 54826 + }, + { + "epoch": 2.997112781954887, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 54827 + }, + { + "epoch": 2.997172932330827, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 54828 + }, + { + "epoch": 2.997233082706767, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 54829 + }, + { + "epoch": 2.997293233082707, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 54830 + }, + { + "epoch": 2.9973533834586465, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 54831 + }, + { + "epoch": 2.997413533834586, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 54832 + }, + { + "epoch": 2.9974736842105263, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 54833 + }, + { + "epoch": 2.9975338345864664, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1032, + "step": 54834 + }, + { + "epoch": 2.997593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 54835 + }, + { + "epoch": 2.9976541353383457, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 54836 + }, + { + "epoch": 2.9977142857142858, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 54837 + }, + { + "epoch": 2.9977744360902254, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 54838 + }, + { + "epoch": 2.9978345864661655, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 54839 + }, + { + "epoch": 2.997894736842105, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 54840 + }, + { + "epoch": 2.9979548872180453, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 54841 + }, + { + "epoch": 2.998015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 54842 + }, + { + "epoch": 2.998075187969925, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 54843 + }, + { + "epoch": 2.9981353383458647, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 54844 + }, + { + "epoch": 2.9981954887218043, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 54845 + }, + { + "epoch": 2.9982556390977444, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 54846 + }, + { + "epoch": 2.998315789473684, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9537, + "step": 54847 + }, + { + "epoch": 2.998375939849624, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 54848 + }, + { + "epoch": 2.998436090225564, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 54849 + }, + { + "epoch": 2.998496240601504, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9291, + "step": 54850 + }, + { + "epoch": 2.9985563909774435, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 54851 + }, + { + "epoch": 2.9986165413533836, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9414, + "step": 54852 + }, + { + "epoch": 2.9986766917293233, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 54853 + }, + { + "epoch": 2.998736842105263, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 54854 + }, + { + "epoch": 2.998796992481203, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 54855 + }, + { + "epoch": 2.998857142857143, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 54856 + }, + { + "epoch": 2.9989172932330828, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 54857 + }, + { + "epoch": 2.9989774436090224, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9178, + "step": 54858 + }, + { + "epoch": 2.9990375939849625, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 54859 + }, + { + "epoch": 2.999097744360902, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 54860 + }, + { + "epoch": 2.9991578947368422, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 54861 + }, + { + "epoch": 2.999218045112782, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 54862 + }, + { + "epoch": 2.9992781954887215, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 54863 + }, + { + "epoch": 2.9993383458646616, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 54864 + }, + { + "epoch": 2.9993984962406017, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1178, + "step": 54865 + }, + { + "epoch": 2.9994586466165414, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 54866 + }, + { + "epoch": 2.999518796992481, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 54867 + }, + { + "epoch": 2.999578947368421, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 54868 + }, + { + "epoch": 2.9996390977443608, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 54869 + }, + { + "epoch": 2.999699248120301, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 54870 + }, + { + "epoch": 2.9997593984962405, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 54871 + }, + { + "epoch": 2.9998195488721806, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 54872 + }, + { + "epoch": 2.9998796992481203, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 54873 + }, + { + "epoch": 2.9999398496240604, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 54874 + }, + { + "epoch": 3.0, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 54875 + }, + { + "epoch": 3.0000601503759396, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 54876 + }, + { + "epoch": 3.0001203007518797, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 54877 + }, + { + "epoch": 3.0001804511278194, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9274, + "step": 54878 + }, + { + "epoch": 3.0002406015037595, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 54879 + }, + { + "epoch": 3.000300751879699, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 54880 + }, + { + "epoch": 3.0003609022556392, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 54881 + }, + { + "epoch": 3.000421052631579, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 54882 + }, + { + "epoch": 3.000481203007519, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 54883 + }, + { + "epoch": 3.0005413533834586, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 54884 + }, + { + "epoch": 3.0006015037593987, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 54885 + }, + { + "epoch": 3.0006616541353384, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.8484, + "step": 54886 + }, + { + "epoch": 3.000721804511278, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1387, + "step": 54887 + }, + { + "epoch": 3.000781954887218, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 54888 + }, + { + "epoch": 3.0008421052631578, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 54889 + }, + { + "epoch": 3.000902255639098, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 54890 + }, + { + "epoch": 3.0009624060150375, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1739, + "step": 54891 + }, + { + "epoch": 3.0010225563909776, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9416, + "step": 54892 + }, + { + "epoch": 3.0010827067669172, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 54893 + }, + { + "epoch": 3.0011428571428573, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9106, + "step": 54894 + }, + { + "epoch": 3.001203007518797, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 54895 + }, + { + "epoch": 3.0012631578947366, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 54896 + }, + { + "epoch": 3.0013233082706767, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 54897 + }, + { + "epoch": 3.0013834586466164, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 54898 + }, + { + "epoch": 3.0014436090225565, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 54899 + }, + { + "epoch": 3.001503759398496, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 54900 + }, + { + "epoch": 3.001563909774436, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 54901 + }, + { + "epoch": 3.001624060150376, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 54902 + }, + { + "epoch": 3.001684210526316, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 54903 + }, + { + "epoch": 3.0017443609022556, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 54904 + }, + { + "epoch": 3.0018045112781957, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 54905 + }, + { + "epoch": 3.0018646616541353, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9325, + "step": 54906 + }, + { + "epoch": 3.001924812030075, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 54907 + }, + { + "epoch": 3.001984962406015, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 54908 + }, + { + "epoch": 3.0020451127819547, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 54909 + }, + { + "epoch": 3.002105263157895, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 54910 + }, + { + "epoch": 3.0021654135338345, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 54911 + }, + { + "epoch": 3.0022255639097746, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 54912 + }, + { + "epoch": 3.0022857142857142, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 54913 + }, + { + "epoch": 3.0023458646616543, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 54914 + }, + { + "epoch": 3.002406015037594, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 54915 + }, + { + "epoch": 3.0024661654135336, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 54916 + }, + { + "epoch": 3.0025263157894737, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0598, + "step": 54917 + }, + { + "epoch": 3.0025864661654134, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 54918 + }, + { + "epoch": 3.0026466165413535, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 54919 + }, + { + "epoch": 3.002706766917293, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1015, + "step": 54920 + }, + { + "epoch": 3.002766917293233, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 54921 + }, + { + "epoch": 3.002827067669173, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 54922 + }, + { + "epoch": 3.002887218045113, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 54923 + }, + { + "epoch": 3.0029473684210526, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 54924 + }, + { + "epoch": 3.0030075187969927, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 54925 + }, + { + "epoch": 3.0030676691729323, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 54926 + }, + { + "epoch": 3.003127819548872, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 54927 + }, + { + "epoch": 3.003187969924812, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 54928 + }, + { + "epoch": 3.0032481203007517, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 54929 + }, + { + "epoch": 3.003308270676692, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 54930 + }, + { + "epoch": 3.0033684210526315, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 54931 + }, + { + "epoch": 3.0034285714285716, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 54932 + }, + { + "epoch": 3.003488721804511, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 54933 + }, + { + "epoch": 3.0035488721804513, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 54934 + }, + { + "epoch": 3.003609022556391, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 54935 + }, + { + "epoch": 3.003669172932331, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 54936 + }, + { + "epoch": 3.0037293233082707, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1102, + "step": 54937 + }, + { + "epoch": 3.0037894736842103, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 54938 + }, + { + "epoch": 3.0038496240601504, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 54939 + }, + { + "epoch": 3.00390977443609, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 54940 + }, + { + "epoch": 3.00396992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1202, + "step": 54941 + }, + { + "epoch": 3.00403007518797, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 54942 + }, + { + "epoch": 3.00409022556391, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 54943 + }, + { + "epoch": 3.0041503759398496, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 54944 + }, + { + "epoch": 3.0042105263157897, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9677, + "step": 54945 + }, + { + "epoch": 3.0042706766917293, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9182, + "step": 54946 + }, + { + "epoch": 3.004330827067669, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 54947 + }, + { + "epoch": 3.004390977443609, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1425, + "step": 54948 + }, + { + "epoch": 3.0044511278195487, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 54949 + }, + { + "epoch": 3.004511278195489, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9345, + "step": 54950 + }, + { + "epoch": 3.0045714285714284, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 54951 + }, + { + "epoch": 3.0046315789473685, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 54952 + }, + { + "epoch": 3.004691729323308, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 54953 + }, + { + "epoch": 3.0047518796992483, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 54954 + }, + { + "epoch": 3.004812030075188, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 54955 + }, + { + "epoch": 3.004872180451128, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 54956 + }, + { + "epoch": 3.0049323308270677, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 54957 + }, + { + "epoch": 3.0049924812030073, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9595, + "step": 54958 + }, + { + "epoch": 3.0050526315789474, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 54959 + }, + { + "epoch": 3.005112781954887, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 54960 + }, + { + "epoch": 3.005172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 54961 + }, + { + "epoch": 3.005233082706767, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 54962 + }, + { + "epoch": 3.005293233082707, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 54963 + }, + { + "epoch": 3.0053533834586466, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 54964 + }, + { + "epoch": 3.0054135338345866, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 54965 + }, + { + "epoch": 3.0054736842105263, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 54966 + }, + { + "epoch": 3.005533834586466, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 54967 + }, + { + "epoch": 3.005593984962406, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 54968 + }, + { + "epoch": 3.0056541353383457, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 54969 + }, + { + "epoch": 3.005714285714286, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9198, + "step": 54970 + }, + { + "epoch": 3.0057744360902254, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 54971 + }, + { + "epoch": 3.0058345864661655, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 54972 + }, + { + "epoch": 3.005894736842105, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 54973 + }, + { + "epoch": 3.0059548872180453, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 54974 + }, + { + "epoch": 3.006015037593985, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 54975 + }, + { + "epoch": 3.006075187969925, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 54976 + }, + { + "epoch": 3.0061353383458647, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 54977 + }, + { + "epoch": 3.0061954887218043, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9238, + "step": 54978 + }, + { + "epoch": 3.0062556390977444, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 54979 + }, + { + "epoch": 3.006315789473684, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9343, + "step": 54980 + }, + { + "epoch": 3.006375939849624, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 54981 + }, + { + "epoch": 3.006436090225564, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9322, + "step": 54982 + }, + { + "epoch": 3.006496240601504, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0702, + "step": 54983 + }, + { + "epoch": 3.0065563909774435, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 54984 + }, + { + "epoch": 3.0066165413533836, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 54985 + }, + { + "epoch": 3.0066766917293233, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 54986 + }, + { + "epoch": 3.0067368421052634, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 54987 + }, + { + "epoch": 3.006796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 54988 + }, + { + "epoch": 3.0068571428571427, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9165, + "step": 54989 + }, + { + "epoch": 3.0069172932330828, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 54990 + }, + { + "epoch": 3.0069774436090224, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 54991 + }, + { + "epoch": 3.0070375939849625, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1462, + "step": 54992 + }, + { + "epoch": 3.007097744360902, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9217, + "step": 54993 + }, + { + "epoch": 3.0071578947368423, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 54994 + }, + { + "epoch": 3.007218045112782, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 54995 + }, + { + "epoch": 3.007278195488722, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9186, + "step": 54996 + }, + { + "epoch": 3.0073383458646616, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 54997 + }, + { + "epoch": 3.0073984962406013, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0706, + "step": 54998 + }, + { + "epoch": 3.0074586466165414, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 54999 + }, + { + "epoch": 3.007518796992481, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 55000 + }, + { + "epoch": 3.007578947368421, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 55001 + }, + { + "epoch": 3.007639097744361, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 55002 + }, + { + "epoch": 3.007699248120301, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 55003 + }, + { + "epoch": 3.0077593984962405, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 55004 + }, + { + "epoch": 3.0078195488721806, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 55005 + }, + { + "epoch": 3.0078796992481203, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 55006 + }, + { + "epoch": 3.0079398496240604, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 55007 + }, + { + "epoch": 3.008, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 55008 + }, + { + "epoch": 3.0080601503759397, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 55009 + }, + { + "epoch": 3.0081203007518797, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 55010 + }, + { + "epoch": 3.0081804511278194, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 55011 + }, + { + "epoch": 3.0082406015037595, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 55012 + }, + { + "epoch": 3.008300751879699, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 55013 + }, + { + "epoch": 3.0083609022556392, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 55014 + }, + { + "epoch": 3.008421052631579, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 55015 + }, + { + "epoch": 3.008481203007519, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 55016 + }, + { + "epoch": 3.0085413533834586, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 55017 + }, + { + "epoch": 3.0086015037593983, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 55018 + }, + { + "epoch": 3.0086616541353384, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 55019 + }, + { + "epoch": 3.008721804511278, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 55020 + }, + { + "epoch": 3.008781954887218, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 55021 + }, + { + "epoch": 3.0088421052631578, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 55022 + }, + { + "epoch": 3.008902255639098, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9297, + "step": 55023 + }, + { + "epoch": 3.0089624060150375, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 55024 + }, + { + "epoch": 3.0090225563909776, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 55025 + }, + { + "epoch": 3.0090827067669172, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 55026 + }, + { + "epoch": 3.0091428571428573, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 55027 + }, + { + "epoch": 3.009203007518797, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 55028 + }, + { + "epoch": 3.0092631578947366, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 55029 + }, + { + "epoch": 3.0093233082706767, + "grad_norm": 1.5078125, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 55030 + }, + { + "epoch": 3.0093834586466164, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 55031 + }, + { + "epoch": 3.0094436090225565, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 55032 + }, + { + "epoch": 3.009503759398496, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 55033 + }, + { + "epoch": 3.009563909774436, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 55034 + }, + { + "epoch": 3.009624060150376, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 55035 + }, + { + "epoch": 3.009684210526316, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 55036 + }, + { + "epoch": 3.0097443609022556, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 55037 + }, + { + "epoch": 3.0098045112781957, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0579, + "step": 55038 + }, + { + "epoch": 3.0098646616541354, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 55039 + }, + { + "epoch": 3.009924812030075, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 55040 + }, + { + "epoch": 3.009984962406015, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 55041 + }, + { + "epoch": 3.0100451127819547, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 55042 + }, + { + "epoch": 3.010105263157895, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 55043 + }, + { + "epoch": 3.0101654135338345, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 55044 + }, + { + "epoch": 3.0102255639097746, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 55045 + }, + { + "epoch": 3.0102857142857142, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 55046 + }, + { + "epoch": 3.0103458646616543, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 55047 + }, + { + "epoch": 3.010406015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9144, + "step": 55048 + }, + { + "epoch": 3.0104661654135336, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 55049 + }, + { + "epoch": 3.0105263157894737, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 55050 + }, + { + "epoch": 3.0105864661654134, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1333, + "step": 55051 + }, + { + "epoch": 3.0106466165413535, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 55052 + }, + { + "epoch": 3.010706766917293, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 55053 + }, + { + "epoch": 3.010766917293233, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0646, + "step": 55054 + }, + { + "epoch": 3.010827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 55055 + }, + { + "epoch": 3.010887218045113, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 55056 + }, + { + "epoch": 3.0109473684210526, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 55057 + }, + { + "epoch": 3.0110075187969927, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 55058 + }, + { + "epoch": 3.0110676691729323, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 55059 + }, + { + "epoch": 3.011127819548872, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 55060 + }, + { + "epoch": 3.011187969924812, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 55061 + }, + { + "epoch": 3.0112481203007517, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 55062 + }, + { + "epoch": 3.011308270676692, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 55063 + }, + { + "epoch": 3.0113684210526315, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 55064 + }, + { + "epoch": 3.0114285714285716, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 55065 + }, + { + "epoch": 3.011488721804511, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 55066 + }, + { + "epoch": 3.0115488721804513, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 55067 + }, + { + "epoch": 3.011609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 55068 + }, + { + "epoch": 3.0116691729323306, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 55069 + }, + { + "epoch": 3.0117293233082707, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 55070 + }, + { + "epoch": 3.0117894736842103, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 55071 + }, + { + "epoch": 3.0118496240601504, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 55072 + }, + { + "epoch": 3.01190977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1088, + "step": 55073 + }, + { + "epoch": 3.01196992481203, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 55074 + }, + { + "epoch": 3.01203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9219, + "step": 55075 + }, + { + "epoch": 3.01209022556391, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 55076 + }, + { + "epoch": 3.0121503759398496, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 55077 + }, + { + "epoch": 3.0122105263157897, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 55078 + }, + { + "epoch": 3.0122706766917293, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 55079 + }, + { + "epoch": 3.012330827067669, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9139, + "step": 55080 + }, + { + "epoch": 3.012390977443609, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 55081 + }, + { + "epoch": 3.0124511278195487, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 55082 + }, + { + "epoch": 3.012511278195489, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 55083 + }, + { + "epoch": 3.0125714285714285, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 55084 + }, + { + "epoch": 3.0126315789473685, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 55085 + }, + { + "epoch": 3.012691729323308, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.947, + "step": 55086 + }, + { + "epoch": 3.0127518796992483, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 55087 + }, + { + "epoch": 3.012812030075188, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9205, + "step": 55088 + }, + { + "epoch": 3.012872180451128, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 55089 + }, + { + "epoch": 3.0129323308270677, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 55090 + }, + { + "epoch": 3.0129924812030073, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1325, + "step": 55091 + }, + { + "epoch": 3.0130526315789474, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 55092 + }, + { + "epoch": 3.013112781954887, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 55093 + }, + { + "epoch": 3.013172932330827, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 55094 + }, + { + "epoch": 3.013233082706767, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 55095 + }, + { + "epoch": 3.013293233082707, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 55096 + }, + { + "epoch": 3.0133533834586466, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 55097 + }, + { + "epoch": 3.0134135338345867, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9443, + "step": 55098 + }, + { + "epoch": 3.0134736842105263, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 55099 + }, + { + "epoch": 3.013533834586466, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 55100 + }, + { + "epoch": 3.013593984962406, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 55101 + }, + { + "epoch": 3.0136541353383457, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 55102 + }, + { + "epoch": 3.013714285714286, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 55103 + }, + { + "epoch": 3.0137744360902254, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 55104 + }, + { + "epoch": 3.0138345864661655, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 55105 + }, + { + "epoch": 3.013894736842105, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 55106 + }, + { + "epoch": 3.0139548872180453, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 55107 + }, + { + "epoch": 3.014015037593985, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 55108 + }, + { + "epoch": 3.014075187969925, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 55109 + }, + { + "epoch": 3.0141353383458647, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 55110 + }, + { + "epoch": 3.0141954887218043, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 55111 + }, + { + "epoch": 3.0142556390977444, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 55112 + }, + { + "epoch": 3.014315789473684, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 55113 + }, + { + "epoch": 3.014375939849624, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 55114 + }, + { + "epoch": 3.014436090225564, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 55115 + }, + { + "epoch": 3.014496240601504, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 55116 + }, + { + "epoch": 3.0145563909774435, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 55117 + }, + { + "epoch": 3.0146165413533836, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 55118 + }, + { + "epoch": 3.0146766917293233, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 55119 + }, + { + "epoch": 3.014736842105263, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 55120 + }, + { + "epoch": 3.014796992481203, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 55121 + }, + { + "epoch": 3.0148571428571427, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 55122 + }, + { + "epoch": 3.0149172932330828, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 55123 + }, + { + "epoch": 3.0149774436090224, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 55124 + }, + { + "epoch": 3.0150375939849625, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9129, + "step": 55125 + }, + { + "epoch": 3.015097744360902, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 55126 + }, + { + "epoch": 3.0151578947368423, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 55127 + }, + { + "epoch": 3.015218045112782, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 55128 + }, + { + "epoch": 3.015278195488722, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 55129 + }, + { + "epoch": 3.0153383458646617, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 55130 + }, + { + "epoch": 3.0153984962406013, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9375, + "step": 55131 + }, + { + "epoch": 3.0154586466165414, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 55132 + }, + { + "epoch": 3.015518796992481, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 55133 + }, + { + "epoch": 3.015578947368421, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.957, + "step": 55134 + }, + { + "epoch": 3.015639097744361, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 55135 + }, + { + "epoch": 3.015699248120301, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 55136 + }, + { + "epoch": 3.0157593984962405, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 55137 + }, + { + "epoch": 3.0158195488721806, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 55138 + }, + { + "epoch": 3.0158796992481203, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 55139 + }, + { + "epoch": 3.0159398496240604, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9316, + "step": 55140 + }, + { + "epoch": 3.016, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.944, + "step": 55141 + }, + { + "epoch": 3.0160601503759397, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 55142 + }, + { + "epoch": 3.0161203007518798, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 55143 + }, + { + "epoch": 3.0161804511278194, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 55144 + }, + { + "epoch": 3.0162406015037595, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.93, + "step": 55145 + }, + { + "epoch": 3.016300751879699, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 55146 + }, + { + "epoch": 3.0163609022556392, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 55147 + }, + { + "epoch": 3.016421052631579, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 55148 + }, + { + "epoch": 3.016481203007519, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 55149 + }, + { + "epoch": 3.0165413533834586, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 55150 + }, + { + "epoch": 3.0166015037593983, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 55151 + }, + { + "epoch": 3.0166616541353384, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 55152 + }, + { + "epoch": 3.016721804511278, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 55153 + }, + { + "epoch": 3.016781954887218, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 55154 + }, + { + "epoch": 3.0168421052631578, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 55155 + }, + { + "epoch": 3.016902255639098, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 55156 + }, + { + "epoch": 3.0169624060150375, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 55157 + }, + { + "epoch": 3.0170225563909776, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 55158 + }, + { + "epoch": 3.0170827067669173, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9162, + "step": 55159 + }, + { + "epoch": 3.0171428571428573, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 55160 + }, + { + "epoch": 3.017203007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 55161 + }, + { + "epoch": 3.0172631578947366, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 55162 + }, + { + "epoch": 3.0173233082706767, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 55163 + }, + { + "epoch": 3.0173834586466164, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 55164 + }, + { + "epoch": 3.0174436090225565, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0814, + "step": 55165 + }, + { + "epoch": 3.017503759398496, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 55166 + }, + { + "epoch": 3.0175639097744362, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 55167 + }, + { + "epoch": 3.017624060150376, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 55168 + }, + { + "epoch": 3.017684210526316, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9125, + "step": 55169 + }, + { + "epoch": 3.0177443609022556, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 55170 + }, + { + "epoch": 3.0178045112781957, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 55171 + }, + { + "epoch": 3.0178646616541354, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 55172 + }, + { + "epoch": 3.017924812030075, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 55173 + }, + { + "epoch": 3.017984962406015, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 55174 + }, + { + "epoch": 3.0180451127819548, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 55175 + }, + { + "epoch": 3.018105263157895, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 55176 + }, + { + "epoch": 3.0181654135338345, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 55177 + }, + { + "epoch": 3.0182255639097746, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 55178 + }, + { + "epoch": 3.0182857142857142, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 55179 + }, + { + "epoch": 3.0183458646616543, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 55180 + }, + { + "epoch": 3.018406015037594, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 55181 + }, + { + "epoch": 3.0184661654135336, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 55182 + }, + { + "epoch": 3.0185263157894737, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 55183 + }, + { + "epoch": 3.0185864661654134, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 55184 + }, + { + "epoch": 3.0186466165413535, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 55185 + }, + { + "epoch": 3.018706766917293, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 55186 + }, + { + "epoch": 3.018766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 55187 + }, + { + "epoch": 3.018827067669173, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 55188 + }, + { + "epoch": 3.018887218045113, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 55189 + }, + { + "epoch": 3.0189473684210526, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 55190 + }, + { + "epoch": 3.0190075187969927, + "grad_norm": 0.4765625, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 55191 + }, + { + "epoch": 3.0190676691729323, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 55192 + }, + { + "epoch": 3.019127819548872, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 55193 + }, + { + "epoch": 3.019187969924812, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 55194 + }, + { + "epoch": 3.0192481203007517, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 55195 + }, + { + "epoch": 3.019308270676692, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 55196 + }, + { + "epoch": 3.0193684210526315, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 55197 + }, + { + "epoch": 3.0194285714285716, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 55198 + }, + { + "epoch": 3.0194887218045112, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 55199 + }, + { + "epoch": 3.0195488721804513, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.1233, + "step": 55200 + }, + { + "epoch": 3.019609022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 55201 + }, + { + "epoch": 3.0196691729323306, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 55202 + }, + { + "epoch": 3.0197293233082707, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9025, + "step": 55203 + }, + { + "epoch": 3.0197894736842104, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 55204 + }, + { + "epoch": 3.0198496240601505, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 55205 + }, + { + "epoch": 3.01990977443609, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 55206 + }, + { + "epoch": 3.01996992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 55207 + }, + { + "epoch": 3.02003007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 55208 + }, + { + "epoch": 3.02009022556391, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 55209 + }, + { + "epoch": 3.0201503759398496, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 55210 + }, + { + "epoch": 3.0202105263157897, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 55211 + }, + { + "epoch": 3.0202706766917293, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 55212 + }, + { + "epoch": 3.020330827067669, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 55213 + }, + { + "epoch": 3.020390977443609, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 55214 + }, + { + "epoch": 3.0204511278195487, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 55215 + }, + { + "epoch": 3.020511278195489, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 55216 + }, + { + "epoch": 3.0205714285714285, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 55217 + }, + { + "epoch": 3.0206315789473686, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 55218 + }, + { + "epoch": 3.020691729323308, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9501, + "step": 55219 + }, + { + "epoch": 3.0207518796992483, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 55220 + }, + { + "epoch": 3.020812030075188, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 55221 + }, + { + "epoch": 3.0208721804511276, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 55222 + }, + { + "epoch": 3.0209323308270677, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 55223 + }, + { + "epoch": 3.0209924812030073, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 55224 + }, + { + "epoch": 3.0210526315789474, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 55225 + }, + { + "epoch": 3.021112781954887, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9332, + "step": 55226 + }, + { + "epoch": 3.021172932330827, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 55227 + }, + { + "epoch": 3.021233082706767, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 55228 + }, + { + "epoch": 3.021293233082707, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 55229 + }, + { + "epoch": 3.0213533834586466, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9283, + "step": 55230 + }, + { + "epoch": 3.0214135338345867, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 55231 + }, + { + "epoch": 3.0214736842105263, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 55232 + }, + { + "epoch": 3.021533834586466, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 55233 + }, + { + "epoch": 3.021593984962406, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 55234 + }, + { + "epoch": 3.0216541353383457, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 55235 + }, + { + "epoch": 3.021714285714286, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 55236 + }, + { + "epoch": 3.0217744360902254, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 55237 + }, + { + "epoch": 3.0218345864661655, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 55238 + }, + { + "epoch": 3.021894736842105, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 55239 + }, + { + "epoch": 3.0219548872180453, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 55240 + }, + { + "epoch": 3.022015037593985, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 55241 + }, + { + "epoch": 3.022075187969925, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 55242 + }, + { + "epoch": 3.0221353383458647, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 55243 + }, + { + "epoch": 3.0221954887218043, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 55244 + }, + { + "epoch": 3.0222556390977444, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 55245 + }, + { + "epoch": 3.022315789473684, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.89, + "step": 55246 + }, + { + "epoch": 3.022375939849624, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9211, + "step": 55247 + }, + { + "epoch": 3.022436090225564, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 55248 + }, + { + "epoch": 3.022496240601504, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 55249 + }, + { + "epoch": 3.0225563909774436, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 55250 + }, + { + "epoch": 3.0226165413533836, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 55251 + }, + { + "epoch": 3.0226766917293233, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 55252 + }, + { + "epoch": 3.022736842105263, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 55253 + }, + { + "epoch": 3.022796992481203, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 55254 + }, + { + "epoch": 3.0228571428571427, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 55255 + }, + { + "epoch": 3.022917293233083, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 55256 + }, + { + "epoch": 3.0229774436090224, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 55257 + }, + { + "epoch": 3.0230375939849625, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 55258 + }, + { + "epoch": 3.023097744360902, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 55259 + }, + { + "epoch": 3.0231578947368423, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 55260 + }, + { + "epoch": 3.023218045112782, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 55261 + }, + { + "epoch": 3.023278195488722, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 55262 + }, + { + "epoch": 3.0233383458646617, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 55263 + }, + { + "epoch": 3.0233984962406013, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 55264 + }, + { + "epoch": 3.0234586466165414, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 55265 + }, + { + "epoch": 3.023518796992481, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9182, + "step": 55266 + }, + { + "epoch": 3.023578947368421, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9228, + "step": 55267 + }, + { + "epoch": 3.023639097744361, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 55268 + }, + { + "epoch": 3.023699248120301, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 55269 + }, + { + "epoch": 3.0237593984962405, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 55270 + }, + { + "epoch": 3.0238195488721806, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 55271 + }, + { + "epoch": 3.0238796992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 55272 + }, + { + "epoch": 3.0239398496240604, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 55273 + }, + { + "epoch": 3.024, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 55274 + }, + { + "epoch": 3.0240601503759397, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 55275 + }, + { + "epoch": 3.0241203007518798, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 55276 + }, + { + "epoch": 3.0241804511278194, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 55277 + }, + { + "epoch": 3.0242406015037595, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 55278 + }, + { + "epoch": 3.024300751879699, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 55279 + }, + { + "epoch": 3.0243609022556392, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 55280 + }, + { + "epoch": 3.024421052631579, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 55281 + }, + { + "epoch": 3.024481203007519, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0618, + "step": 55282 + }, + { + "epoch": 3.0245413533834586, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 55283 + }, + { + "epoch": 3.0246015037593983, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 55284 + }, + { + "epoch": 3.0246616541353384, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 55285 + }, + { + "epoch": 3.024721804511278, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 55286 + }, + { + "epoch": 3.024781954887218, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0739, + "step": 55287 + }, + { + "epoch": 3.0248421052631578, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 55288 + }, + { + "epoch": 3.024902255639098, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 55289 + }, + { + "epoch": 3.0249624060150375, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1376, + "step": 55290 + }, + { + "epoch": 3.0250225563909776, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 55291 + }, + { + "epoch": 3.0250827067669173, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 55292 + }, + { + "epoch": 3.0251428571428574, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 55293 + }, + { + "epoch": 3.025203007518797, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 55294 + }, + { + "epoch": 3.0252631578947367, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 55295 + }, + { + "epoch": 3.0253233082706767, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.8946, + "step": 55296 + }, + { + "epoch": 3.0253834586466164, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 55297 + }, + { + "epoch": 3.0254436090225565, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 55298 + }, + { + "epoch": 3.025503759398496, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 55299 + }, + { + "epoch": 3.0255639097744362, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 55300 + }, + { + "epoch": 3.025624060150376, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 55301 + }, + { + "epoch": 3.025684210526316, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 55302 + }, + { + "epoch": 3.0257443609022556, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 55303 + }, + { + "epoch": 3.0258045112781953, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 55304 + }, + { + "epoch": 3.0258646616541354, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 55305 + }, + { + "epoch": 3.025924812030075, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9051, + "step": 55306 + }, + { + "epoch": 3.025984962406015, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 55307 + }, + { + "epoch": 3.0260451127819548, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 55308 + }, + { + "epoch": 3.026105263157895, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 55309 + }, + { + "epoch": 3.0261654135338345, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 55310 + }, + { + "epoch": 3.0262255639097746, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9226, + "step": 55311 + }, + { + "epoch": 3.0262857142857142, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 55312 + }, + { + "epoch": 3.0263458646616543, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 55313 + }, + { + "epoch": 3.026406015037594, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 55314 + }, + { + "epoch": 3.0264661654135336, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 55315 + }, + { + "epoch": 3.0265263157894737, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 55316 + }, + { + "epoch": 3.0265864661654134, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 55317 + }, + { + "epoch": 3.0266466165413535, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 55318 + }, + { + "epoch": 3.026706766917293, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 55319 + }, + { + "epoch": 3.026766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 55320 + }, + { + "epoch": 3.026827067669173, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 55321 + }, + { + "epoch": 3.026887218045113, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 55322 + }, + { + "epoch": 3.0269473684210526, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 55323 + }, + { + "epoch": 3.0270075187969923, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9186, + "step": 55324 + }, + { + "epoch": 3.0270676691729324, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1016, + "step": 55325 + }, + { + "epoch": 3.027127819548872, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 55326 + }, + { + "epoch": 3.027187969924812, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 55327 + }, + { + "epoch": 3.0272481203007517, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 55328 + }, + { + "epoch": 3.027308270676692, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 55329 + }, + { + "epoch": 3.0273684210526315, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 55330 + }, + { + "epoch": 3.0274285714285716, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 55331 + }, + { + "epoch": 3.0274887218045112, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 55332 + }, + { + "epoch": 3.0275488721804513, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 55333 + }, + { + "epoch": 3.027609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 55334 + }, + { + "epoch": 3.0276691729323306, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 55335 + }, + { + "epoch": 3.0277293233082707, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 55336 + }, + { + "epoch": 3.0277894736842104, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 55337 + }, + { + "epoch": 3.0278496240601505, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 55338 + }, + { + "epoch": 3.02790977443609, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 55339 + }, + { + "epoch": 3.02796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9155, + "step": 55340 + }, + { + "epoch": 3.02803007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 55341 + }, + { + "epoch": 3.02809022556391, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 55342 + }, + { + "epoch": 3.0281503759398496, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 55343 + }, + { + "epoch": 3.0282105263157897, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 55344 + }, + { + "epoch": 3.0282706766917293, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 55345 + }, + { + "epoch": 3.028330827067669, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 55346 + }, + { + "epoch": 3.028390977443609, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 55347 + }, + { + "epoch": 3.0284511278195487, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 55348 + }, + { + "epoch": 3.028511278195489, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 55349 + }, + { + "epoch": 3.0285714285714285, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 55350 + }, + { + "epoch": 3.0286315789473686, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 55351 + }, + { + "epoch": 3.028691729323308, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 55352 + }, + { + "epoch": 3.0287518796992483, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 55353 + }, + { + "epoch": 3.028812030075188, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 55354 + }, + { + "epoch": 3.0288721804511276, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 55355 + }, + { + "epoch": 3.0289323308270677, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 55356 + }, + { + "epoch": 3.0289924812030073, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 55357 + }, + { + "epoch": 3.0290526315789474, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 55358 + }, + { + "epoch": 3.029112781954887, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 55359 + }, + { + "epoch": 3.029172932330827, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 55360 + }, + { + "epoch": 3.029233082706767, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 55361 + }, + { + "epoch": 3.029293233082707, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 55362 + }, + { + "epoch": 3.0293533834586466, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 55363 + }, + { + "epoch": 3.0294135338345867, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 55364 + }, + { + "epoch": 3.0294736842105263, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 55365 + }, + { + "epoch": 3.029533834586466, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 55366 + }, + { + "epoch": 3.029593984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 55367 + }, + { + "epoch": 3.0296541353383457, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 55368 + }, + { + "epoch": 3.029714285714286, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 55369 + }, + { + "epoch": 3.0297744360902255, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 55370 + }, + { + "epoch": 3.0298345864661655, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 55371 + }, + { + "epoch": 3.029894736842105, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 55372 + }, + { + "epoch": 3.0299548872180453, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 55373 + }, + { + "epoch": 3.030015037593985, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 55374 + }, + { + "epoch": 3.030075187969925, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 55375 + }, + { + "epoch": 3.0301353383458647, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 55376 + }, + { + "epoch": 3.0301954887218043, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 55377 + }, + { + "epoch": 3.0302556390977444, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9251, + "step": 55378 + }, + { + "epoch": 3.030315789473684, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9453, + "step": 55379 + }, + { + "epoch": 3.030375939849624, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 55380 + }, + { + "epoch": 3.030436090225564, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9315, + "step": 55381 + }, + { + "epoch": 3.030496240601504, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 55382 + }, + { + "epoch": 3.0305563909774436, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 55383 + }, + { + "epoch": 3.0306165413533837, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 55384 + }, + { + "epoch": 3.0306766917293233, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1275, + "step": 55385 + }, + { + "epoch": 3.030736842105263, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 55386 + }, + { + "epoch": 3.030796992481203, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 55387 + }, + { + "epoch": 3.0308571428571427, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9303, + "step": 55388 + }, + { + "epoch": 3.030917293233083, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 55389 + }, + { + "epoch": 3.0309774436090224, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 55390 + }, + { + "epoch": 3.0310375939849625, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 55391 + }, + { + "epoch": 3.031097744360902, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 55392 + }, + { + "epoch": 3.0311578947368423, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 55393 + }, + { + "epoch": 3.031218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 55394 + }, + { + "epoch": 3.031278195488722, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 55395 + }, + { + "epoch": 3.0313383458646617, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 55396 + }, + { + "epoch": 3.0313984962406013, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 55397 + }, + { + "epoch": 3.0314586466165414, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 55398 + }, + { + "epoch": 3.031518796992481, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 55399 + }, + { + "epoch": 3.031578947368421, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 55400 + }, + { + "epoch": 3.031639097744361, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 55401 + }, + { + "epoch": 3.031699248120301, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 55402 + }, + { + "epoch": 3.0317593984962405, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 55403 + }, + { + "epoch": 3.0318195488721806, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 55404 + }, + { + "epoch": 3.0318796992481203, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 55405 + }, + { + "epoch": 3.03193984962406, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 55406 + }, + { + "epoch": 3.032, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 55407 + }, + { + "epoch": 3.0320601503759397, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 55408 + }, + { + "epoch": 3.0321203007518798, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 55409 + }, + { + "epoch": 3.0321804511278194, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 55410 + }, + { + "epoch": 3.0322406015037595, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 55411 + }, + { + "epoch": 3.032300751879699, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 55412 + }, + { + "epoch": 3.0323609022556393, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.8788, + "step": 55413 + }, + { + "epoch": 3.032421052631579, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 55414 + }, + { + "epoch": 3.032481203007519, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 55415 + }, + { + "epoch": 3.0325413533834586, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1874, + "step": 55416 + }, + { + "epoch": 3.0326015037593983, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9366, + "step": 55417 + }, + { + "epoch": 3.0326616541353384, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 55418 + }, + { + "epoch": 3.032721804511278, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 55419 + }, + { + "epoch": 3.032781954887218, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1113, + "step": 55420 + }, + { + "epoch": 3.032842105263158, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 55421 + }, + { + "epoch": 3.032902255639098, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9296, + "step": 55422 + }, + { + "epoch": 3.0329624060150375, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 55423 + }, + { + "epoch": 3.0330225563909776, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 55424 + }, + { + "epoch": 3.0330827067669173, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 55425 + }, + { + "epoch": 3.033142857142857, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 55426 + }, + { + "epoch": 3.033203007518797, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 55427 + }, + { + "epoch": 3.0332631578947367, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 55428 + }, + { + "epoch": 3.0333233082706768, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 55429 + }, + { + "epoch": 3.0333834586466164, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 55430 + }, + { + "epoch": 3.0334436090225565, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 55431 + }, + { + "epoch": 3.033503759398496, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 55432 + }, + { + "epoch": 3.0335639097744362, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 55433 + }, + { + "epoch": 3.033624060150376, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 55434 + }, + { + "epoch": 3.033684210526316, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 55435 + }, + { + "epoch": 3.0337443609022556, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 55436 + }, + { + "epoch": 3.0338045112781953, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 55437 + }, + { + "epoch": 3.0338646616541354, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 55438 + }, + { + "epoch": 3.033924812030075, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 55439 + }, + { + "epoch": 3.033984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9294, + "step": 55440 + }, + { + "epoch": 3.0340451127819548, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 55441 + }, + { + "epoch": 3.034105263157895, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 55442 + }, + { + "epoch": 3.0341654135338345, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 55443 + }, + { + "epoch": 3.0342255639097746, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 55444 + }, + { + "epoch": 3.0342857142857143, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.948, + "step": 55445 + }, + { + "epoch": 3.0343458646616543, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 55446 + }, + { + "epoch": 3.034406015037594, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 55447 + }, + { + "epoch": 3.0344661654135336, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 55448 + }, + { + "epoch": 3.0345263157894737, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 55449 + }, + { + "epoch": 3.0345864661654134, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 55450 + }, + { + "epoch": 3.0346466165413535, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 55451 + }, + { + "epoch": 3.034706766917293, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 55452 + }, + { + "epoch": 3.0347669172932332, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 55453 + }, + { + "epoch": 3.034827067669173, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 55454 + }, + { + "epoch": 3.034887218045113, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0834, + "step": 55455 + }, + { + "epoch": 3.0349473684210526, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 55456 + }, + { + "epoch": 3.0350075187969923, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 55457 + }, + { + "epoch": 3.0350676691729324, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.8998, + "step": 55458 + }, + { + "epoch": 3.035127819548872, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 55459 + }, + { + "epoch": 3.035187969924812, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 55460 + }, + { + "epoch": 3.0352481203007518, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 55461 + }, + { + "epoch": 3.035308270676692, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 55462 + }, + { + "epoch": 3.0353684210526315, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 55463 + }, + { + "epoch": 3.0354285714285716, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9215, + "step": 55464 + }, + { + "epoch": 3.0354887218045112, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 55465 + }, + { + "epoch": 3.0355488721804513, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 55466 + }, + { + "epoch": 3.035609022556391, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 55467 + }, + { + "epoch": 3.0356691729323306, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9217, + "step": 55468 + }, + { + "epoch": 3.0357293233082707, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 55469 + }, + { + "epoch": 3.0357894736842104, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9282, + "step": 55470 + }, + { + "epoch": 3.0358496240601505, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 55471 + }, + { + "epoch": 3.03590977443609, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 55472 + }, + { + "epoch": 3.03596992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 55473 + }, + { + "epoch": 3.03603007518797, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 55474 + }, + { + "epoch": 3.03609022556391, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 55475 + }, + { + "epoch": 3.0361503759398496, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1128, + "step": 55476 + }, + { + "epoch": 3.0362105263157897, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 55477 + }, + { + "epoch": 3.0362706766917293, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 55478 + }, + { + "epoch": 3.036330827067669, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 55479 + }, + { + "epoch": 3.036390977443609, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 55480 + }, + { + "epoch": 3.0364511278195487, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 55481 + }, + { + "epoch": 3.036511278195489, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 55482 + }, + { + "epoch": 3.0365714285714285, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 55483 + }, + { + "epoch": 3.0366315789473686, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 55484 + }, + { + "epoch": 3.036691729323308, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 55485 + }, + { + "epoch": 3.0367518796992483, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 55486 + }, + { + "epoch": 3.036812030075188, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 55487 + }, + { + "epoch": 3.0368721804511276, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 55488 + }, + { + "epoch": 3.0369323308270677, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9386, + "step": 55489 + }, + { + "epoch": 3.0369924812030074, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 55490 + }, + { + "epoch": 3.0370526315789474, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 55491 + }, + { + "epoch": 3.037112781954887, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 55492 + }, + { + "epoch": 3.037172932330827, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 55493 + }, + { + "epoch": 3.037233082706767, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 55494 + }, + { + "epoch": 3.037293233082707, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.904, + "step": 55495 + }, + { + "epoch": 3.0373533834586466, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1044, + "step": 55496 + }, + { + "epoch": 3.0374135338345867, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 55497 + }, + { + "epoch": 3.0374736842105263, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 55498 + }, + { + "epoch": 3.037533834586466, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 55499 + }, + { + "epoch": 3.037593984962406, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1715, + "step": 55500 + }, + { + "epoch": 3.0376541353383457, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 55501 + }, + { + "epoch": 3.037714285714286, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9317, + "step": 55502 + }, + { + "epoch": 3.0377744360902255, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 55503 + }, + { + "epoch": 3.0378345864661656, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 55504 + }, + { + "epoch": 3.037894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 55505 + }, + { + "epoch": 3.0379548872180453, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9125, + "step": 55506 + }, + { + "epoch": 3.038015037593985, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9362, + "step": 55507 + }, + { + "epoch": 3.0380751879699246, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 55508 + }, + { + "epoch": 3.0381353383458647, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 55509 + }, + { + "epoch": 3.0381954887218043, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 55510 + }, + { + "epoch": 3.0382556390977444, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 55511 + }, + { + "epoch": 3.038315789473684, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 55512 + }, + { + "epoch": 3.038375939849624, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 55513 + }, + { + "epoch": 3.038436090225564, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 55514 + }, + { + "epoch": 3.038496240601504, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 55515 + }, + { + "epoch": 3.0385563909774436, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 55516 + }, + { + "epoch": 3.0386165413533837, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 55517 + }, + { + "epoch": 3.0386766917293233, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 55518 + }, + { + "epoch": 3.038736842105263, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 55519 + }, + { + "epoch": 3.038796992481203, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 55520 + }, + { + "epoch": 3.0388571428571427, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 55521 + }, + { + "epoch": 3.038917293233083, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9342, + "step": 55522 + }, + { + "epoch": 3.0389774436090224, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 55523 + }, + { + "epoch": 3.0390375939849625, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 55524 + }, + { + "epoch": 3.039097744360902, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 55525 + }, + { + "epoch": 3.0391578947368423, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 55526 + }, + { + "epoch": 3.039218045112782, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 55527 + }, + { + "epoch": 3.039278195488722, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 55528 + }, + { + "epoch": 3.0393383458646617, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 55529 + }, + { + "epoch": 3.0393984962406013, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 55530 + }, + { + "epoch": 3.0394586466165414, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 55531 + }, + { + "epoch": 3.039518796992481, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 55532 + }, + { + "epoch": 3.039578947368421, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 55533 + }, + { + "epoch": 3.039639097744361, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 55534 + }, + { + "epoch": 3.039699248120301, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 55535 + }, + { + "epoch": 3.0397593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 55536 + }, + { + "epoch": 3.0398195488721806, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 55537 + }, + { + "epoch": 3.0398796992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1151, + "step": 55538 + }, + { + "epoch": 3.03993984962406, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 55539 + }, + { + "epoch": 3.04, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0805, + "step": 55540 + }, + { + "epoch": 3.0400601503759397, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 55541 + }, + { + "epoch": 3.04012030075188, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 55542 + }, + { + "epoch": 3.0401804511278194, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 55543 + }, + { + "epoch": 3.0402406015037595, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 55544 + }, + { + "epoch": 3.040300751879699, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 55545 + }, + { + "epoch": 3.0403609022556393, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 55546 + }, + { + "epoch": 3.040421052631579, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9459, + "step": 55547 + }, + { + "epoch": 3.040481203007519, + "grad_norm": 1.34375, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 55548 + }, + { + "epoch": 3.0405413533834587, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 55549 + }, + { + "epoch": 3.0406015037593983, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 55550 + }, + { + "epoch": 3.0406616541353384, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0967, + "step": 55551 + }, + { + "epoch": 3.040721804511278, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9322, + "step": 55552 + }, + { + "epoch": 3.040781954887218, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 55553 + }, + { + "epoch": 3.040842105263158, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9451, + "step": 55554 + }, + { + "epoch": 3.040902255639098, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 55555 + }, + { + "epoch": 3.0409624060150375, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9254, + "step": 55556 + }, + { + "epoch": 3.0410225563909776, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 55557 + }, + { + "epoch": 3.0410827067669173, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 55558 + }, + { + "epoch": 3.041142857142857, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 55559 + }, + { + "epoch": 3.041203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 55560 + }, + { + "epoch": 3.0412631578947367, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 55561 + }, + { + "epoch": 3.0413233082706768, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 55562 + }, + { + "epoch": 3.0413834586466164, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 55563 + }, + { + "epoch": 3.0414436090225565, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 55564 + }, + { + "epoch": 3.041503759398496, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9058, + "step": 55565 + }, + { + "epoch": 3.0415639097744362, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 55566 + }, + { + "epoch": 3.041624060150376, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 55567 + }, + { + "epoch": 3.041684210526316, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 55568 + }, + { + "epoch": 3.0417443609022556, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 55569 + }, + { + "epoch": 3.0418045112781953, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1413, + "step": 55570 + }, + { + "epoch": 3.0418646616541354, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 55571 + }, + { + "epoch": 3.041924812030075, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 55572 + }, + { + "epoch": 3.041984962406015, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 55573 + }, + { + "epoch": 3.0420451127819548, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 55574 + }, + { + "epoch": 3.042105263157895, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 55575 + }, + { + "epoch": 3.0421654135338345, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 55576 + }, + { + "epoch": 3.0422255639097746, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 55577 + }, + { + "epoch": 3.0422857142857143, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 55578 + }, + { + "epoch": 3.0423458646616544, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 55579 + }, + { + "epoch": 3.042406015037594, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0916, + "step": 55580 + }, + { + "epoch": 3.0424661654135337, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 55581 + }, + { + "epoch": 3.0425263157894737, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 55582 + }, + { + "epoch": 3.0425864661654134, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 55583 + }, + { + "epoch": 3.0426466165413535, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 55584 + }, + { + "epoch": 3.042706766917293, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 55585 + }, + { + "epoch": 3.0427669172932332, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 55586 + }, + { + "epoch": 3.042827067669173, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 55587 + }, + { + "epoch": 3.042887218045113, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 55588 + }, + { + "epoch": 3.0429473684210526, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 55589 + }, + { + "epoch": 3.0430075187969923, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 55590 + }, + { + "epoch": 3.0430676691729324, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 55591 + }, + { + "epoch": 3.043127819548872, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 55592 + }, + { + "epoch": 3.043187969924812, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 55593 + }, + { + "epoch": 3.0432481203007518, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9193, + "step": 55594 + }, + { + "epoch": 3.043308270676692, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 55595 + }, + { + "epoch": 3.0433684210526315, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 55596 + }, + { + "epoch": 3.0434285714285716, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 55597 + }, + { + "epoch": 3.0434887218045112, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 55598 + }, + { + "epoch": 3.0435488721804513, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 55599 + }, + { + "epoch": 3.043609022556391, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9111, + "step": 55600 + }, + { + "epoch": 3.0436691729323306, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 55601 + }, + { + "epoch": 3.0437293233082707, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 55602 + }, + { + "epoch": 3.0437894736842104, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 55603 + }, + { + "epoch": 3.0438496240601505, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 55604 + }, + { + "epoch": 3.04390977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 55605 + }, + { + "epoch": 3.04396992481203, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 55606 + }, + { + "epoch": 3.04403007518797, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 55607 + }, + { + "epoch": 3.04409022556391, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 55608 + }, + { + "epoch": 3.0441503759398496, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.1341, + "step": 55609 + }, + { + "epoch": 3.0442105263157893, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.8747, + "step": 55610 + }, + { + "epoch": 3.0442706766917293, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 55611 + }, + { + "epoch": 3.044330827067669, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 55612 + }, + { + "epoch": 3.044390977443609, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 55613 + }, + { + "epoch": 3.0444511278195487, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 55614 + }, + { + "epoch": 3.044511278195489, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1176, + "step": 55615 + }, + { + "epoch": 3.0445714285714285, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 55616 + }, + { + "epoch": 3.0446315789473686, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 55617 + }, + { + "epoch": 3.0446917293233082, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 55618 + }, + { + "epoch": 3.0447518796992483, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.899, + "step": 55619 + }, + { + "epoch": 3.044812030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9039, + "step": 55620 + }, + { + "epoch": 3.0448721804511276, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.8786, + "step": 55621 + }, + { + "epoch": 3.0449323308270677, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 55622 + }, + { + "epoch": 3.0449924812030074, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 55623 + }, + { + "epoch": 3.0450526315789475, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 55624 + }, + { + "epoch": 3.045112781954887, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9433, + "step": 55625 + }, + { + "epoch": 3.045172932330827, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.8946, + "step": 55626 + }, + { + "epoch": 3.045233082706767, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 55627 + }, + { + "epoch": 3.045293233082707, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 55628 + }, + { + "epoch": 3.0453533834586466, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.2212, + "step": 55629 + }, + { + "epoch": 3.0454135338345867, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 55630 + }, + { + "epoch": 3.0454736842105263, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9112, + "step": 55631 + }, + { + "epoch": 3.045533834586466, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 55632 + }, + { + "epoch": 3.045593984962406, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 55633 + }, + { + "epoch": 3.0456541353383457, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 55634 + }, + { + "epoch": 3.045714285714286, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 55635 + }, + { + "epoch": 3.0457744360902255, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 55636 + }, + { + "epoch": 3.0458345864661656, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 55637 + }, + { + "epoch": 3.045894736842105, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 55638 + }, + { + "epoch": 3.0459548872180453, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 55639 + }, + { + "epoch": 3.046015037593985, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 55640 + }, + { + "epoch": 3.0460751879699246, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 55641 + }, + { + "epoch": 3.0461353383458647, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 55642 + }, + { + "epoch": 3.0461954887218043, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 55643 + }, + { + "epoch": 3.0462556390977444, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 55644 + }, + { + "epoch": 3.046315789473684, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 55645 + }, + { + "epoch": 3.046375939849624, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 55646 + }, + { + "epoch": 3.046436090225564, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 55647 + }, + { + "epoch": 3.046496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 55648 + }, + { + "epoch": 3.0465563909774436, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 55649 + }, + { + "epoch": 3.0466165413533837, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 55650 + }, + { + "epoch": 3.0466766917293233, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 55651 + }, + { + "epoch": 3.046736842105263, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 55652 + }, + { + "epoch": 3.046796992481203, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 55653 + }, + { + "epoch": 3.0468571428571427, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 55654 + }, + { + "epoch": 3.046917293233083, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 55655 + }, + { + "epoch": 3.0469774436090225, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 55656 + }, + { + "epoch": 3.0470375939849625, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 55657 + }, + { + "epoch": 3.047097744360902, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9281, + "step": 55658 + }, + { + "epoch": 3.0471578947368423, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 55659 + }, + { + "epoch": 3.047218045112782, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9342, + "step": 55660 + }, + { + "epoch": 3.0472781954887216, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1204, + "step": 55661 + }, + { + "epoch": 3.0473383458646617, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 55662 + }, + { + "epoch": 3.0473984962406013, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 55663 + }, + { + "epoch": 3.0474586466165414, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 55664 + }, + { + "epoch": 3.047518796992481, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 55665 + }, + { + "epoch": 3.047578947368421, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 55666 + }, + { + "epoch": 3.047639097744361, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1456, + "step": 55667 + }, + { + "epoch": 3.047699248120301, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 55668 + }, + { + "epoch": 3.0477593984962406, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9208, + "step": 55669 + }, + { + "epoch": 3.0478195488721807, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 55670 + }, + { + "epoch": 3.0478796992481203, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 55671 + }, + { + "epoch": 3.04793984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 55672 + }, + { + "epoch": 3.048, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9447, + "step": 55673 + }, + { + "epoch": 3.0480601503759397, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 55674 + }, + { + "epoch": 3.04812030075188, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 55675 + }, + { + "epoch": 3.0481804511278194, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 55676 + }, + { + "epoch": 3.0482406015037595, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 55677 + }, + { + "epoch": 3.048300751879699, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9497, + "step": 55678 + }, + { + "epoch": 3.0483609022556393, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 55679 + }, + { + "epoch": 3.048421052631579, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 55680 + }, + { + "epoch": 3.048481203007519, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.8999, + "step": 55681 + }, + { + "epoch": 3.0485413533834587, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 55682 + }, + { + "epoch": 3.0486015037593983, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 55683 + }, + { + "epoch": 3.0486616541353384, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 55684 + }, + { + "epoch": 3.048721804511278, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 55685 + }, + { + "epoch": 3.048781954887218, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 55686 + }, + { + "epoch": 3.048842105263158, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9258, + "step": 55687 + }, + { + "epoch": 3.048902255639098, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 55688 + }, + { + "epoch": 3.0489624060150375, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 55689 + }, + { + "epoch": 3.0490225563909776, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 55690 + }, + { + "epoch": 3.0490827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 55691 + }, + { + "epoch": 3.049142857142857, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9206, + "step": 55692 + }, + { + "epoch": 3.049203007518797, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 55693 + }, + { + "epoch": 3.0492631578947367, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 55694 + }, + { + "epoch": 3.0493233082706768, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 55695 + }, + { + "epoch": 3.0493834586466164, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9344, + "step": 55696 + }, + { + "epoch": 3.0494436090225565, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9222, + "step": 55697 + }, + { + "epoch": 3.049503759398496, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9119, + "step": 55698 + }, + { + "epoch": 3.0495639097744363, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 55699 + }, + { + "epoch": 3.049624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 55700 + }, + { + "epoch": 3.049684210526316, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 55701 + }, + { + "epoch": 3.0497443609022556, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0631, + "step": 55702 + }, + { + "epoch": 3.0498045112781953, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 55703 + }, + { + "epoch": 3.0498646616541354, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 55704 + }, + { + "epoch": 3.049924812030075, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 55705 + }, + { + "epoch": 3.049984962406015, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 55706 + }, + { + "epoch": 3.050045112781955, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 55707 + }, + { + "epoch": 3.050105263157895, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 55708 + }, + { + "epoch": 3.0501654135338345, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 55709 + }, + { + "epoch": 3.0502255639097746, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 55710 + }, + { + "epoch": 3.0502857142857143, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 55711 + }, + { + "epoch": 3.0503458646616544, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9332, + "step": 55712 + }, + { + "epoch": 3.050406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 55713 + }, + { + "epoch": 3.0504661654135337, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 55714 + }, + { + "epoch": 3.0505263157894738, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1252, + "step": 55715 + }, + { + "epoch": 3.0505864661654134, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 55716 + }, + { + "epoch": 3.0506466165413535, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 55717 + }, + { + "epoch": 3.050706766917293, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 55718 + }, + { + "epoch": 3.0507669172932332, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0829, + "step": 55719 + }, + { + "epoch": 3.050827067669173, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 55720 + }, + { + "epoch": 3.050887218045113, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 55721 + }, + { + "epoch": 3.0509473684210526, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 55722 + }, + { + "epoch": 3.0510075187969923, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0965, + "step": 55723 + }, + { + "epoch": 3.0510676691729324, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 55724 + }, + { + "epoch": 3.051127819548872, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 55725 + }, + { + "epoch": 3.051187969924812, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 55726 + }, + { + "epoch": 3.0512481203007518, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 55727 + }, + { + "epoch": 3.051308270676692, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 55728 + }, + { + "epoch": 3.0513684210526315, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1054, + "step": 55729 + }, + { + "epoch": 3.0514285714285716, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 55730 + }, + { + "epoch": 3.0514887218045113, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 55731 + }, + { + "epoch": 3.0515488721804513, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.8966, + "step": 55732 + }, + { + "epoch": 3.051609022556391, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 55733 + }, + { + "epoch": 3.0516691729323306, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 55734 + }, + { + "epoch": 3.0517293233082707, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 55735 + }, + { + "epoch": 3.0517894736842104, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 55736 + }, + { + "epoch": 3.0518496240601505, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 55737 + }, + { + "epoch": 3.05190977443609, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 55738 + }, + { + "epoch": 3.05196992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 55739 + }, + { + "epoch": 3.05203007518797, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 55740 + }, + { + "epoch": 3.05209022556391, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 55741 + }, + { + "epoch": 3.0521503759398496, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 55742 + }, + { + "epoch": 3.0522105263157893, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 55743 + }, + { + "epoch": 3.0522706766917294, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 55744 + }, + { + "epoch": 3.052330827067669, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 55745 + }, + { + "epoch": 3.052390977443609, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 55746 + }, + { + "epoch": 3.0524511278195487, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 55747 + }, + { + "epoch": 3.052511278195489, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 55748 + }, + { + "epoch": 3.0525714285714285, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 55749 + }, + { + "epoch": 3.0526315789473686, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 55750 + }, + { + "epoch": 3.0526917293233082, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1812, + "step": 55751 + }, + { + "epoch": 3.0527518796992483, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 55752 + }, + { + "epoch": 3.052812030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 55753 + }, + { + "epoch": 3.0528721804511276, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 55754 + }, + { + "epoch": 3.0529323308270677, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 55755 + }, + { + "epoch": 3.0529924812030074, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 55756 + }, + { + "epoch": 3.0530526315789475, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 55757 + }, + { + "epoch": 3.053112781954887, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 55758 + }, + { + "epoch": 3.053172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 55759 + }, + { + "epoch": 3.053233082706767, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 55760 + }, + { + "epoch": 3.053293233082707, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 55761 + }, + { + "epoch": 3.0533533834586466, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 55762 + }, + { + "epoch": 3.0534135338345862, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1648, + "step": 55763 + }, + { + "epoch": 3.0534736842105263, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 55764 + }, + { + "epoch": 3.053533834586466, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 55765 + }, + { + "epoch": 3.053593984962406, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 55766 + }, + { + "epoch": 3.0536541353383457, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1182, + "step": 55767 + }, + { + "epoch": 3.053714285714286, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 55768 + }, + { + "epoch": 3.0537744360902255, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 55769 + }, + { + "epoch": 3.0538345864661656, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1103, + "step": 55770 + }, + { + "epoch": 3.053894736842105, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 55771 + }, + { + "epoch": 3.0539548872180453, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 55772 + }, + { + "epoch": 3.054015037593985, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.948, + "step": 55773 + }, + { + "epoch": 3.0540751879699246, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 55774 + }, + { + "epoch": 3.0541353383458647, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 55775 + }, + { + "epoch": 3.0541954887218044, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 55776 + }, + { + "epoch": 3.0542556390977444, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 55777 + }, + { + "epoch": 3.054315789473684, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9351, + "step": 55778 + }, + { + "epoch": 3.054375939849624, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 55779 + }, + { + "epoch": 3.054436090225564, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 55780 + }, + { + "epoch": 3.054496240601504, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 55781 + }, + { + "epoch": 3.0545563909774436, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9293, + "step": 55782 + }, + { + "epoch": 3.0546165413533837, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 55783 + }, + { + "epoch": 3.0546766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 55784 + }, + { + "epoch": 3.054736842105263, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0715, + "step": 55785 + }, + { + "epoch": 3.054796992481203, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 55786 + }, + { + "epoch": 3.0548571428571427, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 55787 + }, + { + "epoch": 3.054917293233083, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 55788 + }, + { + "epoch": 3.0549774436090225, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9364, + "step": 55789 + }, + { + "epoch": 3.0550375939849626, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 55790 + }, + { + "epoch": 3.055097744360902, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.938, + "step": 55791 + }, + { + "epoch": 3.0551578947368423, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.8986, + "step": 55792 + }, + { + "epoch": 3.055218045112782, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 55793 + }, + { + "epoch": 3.0552781954887216, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 55794 + }, + { + "epoch": 3.0553383458646617, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 55795 + }, + { + "epoch": 3.0553984962406013, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 55796 + }, + { + "epoch": 3.0554586466165414, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 55797 + }, + { + "epoch": 3.055518796992481, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 55798 + }, + { + "epoch": 3.055578947368421, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 55799 + }, + { + "epoch": 3.055639097744361, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 55800 + }, + { + "epoch": 3.055699248120301, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 55801 + }, + { + "epoch": 3.0557593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 55802 + }, + { + "epoch": 3.0558195488721807, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 55803 + }, + { + "epoch": 3.0558796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 55804 + }, + { + "epoch": 3.05593984962406, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 55805 + }, + { + "epoch": 3.056, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 55806 + }, + { + "epoch": 3.0560601503759397, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 55807 + }, + { + "epoch": 3.05612030075188, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 55808 + }, + { + "epoch": 3.0561804511278194, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 55809 + }, + { + "epoch": 3.0562406015037595, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9412, + "step": 55810 + }, + { + "epoch": 3.056300751879699, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 55811 + }, + { + "epoch": 3.0563609022556393, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 55812 + }, + { + "epoch": 3.056421052631579, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 55813 + }, + { + "epoch": 3.056481203007519, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 55814 + }, + { + "epoch": 3.0565413533834587, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 55815 + }, + { + "epoch": 3.0566015037593983, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.916, + "step": 55816 + }, + { + "epoch": 3.0566616541353384, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 55817 + }, + { + "epoch": 3.056721804511278, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 55818 + }, + { + "epoch": 3.056781954887218, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 55819 + }, + { + "epoch": 3.056842105263158, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 55820 + }, + { + "epoch": 3.056902255639098, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 55821 + }, + { + "epoch": 3.0569624060150375, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 55822 + }, + { + "epoch": 3.0570225563909776, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 55823 + }, + { + "epoch": 3.0570827067669173, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 55824 + }, + { + "epoch": 3.057142857142857, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 55825 + }, + { + "epoch": 3.057203007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 55826 + }, + { + "epoch": 3.0572631578947367, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 55827 + }, + { + "epoch": 3.0573233082706768, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 55828 + }, + { + "epoch": 3.0573834586466164, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 55829 + }, + { + "epoch": 3.0574436090225565, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 55830 + }, + { + "epoch": 3.057503759398496, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9444, + "step": 55831 + }, + { + "epoch": 3.0575639097744363, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 55832 + }, + { + "epoch": 3.057624060150376, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 55833 + }, + { + "epoch": 3.057684210526316, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9309, + "step": 55834 + }, + { + "epoch": 3.0577443609022557, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 55835 + }, + { + "epoch": 3.0578045112781953, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 55836 + }, + { + "epoch": 3.0578646616541354, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 55837 + }, + { + "epoch": 3.057924812030075, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 55838 + }, + { + "epoch": 3.057984962406015, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 55839 + }, + { + "epoch": 3.058045112781955, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 55840 + }, + { + "epoch": 3.058105263157895, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 55841 + }, + { + "epoch": 3.0581654135338345, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 55842 + }, + { + "epoch": 3.0582255639097746, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 55843 + }, + { + "epoch": 3.0582857142857143, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 55844 + }, + { + "epoch": 3.058345864661654, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 55845 + }, + { + "epoch": 3.058406015037594, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.944, + "step": 55846 + }, + { + "epoch": 3.0584661654135337, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9499, + "step": 55847 + }, + { + "epoch": 3.0585263157894738, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 55848 + }, + { + "epoch": 3.0585864661654134, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9529, + "step": 55849 + }, + { + "epoch": 3.0586466165413535, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9153, + "step": 55850 + }, + { + "epoch": 3.058706766917293, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 55851 + }, + { + "epoch": 3.0587669172932332, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 55852 + }, + { + "epoch": 3.058827067669173, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 55853 + }, + { + "epoch": 3.058887218045113, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 55854 + }, + { + "epoch": 3.0589473684210526, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 55855 + }, + { + "epoch": 3.0590075187969923, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 55856 + }, + { + "epoch": 3.0590676691729324, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 55857 + }, + { + "epoch": 3.059127819548872, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 55858 + }, + { + "epoch": 3.059187969924812, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 55859 + }, + { + "epoch": 3.0592481203007518, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9434, + "step": 55860 + }, + { + "epoch": 3.059308270676692, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 55861 + }, + { + "epoch": 3.0593684210526315, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 55862 + }, + { + "epoch": 3.0594285714285716, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 55863 + }, + { + "epoch": 3.0594887218045113, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 55864 + }, + { + "epoch": 3.059548872180451, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9369, + "step": 55865 + }, + { + "epoch": 3.059609022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 55866 + }, + { + "epoch": 3.0596691729323306, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 55867 + }, + { + "epoch": 3.0597293233082707, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 55868 + }, + { + "epoch": 3.0597894736842104, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 55869 + }, + { + "epoch": 3.0598496240601505, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 55870 + }, + { + "epoch": 3.05990977443609, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 55871 + }, + { + "epoch": 3.0599699248120302, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 55872 + }, + { + "epoch": 3.06003007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 55873 + }, + { + "epoch": 3.06009022556391, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 55874 + }, + { + "epoch": 3.0601503759398496, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 55875 + }, + { + "epoch": 3.0602105263157893, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 55876 + }, + { + "epoch": 3.0602706766917294, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9121, + "step": 55877 + }, + { + "epoch": 3.060330827067669, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9504, + "step": 55878 + }, + { + "epoch": 3.060390977443609, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 55879 + }, + { + "epoch": 3.0604511278195488, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 55880 + }, + { + "epoch": 3.060511278195489, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9416, + "step": 55881 + }, + { + "epoch": 3.0605714285714285, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 55882 + }, + { + "epoch": 3.0606315789473686, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9325, + "step": 55883 + }, + { + "epoch": 3.0606917293233082, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 55884 + }, + { + "epoch": 3.0607518796992483, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 55885 + }, + { + "epoch": 3.060812030075188, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 55886 + }, + { + "epoch": 3.0608721804511276, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 55887 + }, + { + "epoch": 3.0609323308270677, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 55888 + }, + { + "epoch": 3.0609924812030074, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 55889 + }, + { + "epoch": 3.0610526315789475, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 55890 + }, + { + "epoch": 3.061112781954887, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 55891 + }, + { + "epoch": 3.061172932330827, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 55892 + }, + { + "epoch": 3.061233082706767, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 55893 + }, + { + "epoch": 3.061293233082707, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 55894 + }, + { + "epoch": 3.0613533834586466, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 55895 + }, + { + "epoch": 3.0614135338345863, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 55896 + }, + { + "epoch": 3.0614736842105263, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 55897 + }, + { + "epoch": 3.061533834586466, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 55898 + }, + { + "epoch": 3.061593984962406, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 55899 + }, + { + "epoch": 3.0616541353383457, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.8994, + "step": 55900 + }, + { + "epoch": 3.061714285714286, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 55901 + }, + { + "epoch": 3.0617744360902255, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 55902 + }, + { + "epoch": 3.0618345864661656, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 55903 + }, + { + "epoch": 3.0618947368421052, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 55904 + }, + { + "epoch": 3.0619548872180453, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 55905 + }, + { + "epoch": 3.062015037593985, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 55906 + }, + { + "epoch": 3.0620751879699246, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 55907 + }, + { + "epoch": 3.0621353383458647, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 55908 + }, + { + "epoch": 3.0621954887218044, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 55909 + }, + { + "epoch": 3.0622556390977445, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 55910 + }, + { + "epoch": 3.062315789473684, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 55911 + }, + { + "epoch": 3.062375939849624, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 55912 + }, + { + "epoch": 3.062436090225564, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 55913 + }, + { + "epoch": 3.062496240601504, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1047, + "step": 55914 + }, + { + "epoch": 3.0625563909774436, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 55915 + }, + { + "epoch": 3.0626165413533837, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 55916 + }, + { + "epoch": 3.0626766917293233, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 55917 + }, + { + "epoch": 3.062736842105263, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 55918 + }, + { + "epoch": 3.062796992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9194, + "step": 55919 + }, + { + "epoch": 3.0628571428571427, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 55920 + }, + { + "epoch": 3.062917293233083, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 55921 + }, + { + "epoch": 3.0629774436090225, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 55922 + }, + { + "epoch": 3.0630375939849626, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 55923 + }, + { + "epoch": 3.063097744360902, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 55924 + }, + { + "epoch": 3.0631578947368423, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 55925 + }, + { + "epoch": 3.063218045112782, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 55926 + }, + { + "epoch": 3.0632781954887216, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 55927 + }, + { + "epoch": 3.0633383458646617, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 55928 + }, + { + "epoch": 3.0633984962406013, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 55929 + }, + { + "epoch": 3.0634586466165414, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 55930 + }, + { + "epoch": 3.063518796992481, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 55931 + }, + { + "epoch": 3.063578947368421, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 55932 + }, + { + "epoch": 3.063639097744361, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 55933 + }, + { + "epoch": 3.063699248120301, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 55934 + }, + { + "epoch": 3.0637593984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 55935 + }, + { + "epoch": 3.0638195488721807, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 55936 + }, + { + "epoch": 3.0638796992481203, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 55937 + }, + { + "epoch": 3.06393984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 55938 + }, + { + "epoch": 3.064, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 55939 + }, + { + "epoch": 3.0640601503759397, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 55940 + }, + { + "epoch": 3.06412030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 55941 + }, + { + "epoch": 3.0641804511278194, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 55942 + }, + { + "epoch": 3.0642406015037595, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1348, + "step": 55943 + }, + { + "epoch": 3.064300751879699, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 55944 + }, + { + "epoch": 3.0643609022556393, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 55945 + }, + { + "epoch": 3.064421052631579, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 55946 + }, + { + "epoch": 3.0644812030075186, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9204, + "step": 55947 + }, + { + "epoch": 3.0645413533834587, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 55948 + }, + { + "epoch": 3.0646015037593983, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9456, + "step": 55949 + }, + { + "epoch": 3.0646616541353384, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1466, + "step": 55950 + }, + { + "epoch": 3.064721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9354, + "step": 55951 + }, + { + "epoch": 3.064781954887218, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.8941, + "step": 55952 + }, + { + "epoch": 3.064842105263158, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 55953 + }, + { + "epoch": 3.064902255639098, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 55954 + }, + { + "epoch": 3.0649624060150376, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 55955 + }, + { + "epoch": 3.0650225563909776, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9249, + "step": 55956 + }, + { + "epoch": 3.0650827067669173, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 55957 + }, + { + "epoch": 3.065142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 55958 + }, + { + "epoch": 3.065203007518797, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 55959 + }, + { + "epoch": 3.0652631578947367, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 55960 + }, + { + "epoch": 3.065323308270677, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 55961 + }, + { + "epoch": 3.0653834586466164, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 55962 + }, + { + "epoch": 3.0654436090225565, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 55963 + }, + { + "epoch": 3.065503759398496, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 55964 + }, + { + "epoch": 3.0655639097744363, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 55965 + }, + { + "epoch": 3.065624060150376, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1003, + "step": 55966 + }, + { + "epoch": 3.0656842105263156, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0183, + "step": 55967 + }, + { + "epoch": 3.0657443609022557, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 55968 + }, + { + "epoch": 3.0658045112781953, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 55969 + }, + { + "epoch": 3.0658646616541354, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 55970 + }, + { + "epoch": 3.065924812030075, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 55971 + }, + { + "epoch": 3.065984962406015, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 55972 + }, + { + "epoch": 3.066045112781955, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 55973 + }, + { + "epoch": 3.066105263157895, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 55974 + }, + { + "epoch": 3.0661654135338345, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 55975 + }, + { + "epoch": 3.0662255639097746, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 55976 + }, + { + "epoch": 3.0662857142857143, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 55977 + }, + { + "epoch": 3.066345864661654, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 55978 + }, + { + "epoch": 3.066406015037594, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 55979 + }, + { + "epoch": 3.0664661654135337, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1136, + "step": 55980 + }, + { + "epoch": 3.0665263157894738, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9487, + "step": 55981 + }, + { + "epoch": 3.0665864661654134, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 55982 + }, + { + "epoch": 3.0666466165413535, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 55983 + }, + { + "epoch": 3.066706766917293, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 55984 + }, + { + "epoch": 3.0667669172932333, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 55985 + }, + { + "epoch": 3.066827067669173, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 55986 + }, + { + "epoch": 3.066887218045113, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9107, + "step": 55987 + }, + { + "epoch": 3.0669473684210526, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.8791, + "step": 55988 + }, + { + "epoch": 3.0670075187969923, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 55989 + }, + { + "epoch": 3.0670676691729324, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 55990 + }, + { + "epoch": 3.067127819548872, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 55991 + }, + { + "epoch": 3.067187969924812, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 55992 + }, + { + "epoch": 3.067248120300752, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 55993 + }, + { + "epoch": 3.067308270676692, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 55994 + }, + { + "epoch": 3.0673684210526315, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 55995 + }, + { + "epoch": 3.0674285714285716, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 55996 + }, + { + "epoch": 3.0674887218045113, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 0.9253, + "step": 55997 + }, + { + "epoch": 3.067548872180451, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 55998 + }, + { + "epoch": 3.067609022556391, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 55999 + }, + { + "epoch": 3.0676691729323307, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 56000 + }, + { + "epoch": 3.0677293233082708, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.105, + "step": 56001 + }, + { + "epoch": 3.0677894736842104, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 56002 + }, + { + "epoch": 3.0678496240601505, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 56003 + }, + { + "epoch": 3.06790977443609, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 56004 + }, + { + "epoch": 3.0679699248120302, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 56005 + }, + { + "epoch": 3.06803007518797, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1778, + "step": 56006 + }, + { + "epoch": 3.06809022556391, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.944, + "step": 56007 + }, + { + "epoch": 3.0681503759398496, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 56008 + }, + { + "epoch": 3.0682105263157893, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 56009 + }, + { + "epoch": 3.0682706766917294, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0907, + "step": 56010 + }, + { + "epoch": 3.068330827067669, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 56011 + }, + { + "epoch": 3.068390977443609, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 56012 + }, + { + "epoch": 3.0684511278195488, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 56013 + }, + { + "epoch": 3.068511278195489, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 56014 + }, + { + "epoch": 3.0685714285714285, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 56015 + }, + { + "epoch": 3.0686315789473686, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 56016 + }, + { + "epoch": 3.0686917293233082, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 56017 + }, + { + "epoch": 3.0687518796992483, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 56018 + }, + { + "epoch": 3.068812030075188, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 56019 + }, + { + "epoch": 3.0688721804511276, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 56020 + }, + { + "epoch": 3.0689323308270677, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 56021 + }, + { + "epoch": 3.0689924812030074, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 56022 + }, + { + "epoch": 3.0690526315789475, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 56023 + }, + { + "epoch": 3.069112781954887, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9292, + "step": 56024 + }, + { + "epoch": 3.069172932330827, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 56025 + }, + { + "epoch": 3.069233082706767, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 56026 + }, + { + "epoch": 3.069293233082707, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 56027 + }, + { + "epoch": 3.0693533834586466, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 56028 + }, + { + "epoch": 3.0694135338345863, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9154, + "step": 56029 + }, + { + "epoch": 3.0694736842105264, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 56030 + }, + { + "epoch": 3.069533834586466, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 56031 + }, + { + "epoch": 3.069593984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 56032 + }, + { + "epoch": 3.0696541353383457, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.94, + "step": 56033 + }, + { + "epoch": 3.069714285714286, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 56034 + }, + { + "epoch": 3.0697744360902255, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 56035 + }, + { + "epoch": 3.0698345864661656, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 56036 + }, + { + "epoch": 3.0698947368421052, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 56037 + }, + { + "epoch": 3.0699548872180453, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 56038 + }, + { + "epoch": 3.070015037593985, + "grad_norm": 1.484375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 56039 + }, + { + "epoch": 3.0700751879699246, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0996, + "step": 56040 + }, + { + "epoch": 3.0701353383458647, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9245, + "step": 56041 + }, + { + "epoch": 3.0701954887218044, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 56042 + }, + { + "epoch": 3.0702556390977445, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 56043 + }, + { + "epoch": 3.070315789473684, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1139, + "step": 56044 + }, + { + "epoch": 3.070375939849624, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 56045 + }, + { + "epoch": 3.070436090225564, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9102, + "step": 56046 + }, + { + "epoch": 3.070496240601504, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9113, + "step": 56047 + }, + { + "epoch": 3.0705563909774436, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9246, + "step": 56048 + }, + { + "epoch": 3.0706165413533832, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 56049 + }, + { + "epoch": 3.0706766917293233, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 56050 + }, + { + "epoch": 3.070736842105263, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 56051 + }, + { + "epoch": 3.070796992481203, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 56052 + }, + { + "epoch": 3.0708571428571427, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 56053 + }, + { + "epoch": 3.070917293233083, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 56054 + }, + { + "epoch": 3.0709774436090225, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9176, + "step": 56055 + }, + { + "epoch": 3.0710375939849626, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 56056 + }, + { + "epoch": 3.071097744360902, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 56057 + }, + { + "epoch": 3.0711578947368423, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 56058 + }, + { + "epoch": 3.071218045112782, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9126, + "step": 56059 + }, + { + "epoch": 3.0712781954887216, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9296, + "step": 56060 + }, + { + "epoch": 3.0713383458646617, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 56061 + }, + { + "epoch": 3.0713984962406014, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 56062 + }, + { + "epoch": 3.0714586466165414, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 56063 + }, + { + "epoch": 3.071518796992481, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 56064 + }, + { + "epoch": 3.071578947368421, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 56065 + }, + { + "epoch": 3.071639097744361, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 56066 + }, + { + "epoch": 3.071699248120301, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 56067 + }, + { + "epoch": 3.0717593984962406, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 56068 + }, + { + "epoch": 3.0718195488721802, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 56069 + }, + { + "epoch": 3.0718796992481203, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 56070 + }, + { + "epoch": 3.07193984962406, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 56071 + }, + { + "epoch": 3.072, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9314, + "step": 56072 + }, + { + "epoch": 3.0720601503759397, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 56073 + }, + { + "epoch": 3.07212030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 56074 + }, + { + "epoch": 3.0721804511278195, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 56075 + }, + { + "epoch": 3.0722406015037595, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9433, + "step": 56076 + }, + { + "epoch": 3.072300751879699, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 56077 + }, + { + "epoch": 3.0723609022556393, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 56078 + }, + { + "epoch": 3.072421052631579, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 56079 + }, + { + "epoch": 3.0724812030075186, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 56080 + }, + { + "epoch": 3.0725413533834587, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 56081 + }, + { + "epoch": 3.0726015037593983, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 56082 + }, + { + "epoch": 3.0726616541353384, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 56083 + }, + { + "epoch": 3.072721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 56084 + }, + { + "epoch": 3.072781954887218, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 56085 + }, + { + "epoch": 3.072842105263158, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 56086 + }, + { + "epoch": 3.072902255639098, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 56087 + }, + { + "epoch": 3.0729624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 56088 + }, + { + "epoch": 3.0730225563909777, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 56089 + }, + { + "epoch": 3.0730827067669173, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 56090 + }, + { + "epoch": 3.073142857142857, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 56091 + }, + { + "epoch": 3.073203007518797, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 56092 + }, + { + "epoch": 3.0732631578947367, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 56093 + }, + { + "epoch": 3.073323308270677, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 56094 + }, + { + "epoch": 3.0733834586466164, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9052, + "step": 56095 + }, + { + "epoch": 3.0734436090225565, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 56096 + }, + { + "epoch": 3.073503759398496, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 56097 + }, + { + "epoch": 3.0735639097744363, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 56098 + }, + { + "epoch": 3.073624060150376, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 56099 + }, + { + "epoch": 3.0736842105263156, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 56100 + }, + { + "epoch": 3.0737443609022557, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.934, + "step": 56101 + }, + { + "epoch": 3.0738045112781953, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 56102 + }, + { + "epoch": 3.0738646616541354, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 56103 + }, + { + "epoch": 3.073924812030075, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0859, + "step": 56104 + }, + { + "epoch": 3.073984962406015, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 56105 + }, + { + "epoch": 3.074045112781955, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 56106 + }, + { + "epoch": 3.074105263157895, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 56107 + }, + { + "epoch": 3.0741654135338345, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 56108 + }, + { + "epoch": 3.0742255639097746, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 56109 + }, + { + "epoch": 3.0742857142857143, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 56110 + }, + { + "epoch": 3.074345864661654, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 56111 + }, + { + "epoch": 3.074406015037594, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 56112 + }, + { + "epoch": 3.0744661654135337, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 56113 + }, + { + "epoch": 3.0745263157894738, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9107, + "step": 56114 + }, + { + "epoch": 3.0745864661654134, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 56115 + }, + { + "epoch": 3.0746466165413535, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 56116 + }, + { + "epoch": 3.074706766917293, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 56117 + }, + { + "epoch": 3.0747669172932333, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9015, + "step": 56118 + }, + { + "epoch": 3.074827067669173, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 56119 + }, + { + "epoch": 3.074887218045113, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 56120 + }, + { + "epoch": 3.0749473684210527, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 56121 + }, + { + "epoch": 3.0750075187969923, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 56122 + }, + { + "epoch": 3.0750676691729324, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 56123 + }, + { + "epoch": 3.075127819548872, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 56124 + }, + { + "epoch": 3.075187969924812, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 56125 + }, + { + "epoch": 3.075248120300752, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 56126 + }, + { + "epoch": 3.075308270676692, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 56127 + }, + { + "epoch": 3.0753684210526315, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 56128 + }, + { + "epoch": 3.0754285714285716, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 56129 + }, + { + "epoch": 3.0754887218045113, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 56130 + }, + { + "epoch": 3.075548872180451, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 56131 + }, + { + "epoch": 3.075609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 56132 + }, + { + "epoch": 3.0756691729323307, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 56133 + }, + { + "epoch": 3.0757293233082708, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1207, + "step": 56134 + }, + { + "epoch": 3.0757894736842104, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 56135 + }, + { + "epoch": 3.0758496240601505, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 56136 + }, + { + "epoch": 3.07590977443609, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9509, + "step": 56137 + }, + { + "epoch": 3.0759699248120302, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9077, + "step": 56138 + }, + { + "epoch": 3.07603007518797, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 56139 + }, + { + "epoch": 3.07609022556391, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9303, + "step": 56140 + }, + { + "epoch": 3.0761503759398496, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9183, + "step": 56141 + }, + { + "epoch": 3.0762105263157893, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 56142 + }, + { + "epoch": 3.0762706766917294, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 56143 + }, + { + "epoch": 3.076330827067669, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 56144 + }, + { + "epoch": 3.076390977443609, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 56145 + }, + { + "epoch": 3.0764511278195488, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 56146 + }, + { + "epoch": 3.076511278195489, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1237, + "step": 56147 + }, + { + "epoch": 3.0765714285714285, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9107, + "step": 56148 + }, + { + "epoch": 3.0766315789473686, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 56149 + }, + { + "epoch": 3.0766917293233083, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 56150 + }, + { + "epoch": 3.0767518796992483, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 56151 + }, + { + "epoch": 3.076812030075188, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 56152 + }, + { + "epoch": 3.0768721804511276, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 56153 + }, + { + "epoch": 3.0769323308270677, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 56154 + }, + { + "epoch": 3.0769924812030074, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 56155 + }, + { + "epoch": 3.0770526315789475, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9171, + "step": 56156 + }, + { + "epoch": 3.077112781954887, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 56157 + }, + { + "epoch": 3.0771729323308272, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 56158 + }, + { + "epoch": 3.077233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 56159 + }, + { + "epoch": 3.077293233082707, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 56160 + }, + { + "epoch": 3.0773533834586466, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1486, + "step": 56161 + }, + { + "epoch": 3.0774135338345863, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 56162 + }, + { + "epoch": 3.0774736842105264, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 56163 + }, + { + "epoch": 3.077533834586466, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 56164 + }, + { + "epoch": 3.077593984962406, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 56165 + }, + { + "epoch": 3.0776541353383458, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 56166 + }, + { + "epoch": 3.077714285714286, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 56167 + }, + { + "epoch": 3.0777744360902255, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9262, + "step": 56168 + }, + { + "epoch": 3.0778345864661656, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 56169 + }, + { + "epoch": 3.0778947368421052, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 56170 + }, + { + "epoch": 3.077954887218045, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 56171 + }, + { + "epoch": 3.078015037593985, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9361, + "step": 56172 + }, + { + "epoch": 3.0780751879699246, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 56173 + }, + { + "epoch": 3.0781353383458647, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 56174 + }, + { + "epoch": 3.0781954887218044, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 56175 + }, + { + "epoch": 3.0782556390977445, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 56176 + }, + { + "epoch": 3.078315789473684, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9394, + "step": 56177 + }, + { + "epoch": 3.078375939849624, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 56178 + }, + { + "epoch": 3.078436090225564, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 56179 + }, + { + "epoch": 3.078496240601504, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 56180 + }, + { + "epoch": 3.0785563909774436, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 56181 + }, + { + "epoch": 3.0786165413533833, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 56182 + }, + { + "epoch": 3.0786766917293233, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9445, + "step": 56183 + }, + { + "epoch": 3.078736842105263, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 56184 + }, + { + "epoch": 3.078796992481203, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 56185 + }, + { + "epoch": 3.0788571428571427, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9279, + "step": 56186 + }, + { + "epoch": 3.078917293233083, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 56187 + }, + { + "epoch": 3.0789774436090225, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 56188 + }, + { + "epoch": 3.0790375939849626, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 56189 + }, + { + "epoch": 3.0790977443609022, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 56190 + }, + { + "epoch": 3.0791578947368423, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 56191 + }, + { + "epoch": 3.079218045112782, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 56192 + }, + { + "epoch": 3.0792781954887216, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 56193 + }, + { + "epoch": 3.0793383458646617, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 56194 + }, + { + "epoch": 3.0793984962406014, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 56195 + }, + { + "epoch": 3.0794586466165415, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 56196 + }, + { + "epoch": 3.079518796992481, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 56197 + }, + { + "epoch": 3.079578947368421, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 56198 + }, + { + "epoch": 3.079639097744361, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 56199 + }, + { + "epoch": 3.079699248120301, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 56200 + }, + { + "epoch": 3.0797593984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 56201 + }, + { + "epoch": 3.0798195488721802, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 56202 + }, + { + "epoch": 3.0798796992481203, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 56203 + }, + { + "epoch": 3.07993984962406, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9281, + "step": 56204 + }, + { + "epoch": 3.08, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 56205 + }, + { + "epoch": 3.0800601503759397, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 56206 + }, + { + "epoch": 3.08012030075188, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 56207 + }, + { + "epoch": 3.0801804511278195, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 56208 + }, + { + "epoch": 3.0802406015037596, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 56209 + }, + { + "epoch": 3.080300751879699, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 56210 + }, + { + "epoch": 3.0803609022556393, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 56211 + }, + { + "epoch": 3.080421052631579, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 56212 + }, + { + "epoch": 3.0804812030075186, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 56213 + }, + { + "epoch": 3.0805413533834587, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 56214 + }, + { + "epoch": 3.0806015037593983, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 56215 + }, + { + "epoch": 3.0806616541353384, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 56216 + }, + { + "epoch": 3.080721804511278, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9345, + "step": 56217 + }, + { + "epoch": 3.080781954887218, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 56218 + }, + { + "epoch": 3.080842105263158, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 56219 + }, + { + "epoch": 3.080902255639098, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 56220 + }, + { + "epoch": 3.0809624060150376, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 56221 + }, + { + "epoch": 3.0810225563909777, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9152, + "step": 56222 + }, + { + "epoch": 3.0810827067669173, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 56223 + }, + { + "epoch": 3.081142857142857, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 56224 + }, + { + "epoch": 3.081203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 56225 + }, + { + "epoch": 3.0812631578947367, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 56226 + }, + { + "epoch": 3.081323308270677, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 56227 + }, + { + "epoch": 3.0813834586466164, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0566, + "step": 56228 + }, + { + "epoch": 3.0814436090225565, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 56229 + }, + { + "epoch": 3.081503759398496, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 56230 + }, + { + "epoch": 3.0815639097744363, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 56231 + }, + { + "epoch": 3.081624060150376, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 56232 + }, + { + "epoch": 3.0816842105263156, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 56233 + }, + { + "epoch": 3.0817443609022557, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 56234 + }, + { + "epoch": 3.0818045112781953, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 56235 + }, + { + "epoch": 3.0818646616541354, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 56236 + }, + { + "epoch": 3.081924812030075, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0284, + "step": 56237 + }, + { + "epoch": 3.081984962406015, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 56238 + }, + { + "epoch": 3.082045112781955, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 56239 + }, + { + "epoch": 3.082105263157895, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 56240 + }, + { + "epoch": 3.0821654135338346, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 56241 + }, + { + "epoch": 3.0822255639097746, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 56242 + }, + { + "epoch": 3.0822857142857143, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.2764, + "step": 56243 + }, + { + "epoch": 3.082345864661654, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 56244 + }, + { + "epoch": 3.082406015037594, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 56245 + }, + { + "epoch": 3.0824661654135337, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9509, + "step": 56246 + }, + { + "epoch": 3.082526315789474, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 56247 + }, + { + "epoch": 3.0825864661654134, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 56248 + }, + { + "epoch": 3.0826466165413535, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 56249 + }, + { + "epoch": 3.082706766917293, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 56250 + }, + { + "epoch": 3.0827669172932333, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 56251 + }, + { + "epoch": 3.082827067669173, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 56252 + }, + { + "epoch": 3.082887218045113, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 56253 + }, + { + "epoch": 3.0829473684210527, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 56254 + }, + { + "epoch": 3.0830075187969923, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 56255 + }, + { + "epoch": 3.0830676691729324, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 56256 + }, + { + "epoch": 3.083127819548872, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 56257 + }, + { + "epoch": 3.083187969924812, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 56258 + }, + { + "epoch": 3.083248120300752, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 56259 + }, + { + "epoch": 3.083308270676692, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 56260 + }, + { + "epoch": 3.0833684210526315, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 0.9729, + "step": 56261 + }, + { + "epoch": 3.0834285714285716, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 56262 + }, + { + "epoch": 3.0834887218045113, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 56263 + }, + { + "epoch": 3.083548872180451, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 56264 + }, + { + "epoch": 3.083609022556391, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 56265 + }, + { + "epoch": 3.0836691729323307, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 56266 + }, + { + "epoch": 3.0837293233082708, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 56267 + }, + { + "epoch": 3.0837894736842104, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 56268 + }, + { + "epoch": 3.0838496240601505, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 56269 + }, + { + "epoch": 3.08390977443609, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 56270 + }, + { + "epoch": 3.0839699248120302, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 56271 + }, + { + "epoch": 3.08403007518797, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 56272 + }, + { + "epoch": 3.0840902255639095, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 56273 + }, + { + "epoch": 3.0841503759398496, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9416, + "step": 56274 + }, + { + "epoch": 3.0842105263157893, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 56275 + }, + { + "epoch": 3.0842706766917294, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 56276 + }, + { + "epoch": 3.084330827067669, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 56277 + }, + { + "epoch": 3.084390977443609, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1475, + "step": 56278 + }, + { + "epoch": 3.0844511278195488, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 56279 + }, + { + "epoch": 3.084511278195489, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 56280 + }, + { + "epoch": 3.0845714285714285, + "grad_norm": 1.8984375, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 56281 + }, + { + "epoch": 3.0846315789473686, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 56282 + }, + { + "epoch": 3.0846917293233083, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 56283 + }, + { + "epoch": 3.084751879699248, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 56284 + }, + { + "epoch": 3.084812030075188, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 56285 + }, + { + "epoch": 3.0848721804511277, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1116, + "step": 56286 + }, + { + "epoch": 3.0849323308270677, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 56287 + }, + { + "epoch": 3.0849924812030074, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 56288 + }, + { + "epoch": 3.0850526315789475, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 56289 + }, + { + "epoch": 3.085112781954887, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 56290 + }, + { + "epoch": 3.0851729323308272, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 56291 + }, + { + "epoch": 3.085233082706767, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 56292 + }, + { + "epoch": 3.085293233082707, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 56293 + }, + { + "epoch": 3.0853533834586466, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 56294 + }, + { + "epoch": 3.0854135338345863, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 56295 + }, + { + "epoch": 3.0854736842105264, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9339, + "step": 56296 + }, + { + "epoch": 3.085533834586466, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 56297 + }, + { + "epoch": 3.085593984962406, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.8999, + "step": 56298 + }, + { + "epoch": 3.0856541353383458, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 56299 + }, + { + "epoch": 3.085714285714286, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 56300 + }, + { + "epoch": 3.0857744360902255, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 56301 + }, + { + "epoch": 3.0858345864661656, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 56302 + }, + { + "epoch": 3.0858947368421052, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 56303 + }, + { + "epoch": 3.085954887218045, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 56304 + }, + { + "epoch": 3.086015037593985, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 56305 + }, + { + "epoch": 3.0860751879699246, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 56306 + }, + { + "epoch": 3.0861353383458647, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 56307 + }, + { + "epoch": 3.0861954887218044, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1195, + "step": 56308 + }, + { + "epoch": 3.0862556390977445, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 56309 + }, + { + "epoch": 3.086315789473684, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 56310 + }, + { + "epoch": 3.086375939849624, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 56311 + }, + { + "epoch": 3.086436090225564, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 56312 + }, + { + "epoch": 3.086496240601504, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 56313 + }, + { + "epoch": 3.0865563909774436, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9245, + "step": 56314 + }, + { + "epoch": 3.0866165413533833, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 56315 + }, + { + "epoch": 3.0866766917293234, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 56316 + }, + { + "epoch": 3.086736842105263, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 56317 + }, + { + "epoch": 3.086796992481203, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 56318 + }, + { + "epoch": 3.0868571428571427, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 56319 + }, + { + "epoch": 3.086917293233083, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 56320 + }, + { + "epoch": 3.0869774436090225, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 56321 + }, + { + "epoch": 3.0870375939849626, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 56322 + }, + { + "epoch": 3.0870977443609022, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 56323 + }, + { + "epoch": 3.0871578947368423, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 56324 + }, + { + "epoch": 3.087218045112782, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.967, + "step": 56325 + }, + { + "epoch": 3.0872781954887216, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1215, + "step": 56326 + }, + { + "epoch": 3.0873383458646617, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1251, + "step": 56327 + }, + { + "epoch": 3.0873984962406014, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 56328 + }, + { + "epoch": 3.0874586466165415, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 56329 + }, + { + "epoch": 3.087518796992481, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 56330 + }, + { + "epoch": 3.087578947368421, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1156, + "step": 56331 + }, + { + "epoch": 3.087639097744361, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 56332 + }, + { + "epoch": 3.087699248120301, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 56333 + }, + { + "epoch": 3.0877593984962406, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 56334 + }, + { + "epoch": 3.0878195488721802, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 56335 + }, + { + "epoch": 3.0878796992481203, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 56336 + }, + { + "epoch": 3.08793984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0521, + "step": 56337 + }, + { + "epoch": 3.088, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 56338 + }, + { + "epoch": 3.0880601503759397, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 56339 + }, + { + "epoch": 3.08812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 56340 + }, + { + "epoch": 3.0881804511278195, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 56341 + }, + { + "epoch": 3.0882406015037596, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 56342 + }, + { + "epoch": 3.088300751879699, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 56343 + }, + { + "epoch": 3.0883609022556393, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 56344 + }, + { + "epoch": 3.088421052631579, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 56345 + }, + { + "epoch": 3.0884812030075186, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 56346 + }, + { + "epoch": 3.0885413533834587, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 56347 + }, + { + "epoch": 3.0886015037593983, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9321, + "step": 56348 + }, + { + "epoch": 3.0886616541353384, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 56349 + }, + { + "epoch": 3.088721804511278, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 56350 + }, + { + "epoch": 3.088781954887218, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 56351 + }, + { + "epoch": 3.088842105263158, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 56352 + }, + { + "epoch": 3.088902255639098, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 56353 + }, + { + "epoch": 3.0889624060150376, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 56354 + }, + { + "epoch": 3.0890225563909777, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 56355 + }, + { + "epoch": 3.0890827067669173, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9313, + "step": 56356 + }, + { + "epoch": 3.089142857142857, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 56357 + }, + { + "epoch": 3.089203007518797, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 56358 + }, + { + "epoch": 3.0892631578947367, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 56359 + }, + { + "epoch": 3.089323308270677, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 56360 + }, + { + "epoch": 3.0893834586466165, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 56361 + }, + { + "epoch": 3.0894436090225565, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 56362 + }, + { + "epoch": 3.089503759398496, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.8992, + "step": 56363 + }, + { + "epoch": 3.0895639097744363, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 56364 + }, + { + "epoch": 3.089624060150376, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 56365 + }, + { + "epoch": 3.0896842105263156, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 56366 + }, + { + "epoch": 3.0897443609022557, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 56367 + }, + { + "epoch": 3.0898045112781953, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 56368 + }, + { + "epoch": 3.0898646616541354, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 56369 + }, + { + "epoch": 3.089924812030075, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 56370 + }, + { + "epoch": 3.089984962406015, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 56371 + }, + { + "epoch": 3.090045112781955, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 56372 + }, + { + "epoch": 3.090105263157895, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 56373 + }, + { + "epoch": 3.0901654135338346, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 56374 + }, + { + "epoch": 3.090225563909774, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 56375 + }, + { + "epoch": 3.0902857142857143, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 56376 + }, + { + "epoch": 3.090345864661654, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 56377 + }, + { + "epoch": 3.090406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 56378 + }, + { + "epoch": 3.0904661654135337, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 56379 + }, + { + "epoch": 3.090526315789474, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 56380 + }, + { + "epoch": 3.0905864661654134, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 56381 + }, + { + "epoch": 3.0906466165413535, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 56382 + }, + { + "epoch": 3.090706766917293, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9324, + "step": 56383 + }, + { + "epoch": 3.0907669172932333, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 56384 + }, + { + "epoch": 3.090827067669173, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 56385 + }, + { + "epoch": 3.0908872180451126, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.932, + "step": 56386 + }, + { + "epoch": 3.0909473684210527, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 56387 + }, + { + "epoch": 3.0910075187969923, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 56388 + }, + { + "epoch": 3.0910676691729324, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 56389 + }, + { + "epoch": 3.091127819548872, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 56390 + }, + { + "epoch": 3.091187969924812, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 56391 + }, + { + "epoch": 3.091248120300752, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 56392 + }, + { + "epoch": 3.091308270676692, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 56393 + }, + { + "epoch": 3.0913684210526315, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.915, + "step": 56394 + }, + { + "epoch": 3.0914285714285716, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 56395 + }, + { + "epoch": 3.0914887218045113, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 56396 + }, + { + "epoch": 3.091548872180451, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.8978, + "step": 56397 + }, + { + "epoch": 3.091609022556391, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 56398 + }, + { + "epoch": 3.0916691729323307, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 56399 + }, + { + "epoch": 3.0917293233082708, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 56400 + }, + { + "epoch": 3.0917894736842104, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 56401 + }, + { + "epoch": 3.0918496240601505, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 56402 + }, + { + "epoch": 3.09190977443609, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 56403 + }, + { + "epoch": 3.0919699248120303, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.8998, + "step": 56404 + }, + { + "epoch": 3.09203007518797, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.8881, + "step": 56405 + }, + { + "epoch": 3.0920902255639096, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 56406 + }, + { + "epoch": 3.0921503759398496, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 56407 + }, + { + "epoch": 3.0922105263157893, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 56408 + }, + { + "epoch": 3.0922706766917294, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 56409 + }, + { + "epoch": 3.092330827067669, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 56410 + }, + { + "epoch": 3.092390977443609, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 56411 + }, + { + "epoch": 3.092451127819549, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 56412 + }, + { + "epoch": 3.092511278195489, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 56413 + }, + { + "epoch": 3.0925714285714285, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 56414 + }, + { + "epoch": 3.0926315789473686, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 56415 + }, + { + "epoch": 3.0926917293233083, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 56416 + }, + { + "epoch": 3.092751879699248, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 56417 + }, + { + "epoch": 3.092812030075188, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 56418 + }, + { + "epoch": 3.0928721804511277, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9337, + "step": 56419 + }, + { + "epoch": 3.0929323308270678, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 56420 + }, + { + "epoch": 3.0929924812030074, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 56421 + }, + { + "epoch": 3.0930526315789475, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 56422 + }, + { + "epoch": 3.093112781954887, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 56423 + }, + { + "epoch": 3.0931729323308272, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 56424 + }, + { + "epoch": 3.093233082706767, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 56425 + }, + { + "epoch": 3.093293233082707, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 56426 + }, + { + "epoch": 3.0933533834586466, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 56427 + }, + { + "epoch": 3.0934135338345863, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0874, + "step": 56428 + }, + { + "epoch": 3.0934736842105264, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 56429 + }, + { + "epoch": 3.093533834586466, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 56430 + }, + { + "epoch": 3.093593984962406, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 56431 + }, + { + "epoch": 3.0936541353383458, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.154, + "step": 56432 + }, + { + "epoch": 3.093714285714286, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 56433 + }, + { + "epoch": 3.0937744360902255, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9392, + "step": 56434 + }, + { + "epoch": 3.0938345864661656, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 56435 + }, + { + "epoch": 3.0938947368421053, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 56436 + }, + { + "epoch": 3.093954887218045, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 56437 + }, + { + "epoch": 3.094015037593985, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 56438 + }, + { + "epoch": 3.0940751879699246, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 56439 + }, + { + "epoch": 3.0941353383458647, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0627, + "step": 56440 + }, + { + "epoch": 3.0941954887218044, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 56441 + }, + { + "epoch": 3.0942556390977445, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 56442 + }, + { + "epoch": 3.094315789473684, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 56443 + }, + { + "epoch": 3.0943759398496242, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 56444 + }, + { + "epoch": 3.094436090225564, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9245, + "step": 56445 + }, + { + "epoch": 3.094496240601504, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 0.9033, + "step": 56446 + }, + { + "epoch": 3.0945563909774436, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 56447 + }, + { + "epoch": 3.0946165413533833, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 56448 + }, + { + "epoch": 3.0946766917293234, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 56449 + }, + { + "epoch": 3.094736842105263, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 56450 + }, + { + "epoch": 3.094796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 56451 + }, + { + "epoch": 3.0948571428571428, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 56452 + }, + { + "epoch": 3.094917293233083, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 56453 + }, + { + "epoch": 3.0949774436090225, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 56454 + }, + { + "epoch": 3.0950375939849626, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 56455 + }, + { + "epoch": 3.0950977443609022, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 56456 + }, + { + "epoch": 3.0951578947368423, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 56457 + }, + { + "epoch": 3.095218045112782, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0865, + "step": 56458 + }, + { + "epoch": 3.0952781954887216, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9273, + "step": 56459 + }, + { + "epoch": 3.0953383458646617, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 56460 + }, + { + "epoch": 3.0953984962406014, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1127, + "step": 56461 + }, + { + "epoch": 3.0954586466165415, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 56462 + }, + { + "epoch": 3.095518796992481, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 56463 + }, + { + "epoch": 3.095578947368421, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 56464 + }, + { + "epoch": 3.095639097744361, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 56465 + }, + { + "epoch": 3.095699248120301, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 56466 + }, + { + "epoch": 3.0957593984962406, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 56467 + }, + { + "epoch": 3.0958195488721802, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 56468 + }, + { + "epoch": 3.0958796992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 56469 + }, + { + "epoch": 3.09593984962406, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 56470 + }, + { + "epoch": 3.096, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 56471 + }, + { + "epoch": 3.0960601503759397, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 56472 + }, + { + "epoch": 3.09612030075188, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 56473 + }, + { + "epoch": 3.0961804511278195, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 56474 + }, + { + "epoch": 3.0962406015037596, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 56475 + }, + { + "epoch": 3.096300751879699, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 56476 + }, + { + "epoch": 3.096360902255639, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 56477 + }, + { + "epoch": 3.096421052631579, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 56478 + }, + { + "epoch": 3.0964812030075186, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 56479 + }, + { + "epoch": 3.0965413533834587, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 56480 + }, + { + "epoch": 3.0966015037593984, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 56481 + }, + { + "epoch": 3.0966616541353384, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 56482 + }, + { + "epoch": 3.096721804511278, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 56483 + }, + { + "epoch": 3.096781954887218, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 56484 + }, + { + "epoch": 3.096842105263158, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.8997, + "step": 56485 + }, + { + "epoch": 3.096902255639098, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 56486 + }, + { + "epoch": 3.0969624060150376, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 56487 + }, + { + "epoch": 3.0970225563909772, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 56488 + }, + { + "epoch": 3.0970827067669173, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 56489 + }, + { + "epoch": 3.097142857142857, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 56490 + }, + { + "epoch": 3.097203007518797, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 56491 + }, + { + "epoch": 3.0972631578947367, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.2002, + "step": 56492 + }, + { + "epoch": 3.097323308270677, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 56493 + }, + { + "epoch": 3.0973834586466165, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 56494 + }, + { + "epoch": 3.0974436090225566, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 56495 + }, + { + "epoch": 3.097503759398496, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9104, + "step": 56496 + }, + { + "epoch": 3.0975639097744363, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 56497 + }, + { + "epoch": 3.097624060150376, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 56498 + }, + { + "epoch": 3.0976842105263156, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 56499 + }, + { + "epoch": 3.0977443609022557, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 56500 + }, + { + "epoch": 3.0978045112781953, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 56501 + }, + { + "epoch": 3.0978646616541354, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 56502 + }, + { + "epoch": 3.097924812030075, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 0.8764, + "step": 56503 + }, + { + "epoch": 3.097984962406015, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0524, + "step": 56504 + }, + { + "epoch": 3.098045112781955, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 56505 + }, + { + "epoch": 3.098105263157895, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9527, + "step": 56506 + }, + { + "epoch": 3.0981654135338346, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 56507 + }, + { + "epoch": 3.098225563909774, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0978, + "step": 56508 + }, + { + "epoch": 3.0982857142857143, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 56509 + }, + { + "epoch": 3.098345864661654, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 56510 + }, + { + "epoch": 3.098406015037594, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 56511 + }, + { + "epoch": 3.0984661654135337, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 56512 + }, + { + "epoch": 3.098526315789474, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 56513 + }, + { + "epoch": 3.0985864661654134, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 56514 + }, + { + "epoch": 3.0986466165413535, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 56515 + }, + { + "epoch": 3.098706766917293, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 56516 + }, + { + "epoch": 3.0987669172932333, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 56517 + }, + { + "epoch": 3.098827067669173, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 56518 + }, + { + "epoch": 3.0988872180451126, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 56519 + }, + { + "epoch": 3.0989473684210527, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 56520 + }, + { + "epoch": 3.0990075187969923, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 56521 + }, + { + "epoch": 3.0990676691729324, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.1983, + "step": 56522 + }, + { + "epoch": 3.099127819548872, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 56523 + }, + { + "epoch": 3.099187969924812, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 56524 + }, + { + "epoch": 3.099248120300752, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 56525 + }, + { + "epoch": 3.099308270676692, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 56526 + }, + { + "epoch": 3.0993684210526316, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 56527 + }, + { + "epoch": 3.0994285714285716, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9263, + "step": 56528 + }, + { + "epoch": 3.0994887218045113, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 56529 + }, + { + "epoch": 3.099548872180451, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 56530 + }, + { + "epoch": 3.099609022556391, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 56531 + }, + { + "epoch": 3.0996691729323307, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 56532 + }, + { + "epoch": 3.099729323308271, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 56533 + }, + { + "epoch": 3.0997894736842104, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 56534 + }, + { + "epoch": 3.0998496240601505, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9115, + "step": 56535 + }, + { + "epoch": 3.09990977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 56536 + }, + { + "epoch": 3.0999699248120303, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 56537 + }, + { + "epoch": 3.10003007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 56538 + }, + { + "epoch": 3.1000902255639096, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 56539 + }, + { + "epoch": 3.1001503759398497, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 56540 + }, + { + "epoch": 3.1002105263157893, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 56541 + }, + { + "epoch": 3.1002706766917294, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 56542 + }, + { + "epoch": 3.100330827067669, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 56543 + }, + { + "epoch": 3.100390977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 56544 + }, + { + "epoch": 3.100451127819549, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 56545 + }, + { + "epoch": 3.100511278195489, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 56546 + }, + { + "epoch": 3.1005714285714285, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 56547 + }, + { + "epoch": 3.1006315789473686, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9215, + "step": 56548 + }, + { + "epoch": 3.1006917293233083, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 56549 + }, + { + "epoch": 3.100751879699248, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 56550 + }, + { + "epoch": 3.100812030075188, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9144, + "step": 56551 + }, + { + "epoch": 3.1008721804511277, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 56552 + }, + { + "epoch": 3.1009323308270678, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 56553 + }, + { + "epoch": 3.1009924812030074, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 56554 + }, + { + "epoch": 3.1010526315789475, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 56555 + }, + { + "epoch": 3.101112781954887, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 56556 + }, + { + "epoch": 3.1011729323308272, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 56557 + }, + { + "epoch": 3.101233082706767, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 56558 + }, + { + "epoch": 3.101293233082707, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 56559 + }, + { + "epoch": 3.1013533834586466, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1045, + "step": 56560 + }, + { + "epoch": 3.1014135338345863, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 56561 + }, + { + "epoch": 3.1014736842105264, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 56562 + }, + { + "epoch": 3.101533834586466, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 56563 + }, + { + "epoch": 3.101593984962406, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 56564 + }, + { + "epoch": 3.1016541353383458, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 56565 + }, + { + "epoch": 3.101714285714286, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.8936, + "step": 56566 + }, + { + "epoch": 3.1017744360902255, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 56567 + }, + { + "epoch": 3.1018345864661656, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 56568 + }, + { + "epoch": 3.1018947368421053, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 56569 + }, + { + "epoch": 3.101954887218045, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.8992, + "step": 56570 + }, + { + "epoch": 3.102015037593985, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 56571 + }, + { + "epoch": 3.1020751879699247, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 56572 + }, + { + "epoch": 3.1021353383458647, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 56573 + }, + { + "epoch": 3.1021954887218044, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 56574 + }, + { + "epoch": 3.1022556390977445, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 56575 + }, + { + "epoch": 3.102315789473684, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 56576 + }, + { + "epoch": 3.1023759398496242, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 56577 + }, + { + "epoch": 3.102436090225564, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 56578 + }, + { + "epoch": 3.102496240601504, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 56579 + }, + { + "epoch": 3.1025563909774436, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1331, + "step": 56580 + }, + { + "epoch": 3.1026165413533833, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 56581 + }, + { + "epoch": 3.1026766917293234, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9527, + "step": 56582 + }, + { + "epoch": 3.102736842105263, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 56583 + }, + { + "epoch": 3.102796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 56584 + }, + { + "epoch": 3.1028571428571428, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 56585 + }, + { + "epoch": 3.102917293233083, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0882, + "step": 56586 + }, + { + "epoch": 3.1029774436090225, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0794, + "step": 56587 + }, + { + "epoch": 3.1030375939849626, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 56588 + }, + { + "epoch": 3.1030977443609022, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1432, + "step": 56589 + }, + { + "epoch": 3.103157894736842, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 56590 + }, + { + "epoch": 3.103218045112782, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 56591 + }, + { + "epoch": 3.1032781954887216, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.2341, + "step": 56592 + }, + { + "epoch": 3.1033383458646617, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 56593 + }, + { + "epoch": 3.1033984962406014, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 56594 + }, + { + "epoch": 3.1034586466165415, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 56595 + }, + { + "epoch": 3.103518796992481, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 56596 + }, + { + "epoch": 3.103578947368421, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 56597 + }, + { + "epoch": 3.103639097744361, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.895, + "step": 56598 + }, + { + "epoch": 3.103699248120301, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 56599 + }, + { + "epoch": 3.1037593984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 56600 + }, + { + "epoch": 3.1038195488721803, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 56601 + }, + { + "epoch": 3.1038796992481203, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 56602 + }, + { + "epoch": 3.10393984962406, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9348, + "step": 56603 + }, + { + "epoch": 3.104, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 56604 + }, + { + "epoch": 3.1040601503759397, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 56605 + }, + { + "epoch": 3.10412030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 56606 + }, + { + "epoch": 3.1041804511278195, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 56607 + }, + { + "epoch": 3.1042406015037596, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 56608 + }, + { + "epoch": 3.1043007518796992, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9176, + "step": 56609 + }, + { + "epoch": 3.104360902255639, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 56610 + }, + { + "epoch": 3.104421052631579, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 56611 + }, + { + "epoch": 3.1044812030075186, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 56612 + }, + { + "epoch": 3.1045413533834587, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 56613 + }, + { + "epoch": 3.1046015037593984, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 56614 + }, + { + "epoch": 3.1046616541353385, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 56615 + }, + { + "epoch": 3.104721804511278, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 56616 + }, + { + "epoch": 3.104781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 56617 + }, + { + "epoch": 3.104842105263158, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 56618 + }, + { + "epoch": 3.104902255639098, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 56619 + }, + { + "epoch": 3.1049624060150376, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 56620 + }, + { + "epoch": 3.1050225563909772, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 56621 + }, + { + "epoch": 3.1050827067669173, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0954, + "step": 56622 + }, + { + "epoch": 3.105142857142857, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 56623 + }, + { + "epoch": 3.105203007518797, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 56624 + }, + { + "epoch": 3.1052631578947367, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9334, + "step": 56625 + }, + { + "epoch": 3.105323308270677, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 56626 + }, + { + "epoch": 3.1053834586466165, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9288, + "step": 56627 + }, + { + "epoch": 3.1054436090225566, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9527, + "step": 56628 + }, + { + "epoch": 3.105503759398496, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 56629 + }, + { + "epoch": 3.1055639097744363, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 56630 + }, + { + "epoch": 3.105624060150376, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 56631 + }, + { + "epoch": 3.1056842105263156, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 56632 + }, + { + "epoch": 3.1057443609022557, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9635, + "step": 56633 + }, + { + "epoch": 3.1058045112781953, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 56634 + }, + { + "epoch": 3.1058646616541354, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 56635 + }, + { + "epoch": 3.105924812030075, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 56636 + }, + { + "epoch": 3.105984962406015, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 56637 + }, + { + "epoch": 3.106045112781955, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 56638 + }, + { + "epoch": 3.106105263157895, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 56639 + }, + { + "epoch": 3.1061654135338346, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 56640 + }, + { + "epoch": 3.1062255639097742, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 56641 + }, + { + "epoch": 3.1062857142857143, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 56642 + }, + { + "epoch": 3.106345864661654, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 56643 + }, + { + "epoch": 3.106406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9059, + "step": 56644 + }, + { + "epoch": 3.1064661654135337, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 56645 + }, + { + "epoch": 3.106526315789474, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 56646 + }, + { + "epoch": 3.1065864661654135, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 56647 + }, + { + "epoch": 3.1066466165413535, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 56648 + }, + { + "epoch": 3.106706766917293, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 56649 + }, + { + "epoch": 3.1067669172932333, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 56650 + }, + { + "epoch": 3.106827067669173, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9647, + "step": 56651 + }, + { + "epoch": 3.1068872180451126, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.908, + "step": 56652 + }, + { + "epoch": 3.1069473684210527, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 56653 + }, + { + "epoch": 3.1070075187969923, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 56654 + }, + { + "epoch": 3.1070676691729324, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 56655 + }, + { + "epoch": 3.107127819548872, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.922, + "step": 56656 + }, + { + "epoch": 3.107187969924812, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1859, + "step": 56657 + }, + { + "epoch": 3.107248120300752, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 56658 + }, + { + "epoch": 3.107308270676692, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 56659 + }, + { + "epoch": 3.1073684210526316, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 56660 + }, + { + "epoch": 3.1074285714285717, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 56661 + }, + { + "epoch": 3.1074887218045113, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9178, + "step": 56662 + }, + { + "epoch": 3.107548872180451, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 56663 + }, + { + "epoch": 3.107609022556391, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 56664 + }, + { + "epoch": 3.1076691729323307, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 56665 + }, + { + "epoch": 3.107729323308271, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 56666 + }, + { + "epoch": 3.1077894736842104, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 56667 + }, + { + "epoch": 3.1078496240601505, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.1013, + "step": 56668 + }, + { + "epoch": 3.10790977443609, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 56669 + }, + { + "epoch": 3.1079699248120303, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 56670 + }, + { + "epoch": 3.10803007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 56671 + }, + { + "epoch": 3.1080902255639096, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 56672 + }, + { + "epoch": 3.1081503759398497, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 56673 + }, + { + "epoch": 3.1082105263157893, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.956, + "step": 56674 + }, + { + "epoch": 3.1082706766917294, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 56675 + }, + { + "epoch": 3.108330827067669, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 56676 + }, + { + "epoch": 3.108390977443609, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 56677 + }, + { + "epoch": 3.108451127819549, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 56678 + }, + { + "epoch": 3.108511278195489, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9268, + "step": 56679 + }, + { + "epoch": 3.1085714285714285, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 56680 + }, + { + "epoch": 3.1086315789473686, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 56681 + }, + { + "epoch": 3.1086917293233083, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 56682 + }, + { + "epoch": 3.108751879699248, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 56683 + }, + { + "epoch": 3.108812030075188, + "grad_norm": 1.390625, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 56684 + }, + { + "epoch": 3.1088721804511277, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 56685 + }, + { + "epoch": 3.1089323308270678, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 56686 + }, + { + "epoch": 3.1089924812030074, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 56687 + }, + { + "epoch": 3.1090526315789475, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 56688 + }, + { + "epoch": 3.109112781954887, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 56689 + }, + { + "epoch": 3.1091729323308273, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 56690 + }, + { + "epoch": 3.109233082706767, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 56691 + }, + { + "epoch": 3.1092932330827066, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 56692 + }, + { + "epoch": 3.1093533834586466, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 56693 + }, + { + "epoch": 3.1094135338345863, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 56694 + }, + { + "epoch": 3.1094736842105264, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 56695 + }, + { + "epoch": 3.109533834586466, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 56696 + }, + { + "epoch": 3.109593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 56697 + }, + { + "epoch": 3.109654135338346, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.2061, + "step": 56698 + }, + { + "epoch": 3.109714285714286, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 56699 + }, + { + "epoch": 3.1097744360902255, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 56700 + }, + { + "epoch": 3.1098345864661656, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 56701 + }, + { + "epoch": 3.1098947368421053, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 56702 + }, + { + "epoch": 3.109954887218045, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 56703 + }, + { + "epoch": 3.110015037593985, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9497, + "step": 56704 + }, + { + "epoch": 3.1100751879699247, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 56705 + }, + { + "epoch": 3.1101353383458648, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 56706 + }, + { + "epoch": 3.1101954887218044, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 56707 + }, + { + "epoch": 3.1102556390977445, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 56708 + }, + { + "epoch": 3.110315789473684, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 56709 + }, + { + "epoch": 3.1103759398496242, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 56710 + }, + { + "epoch": 3.110436090225564, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 56711 + }, + { + "epoch": 3.1104962406015035, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 56712 + }, + { + "epoch": 3.1105563909774436, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 56713 + }, + { + "epoch": 3.1106165413533833, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 56714 + }, + { + "epoch": 3.1106766917293234, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 56715 + }, + { + "epoch": 3.110736842105263, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 56716 + }, + { + "epoch": 3.110796992481203, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 56717 + }, + { + "epoch": 3.1108571428571428, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 56718 + }, + { + "epoch": 3.110917293233083, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9197, + "step": 56719 + }, + { + "epoch": 3.1109774436090225, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1206, + "step": 56720 + }, + { + "epoch": 3.1110375939849626, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9236, + "step": 56721 + }, + { + "epoch": 3.1110977443609023, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 56722 + }, + { + "epoch": 3.111157894736842, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 56723 + }, + { + "epoch": 3.111218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 56724 + }, + { + "epoch": 3.1112781954887216, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 56725 + }, + { + "epoch": 3.1113383458646617, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 56726 + }, + { + "epoch": 3.1113984962406014, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 56727 + }, + { + "epoch": 3.1114586466165415, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.8473, + "step": 56728 + }, + { + "epoch": 3.111518796992481, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 56729 + }, + { + "epoch": 3.111578947368421, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 56730 + }, + { + "epoch": 3.111639097744361, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.1261, + "step": 56731 + }, + { + "epoch": 3.111699248120301, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 56732 + }, + { + "epoch": 3.1117593984962406, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 56733 + }, + { + "epoch": 3.1118195488721803, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 56734 + }, + { + "epoch": 3.1118796992481204, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 56735 + }, + { + "epoch": 3.11193984962406, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 56736 + }, + { + "epoch": 3.112, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 56737 + }, + { + "epoch": 3.1120601503759397, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 56738 + }, + { + "epoch": 3.11212030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 56739 + }, + { + "epoch": 3.1121804511278195, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 56740 + }, + { + "epoch": 3.1122406015037596, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 56741 + }, + { + "epoch": 3.1123007518796992, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 56742 + }, + { + "epoch": 3.112360902255639, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.1986, + "step": 56743 + }, + { + "epoch": 3.112421052631579, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 56744 + }, + { + "epoch": 3.1124812030075186, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 56745 + }, + { + "epoch": 3.1125413533834587, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 56746 + }, + { + "epoch": 3.1126015037593984, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 56747 + }, + { + "epoch": 3.1126616541353385, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 56748 + }, + { + "epoch": 3.112721804511278, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 56749 + }, + { + "epoch": 3.112781954887218, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 56750 + }, + { + "epoch": 3.112842105263158, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 56751 + }, + { + "epoch": 3.112902255639098, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 56752 + }, + { + "epoch": 3.1129624060150376, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 56753 + }, + { + "epoch": 3.1130225563909772, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 56754 + }, + { + "epoch": 3.1130827067669173, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 56755 + }, + { + "epoch": 3.113142857142857, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 56756 + }, + { + "epoch": 3.113203007518797, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9501, + "step": 56757 + }, + { + "epoch": 3.1132631578947367, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 56758 + }, + { + "epoch": 3.113323308270677, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 56759 + }, + { + "epoch": 3.1133834586466165, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 56760 + }, + { + "epoch": 3.1134436090225566, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 56761 + }, + { + "epoch": 3.113503759398496, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 56762 + }, + { + "epoch": 3.1135639097744363, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 56763 + }, + { + "epoch": 3.113624060150376, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.87, + "step": 56764 + }, + { + "epoch": 3.1136842105263156, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 56765 + }, + { + "epoch": 3.1137443609022557, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 56766 + }, + { + "epoch": 3.1138045112781954, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 56767 + }, + { + "epoch": 3.1138646616541354, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 56768 + }, + { + "epoch": 3.113924812030075, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 56769 + }, + { + "epoch": 3.113984962406015, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 56770 + }, + { + "epoch": 3.114045112781955, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 56771 + }, + { + "epoch": 3.114105263157895, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 56772 + }, + { + "epoch": 3.1141654135338346, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 56773 + }, + { + "epoch": 3.1142255639097742, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 56774 + }, + { + "epoch": 3.1142857142857143, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 56775 + }, + { + "epoch": 3.114345864661654, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0673, + "step": 56776 + }, + { + "epoch": 3.114406015037594, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 56777 + }, + { + "epoch": 3.1144661654135337, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 56778 + }, + { + "epoch": 3.114526315789474, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.2272, + "step": 56779 + }, + { + "epoch": 3.1145864661654135, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 56780 + }, + { + "epoch": 3.1146466165413536, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 56781 + }, + { + "epoch": 3.114706766917293, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 56782 + }, + { + "epoch": 3.1147669172932333, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 56783 + }, + { + "epoch": 3.114827067669173, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 56784 + }, + { + "epoch": 3.1148872180451126, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 56785 + }, + { + "epoch": 3.1149473684210527, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 56786 + }, + { + "epoch": 3.1150075187969923, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 56787 + }, + { + "epoch": 3.1150676691729324, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 56788 + }, + { + "epoch": 3.115127819548872, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 56789 + }, + { + "epoch": 3.115187969924812, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 56790 + }, + { + "epoch": 3.115248120300752, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9244, + "step": 56791 + }, + { + "epoch": 3.115308270676692, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 56792 + }, + { + "epoch": 3.1153684210526316, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 56793 + }, + { + "epoch": 3.115428571428571, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1433, + "step": 56794 + }, + { + "epoch": 3.1154887218045113, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 56795 + }, + { + "epoch": 3.115548872180451, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1042, + "step": 56796 + }, + { + "epoch": 3.115609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 56797 + }, + { + "epoch": 3.1156691729323307, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 56798 + }, + { + "epoch": 3.115729323308271, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9284, + "step": 56799 + }, + { + "epoch": 3.1157894736842104, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 56800 + }, + { + "epoch": 3.1158496240601505, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9667, + "step": 56801 + }, + { + "epoch": 3.11590977443609, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 56802 + }, + { + "epoch": 3.1159699248120303, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 56803 + }, + { + "epoch": 3.11603007518797, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 56804 + }, + { + "epoch": 3.1160902255639096, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 56805 + }, + { + "epoch": 3.1161503759398497, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 56806 + }, + { + "epoch": 3.1162105263157893, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 56807 + }, + { + "epoch": 3.1162706766917294, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 56808 + }, + { + "epoch": 3.116330827067669, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 56809 + }, + { + "epoch": 3.116390977443609, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 56810 + }, + { + "epoch": 3.116451127819549, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 56811 + }, + { + "epoch": 3.116511278195489, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 56812 + }, + { + "epoch": 3.1165714285714285, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.1446, + "step": 56813 + }, + { + "epoch": 3.116631578947368, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 56814 + }, + { + "epoch": 3.1166917293233083, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 56815 + }, + { + "epoch": 3.116751879699248, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 56816 + }, + { + "epoch": 3.116812030075188, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 56817 + }, + { + "epoch": 3.1168721804511277, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 56818 + }, + { + "epoch": 3.1169323308270678, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 56819 + }, + { + "epoch": 3.1169924812030074, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 56820 + }, + { + "epoch": 3.1170526315789475, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 56821 + }, + { + "epoch": 3.117112781954887, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 56822 + }, + { + "epoch": 3.1171729323308273, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 56823 + }, + { + "epoch": 3.117233082706767, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9393, + "step": 56824 + }, + { + "epoch": 3.1172932330827066, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 56825 + }, + { + "epoch": 3.1173533834586467, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 56826 + }, + { + "epoch": 3.1174135338345863, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 56827 + }, + { + "epoch": 3.1174736842105264, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9434, + "step": 56828 + }, + { + "epoch": 3.117533834586466, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 56829 + }, + { + "epoch": 3.117593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 56830 + }, + { + "epoch": 3.117654135338346, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 56831 + }, + { + "epoch": 3.117714285714286, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9331, + "step": 56832 + }, + { + "epoch": 3.1177744360902255, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 56833 + }, + { + "epoch": 3.1178345864661656, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 56834 + }, + { + "epoch": 3.1178947368421053, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0885, + "step": 56835 + }, + { + "epoch": 3.117954887218045, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 56836 + }, + { + "epoch": 3.118015037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 56837 + }, + { + "epoch": 3.1180751879699247, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 56838 + }, + { + "epoch": 3.1181353383458648, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1123, + "step": 56839 + }, + { + "epoch": 3.1181954887218044, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 56840 + }, + { + "epoch": 3.1182556390977445, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 56841 + }, + { + "epoch": 3.118315789473684, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 56842 + }, + { + "epoch": 3.1183759398496242, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 56843 + }, + { + "epoch": 3.118436090225564, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 56844 + }, + { + "epoch": 3.1184962406015035, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 56845 + }, + { + "epoch": 3.1185563909774436, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 56846 + }, + { + "epoch": 3.1186165413533833, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 56847 + }, + { + "epoch": 3.1186766917293234, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 56848 + }, + { + "epoch": 3.118736842105263, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 56849 + }, + { + "epoch": 3.118796992481203, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 56850 + }, + { + "epoch": 3.1188571428571428, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 56851 + }, + { + "epoch": 3.118917293233083, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 56852 + }, + { + "epoch": 3.1189774436090225, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 56853 + }, + { + "epoch": 3.1190375939849626, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 56854 + }, + { + "epoch": 3.1190977443609023, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 56855 + }, + { + "epoch": 3.119157894736842, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 56856 + }, + { + "epoch": 3.119218045112782, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 56857 + }, + { + "epoch": 3.1192781954887217, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 56858 + }, + { + "epoch": 3.1193383458646617, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 56859 + }, + { + "epoch": 3.1193984962406014, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 56860 + }, + { + "epoch": 3.1194586466165415, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 56861 + }, + { + "epoch": 3.119518796992481, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.8667, + "step": 56862 + }, + { + "epoch": 3.1195789473684212, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 56863 + }, + { + "epoch": 3.119639097744361, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 56864 + }, + { + "epoch": 3.119699248120301, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 56865 + }, + { + "epoch": 3.1197593984962406, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 56866 + }, + { + "epoch": 3.1198195488721803, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 56867 + }, + { + "epoch": 3.1198796992481204, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 56868 + }, + { + "epoch": 3.11993984962406, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 56869 + }, + { + "epoch": 3.12, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 56870 + }, + { + "epoch": 3.1200601503759398, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 56871 + }, + { + "epoch": 3.12012030075188, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 56872 + }, + { + "epoch": 3.1201804511278195, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 56873 + }, + { + "epoch": 3.1202406015037596, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 56874 + }, + { + "epoch": 3.1203007518796992, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1467, + "step": 56875 + }, + { + "epoch": 3.120360902255639, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 56876 + }, + { + "epoch": 3.120421052631579, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 56877 + }, + { + "epoch": 3.1204812030075186, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.8318, + "step": 56878 + }, + { + "epoch": 3.1205413533834587, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9115, + "step": 56879 + }, + { + "epoch": 3.1206015037593984, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 56880 + }, + { + "epoch": 3.1206616541353385, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 56881 + }, + { + "epoch": 3.120721804511278, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 56882 + }, + { + "epoch": 3.120781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 56883 + }, + { + "epoch": 3.120842105263158, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 56884 + }, + { + "epoch": 3.120902255639098, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 56885 + }, + { + "epoch": 3.1209624060150376, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 56886 + }, + { + "epoch": 3.1210225563909773, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9349, + "step": 56887 + }, + { + "epoch": 3.1210827067669173, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 56888 + }, + { + "epoch": 3.121142857142857, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 56889 + }, + { + "epoch": 3.121203007518797, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.8701, + "step": 56890 + }, + { + "epoch": 3.1212631578947367, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1274, + "step": 56891 + }, + { + "epoch": 3.121323308270677, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 56892 + }, + { + "epoch": 3.1213834586466165, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 56893 + }, + { + "epoch": 3.1214436090225566, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 56894 + }, + { + "epoch": 3.1215037593984962, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 56895 + }, + { + "epoch": 3.1215639097744363, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 56896 + }, + { + "epoch": 3.121624060150376, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 56897 + }, + { + "epoch": 3.1216842105263156, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 56898 + }, + { + "epoch": 3.1217443609022557, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 56899 + }, + { + "epoch": 3.1218045112781954, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 56900 + }, + { + "epoch": 3.1218646616541355, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.8964, + "step": 56901 + }, + { + "epoch": 3.121924812030075, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 56902 + }, + { + "epoch": 3.121984962406015, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 56903 + }, + { + "epoch": 3.122045112781955, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 56904 + }, + { + "epoch": 3.122105263157895, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9017, + "step": 56905 + }, + { + "epoch": 3.1221654135338346, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 56906 + }, + { + "epoch": 3.1222255639097742, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 56907 + }, + { + "epoch": 3.1222857142857143, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9161, + "step": 56908 + }, + { + "epoch": 3.122345864661654, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1536, + "step": 56909 + }, + { + "epoch": 3.122406015037594, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.944, + "step": 56910 + }, + { + "epoch": 3.1224661654135337, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 56911 + }, + { + "epoch": 3.122526315789474, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 56912 + }, + { + "epoch": 3.1225864661654135, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 56913 + }, + { + "epoch": 3.1226466165413536, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 56914 + }, + { + "epoch": 3.122706766917293, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 56915 + }, + { + "epoch": 3.122766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9201, + "step": 56916 + }, + { + "epoch": 3.122827067669173, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1133, + "step": 56917 + }, + { + "epoch": 3.1228872180451126, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 56918 + }, + { + "epoch": 3.1229473684210527, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 56919 + }, + { + "epoch": 3.1230075187969923, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 56920 + }, + { + "epoch": 3.1230676691729324, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9306, + "step": 56921 + }, + { + "epoch": 3.123127819548872, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 56922 + }, + { + "epoch": 3.123187969924812, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 56923 + }, + { + "epoch": 3.123248120300752, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 56924 + }, + { + "epoch": 3.123308270676692, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 56925 + }, + { + "epoch": 3.1233684210526316, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9139, + "step": 56926 + }, + { + "epoch": 3.123428571428571, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 56927 + }, + { + "epoch": 3.1234887218045113, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 56928 + }, + { + "epoch": 3.123548872180451, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 56929 + }, + { + "epoch": 3.123609022556391, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 56930 + }, + { + "epoch": 3.1236691729323307, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 56931 + }, + { + "epoch": 3.123729323308271, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 56932 + }, + { + "epoch": 3.1237894736842104, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 56933 + }, + { + "epoch": 3.1238496240601505, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 56934 + }, + { + "epoch": 3.12390977443609, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 56935 + }, + { + "epoch": 3.1239699248120303, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 56936 + }, + { + "epoch": 3.12403007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 56937 + }, + { + "epoch": 3.1240902255639096, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 56938 + }, + { + "epoch": 3.1241503759398497, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 56939 + }, + { + "epoch": 3.1242105263157893, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 56940 + }, + { + "epoch": 3.1242706766917294, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 56941 + }, + { + "epoch": 3.124330827067669, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9282, + "step": 56942 + }, + { + "epoch": 3.124390977443609, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 56943 + }, + { + "epoch": 3.124451127819549, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 56944 + }, + { + "epoch": 3.124511278195489, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 56945 + }, + { + "epoch": 3.1245714285714286, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 56946 + }, + { + "epoch": 3.124631578947368, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0487, + "step": 56947 + }, + { + "epoch": 3.1246917293233083, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 56948 + }, + { + "epoch": 3.124751879699248, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 56949 + }, + { + "epoch": 3.124812030075188, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 56950 + }, + { + "epoch": 3.1248721804511277, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 56951 + }, + { + "epoch": 3.124932330827068, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 56952 + }, + { + "epoch": 3.1249924812030074, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 56953 + }, + { + "epoch": 3.1250526315789475, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 56954 + }, + { + "epoch": 3.125112781954887, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 56955 + }, + { + "epoch": 3.1251729323308273, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 56956 + }, + { + "epoch": 3.125233082706767, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 56957 + }, + { + "epoch": 3.1252932330827066, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 56958 + }, + { + "epoch": 3.1253533834586467, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9421, + "step": 56959 + }, + { + "epoch": 3.1254135338345863, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 56960 + }, + { + "epoch": 3.1254736842105264, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 56961 + }, + { + "epoch": 3.125533834586466, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1639, + "step": 56962 + }, + { + "epoch": 3.125593984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 56963 + }, + { + "epoch": 3.125654135338346, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9348, + "step": 56964 + }, + { + "epoch": 3.125714285714286, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 56965 + }, + { + "epoch": 3.1257744360902255, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 56966 + }, + { + "epoch": 3.1258345864661656, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 56967 + }, + { + "epoch": 3.1258947368421053, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.16, + "step": 56968 + }, + { + "epoch": 3.125954887218045, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 56969 + }, + { + "epoch": 3.126015037593985, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 56970 + }, + { + "epoch": 3.1260751879699247, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 56971 + }, + { + "epoch": 3.1261353383458648, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 56972 + }, + { + "epoch": 3.1261954887218044, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 56973 + }, + { + "epoch": 3.1262556390977445, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 56974 + }, + { + "epoch": 3.126315789473684, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 56975 + }, + { + "epoch": 3.1263759398496243, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 56976 + }, + { + "epoch": 3.126436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 56977 + }, + { + "epoch": 3.1264962406015036, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 56978 + }, + { + "epoch": 3.1265563909774436, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.918, + "step": 56979 + }, + { + "epoch": 3.1266165413533833, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 56980 + }, + { + "epoch": 3.1266766917293234, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 56981 + }, + { + "epoch": 3.126736842105263, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 56982 + }, + { + "epoch": 3.126796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 56983 + }, + { + "epoch": 3.126857142857143, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 56984 + }, + { + "epoch": 3.126917293233083, + "grad_norm": 0.478515625, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 56985 + }, + { + "epoch": 3.1269774436090225, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 56986 + }, + { + "epoch": 3.127037593984962, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 56987 + }, + { + "epoch": 3.1270977443609023, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9393, + "step": 56988 + }, + { + "epoch": 3.127157894736842, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 56989 + }, + { + "epoch": 3.127218045112782, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 56990 + }, + { + "epoch": 3.1272781954887217, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 56991 + }, + { + "epoch": 3.1273383458646618, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 56992 + }, + { + "epoch": 3.1273984962406014, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 56993 + }, + { + "epoch": 3.1274586466165415, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 56994 + }, + { + "epoch": 3.127518796992481, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 56995 + }, + { + "epoch": 3.1275789473684212, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 56996 + }, + { + "epoch": 3.127639097744361, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 56997 + }, + { + "epoch": 3.127699248120301, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 56998 + }, + { + "epoch": 3.1277593984962406, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 56999 + }, + { + "epoch": 3.1278195488721803, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9158, + "step": 57000 + }, + { + "epoch": 3.1278796992481204, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 57001 + }, + { + "epoch": 3.12793984962406, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 57002 + }, + { + "epoch": 3.128, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 57003 + }, + { + "epoch": 3.1280601503759398, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 57004 + }, + { + "epoch": 3.12812030075188, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 57005 + }, + { + "epoch": 3.1281804511278195, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 57006 + }, + { + "epoch": 3.1282406015037596, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9152, + "step": 57007 + }, + { + "epoch": 3.1283007518796992, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 57008 + }, + { + "epoch": 3.128360902255639, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 57009 + }, + { + "epoch": 3.128421052631579, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0776, + "step": 57010 + }, + { + "epoch": 3.1284812030075186, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 57011 + }, + { + "epoch": 3.1285413533834587, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 57012 + }, + { + "epoch": 3.1286015037593984, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 57013 + }, + { + "epoch": 3.1286616541353385, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 57014 + }, + { + "epoch": 3.128721804511278, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9337, + "step": 57015 + }, + { + "epoch": 3.128781954887218, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 57016 + }, + { + "epoch": 3.128842105263158, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 57017 + }, + { + "epoch": 3.1289022556390975, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 57018 + }, + { + "epoch": 3.1289624060150376, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 57019 + }, + { + "epoch": 3.1290225563909773, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 57020 + }, + { + "epoch": 3.1290827067669174, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 57021 + }, + { + "epoch": 3.129142857142857, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 57022 + }, + { + "epoch": 3.129203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 57023 + }, + { + "epoch": 3.1292631578947367, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 57024 + }, + { + "epoch": 3.129323308270677, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 57025 + }, + { + "epoch": 3.1293834586466165, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 57026 + }, + { + "epoch": 3.1294436090225566, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 57027 + }, + { + "epoch": 3.1295037593984962, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 57028 + }, + { + "epoch": 3.1295639097744363, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 57029 + }, + { + "epoch": 3.129624060150376, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9469, + "step": 57030 + }, + { + "epoch": 3.1296842105263156, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 57031 + }, + { + "epoch": 3.1297443609022557, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 57032 + }, + { + "epoch": 3.1298045112781954, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 57033 + }, + { + "epoch": 3.1298646616541355, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9322, + "step": 57034 + }, + { + "epoch": 3.129924812030075, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9316, + "step": 57035 + }, + { + "epoch": 3.129984962406015, + "grad_norm": 1.53125, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 57036 + }, + { + "epoch": 3.130045112781955, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0617, + "step": 57037 + }, + { + "epoch": 3.130105263157895, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 57038 + }, + { + "epoch": 3.1301654135338346, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 57039 + }, + { + "epoch": 3.1302255639097742, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 57040 + }, + { + "epoch": 3.1302857142857143, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 57041 + }, + { + "epoch": 3.130345864661654, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 57042 + }, + { + "epoch": 3.130406015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 57043 + }, + { + "epoch": 3.1304661654135337, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 57044 + }, + { + "epoch": 3.130526315789474, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 57045 + }, + { + "epoch": 3.1305864661654135, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 57046 + }, + { + "epoch": 3.1306466165413536, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 57047 + }, + { + "epoch": 3.130706766917293, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 57048 + }, + { + "epoch": 3.130766917293233, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 57049 + }, + { + "epoch": 3.130827067669173, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 57050 + }, + { + "epoch": 3.1308872180451126, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 57051 + }, + { + "epoch": 3.1309473684210527, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9378, + "step": 57052 + }, + { + "epoch": 3.1310075187969924, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 57053 + }, + { + "epoch": 3.1310676691729324, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 57054 + }, + { + "epoch": 3.131127819548872, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 57055 + }, + { + "epoch": 3.131187969924812, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 57056 + }, + { + "epoch": 3.131248120300752, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 57057 + }, + { + "epoch": 3.131308270676692, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 57058 + }, + { + "epoch": 3.1313684210526316, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 57059 + }, + { + "epoch": 3.1314285714285712, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 57060 + }, + { + "epoch": 3.1314887218045113, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 57061 + }, + { + "epoch": 3.131548872180451, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 57062 + }, + { + "epoch": 3.131609022556391, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 57063 + }, + { + "epoch": 3.1316691729323307, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 57064 + }, + { + "epoch": 3.131729323308271, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 57065 + }, + { + "epoch": 3.1317894736842105, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 57066 + }, + { + "epoch": 3.1318496240601505, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 57067 + }, + { + "epoch": 3.13190977443609, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 57068 + }, + { + "epoch": 3.1319699248120303, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0758, + "step": 57069 + }, + { + "epoch": 3.13203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 57070 + }, + { + "epoch": 3.1320902255639096, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 57071 + }, + { + "epoch": 3.1321503759398497, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 57072 + }, + { + "epoch": 3.1322105263157893, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 57073 + }, + { + "epoch": 3.1322706766917294, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 57074 + }, + { + "epoch": 3.132330827067669, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1183, + "step": 57075 + }, + { + "epoch": 3.132390977443609, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 57076 + }, + { + "epoch": 3.132451127819549, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 57077 + }, + { + "epoch": 3.132511278195489, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 57078 + }, + { + "epoch": 3.1325714285714286, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 57079 + }, + { + "epoch": 3.132631578947368, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1014, + "step": 57080 + }, + { + "epoch": 3.1326917293233083, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9941, + "step": 57081 + }, + { + "epoch": 3.132751879699248, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 57082 + }, + { + "epoch": 3.132812030075188, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 57083 + }, + { + "epoch": 3.1328721804511277, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 57084 + }, + { + "epoch": 3.132932330827068, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 57085 + }, + { + "epoch": 3.1329924812030074, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 57086 + }, + { + "epoch": 3.1330526315789475, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 57087 + }, + { + "epoch": 3.133112781954887, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 57088 + }, + { + "epoch": 3.133172932330827, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 57089 + }, + { + "epoch": 3.133233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 57090 + }, + { + "epoch": 3.1332932330827066, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.8932, + "step": 57091 + }, + { + "epoch": 3.1333533834586467, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 57092 + }, + { + "epoch": 3.1334135338345863, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 57093 + }, + { + "epoch": 3.1334736842105264, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 57094 + }, + { + "epoch": 3.133533834586466, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 57095 + }, + { + "epoch": 3.133593984962406, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 57096 + }, + { + "epoch": 3.133654135338346, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 57097 + }, + { + "epoch": 3.133714285714286, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0812, + "step": 57098 + }, + { + "epoch": 3.1337744360902255, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 57099 + }, + { + "epoch": 3.1338345864661656, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 57100 + }, + { + "epoch": 3.1338947368421053, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 57101 + }, + { + "epoch": 3.133954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 57102 + }, + { + "epoch": 3.134015037593985, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 57103 + }, + { + "epoch": 3.1340751879699247, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 57104 + }, + { + "epoch": 3.1341353383458648, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 57105 + }, + { + "epoch": 3.1341954887218044, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 57106 + }, + { + "epoch": 3.1342556390977445, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2024, + "step": 57107 + }, + { + "epoch": 3.134315789473684, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 57108 + }, + { + "epoch": 3.1343759398496243, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 57109 + }, + { + "epoch": 3.134436090225564, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 57110 + }, + { + "epoch": 3.1344962406015036, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 57111 + }, + { + "epoch": 3.1345563909774437, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 57112 + }, + { + "epoch": 3.1346165413533833, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 57113 + }, + { + "epoch": 3.1346766917293234, + "grad_norm": 1.3671875, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 57114 + }, + { + "epoch": 3.134736842105263, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 57115 + }, + { + "epoch": 3.134796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 57116 + }, + { + "epoch": 3.134857142857143, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 57117 + }, + { + "epoch": 3.134917293233083, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 57118 + }, + { + "epoch": 3.1349774436090225, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 57119 + }, + { + "epoch": 3.135037593984962, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 57120 + }, + { + "epoch": 3.1350977443609023, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 57121 + }, + { + "epoch": 3.135157894736842, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 57122 + }, + { + "epoch": 3.135218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.1012, + "step": 57123 + }, + { + "epoch": 3.1352781954887217, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 57124 + }, + { + "epoch": 3.1353383458646618, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 57125 + }, + { + "epoch": 3.1353984962406014, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 57126 + }, + { + "epoch": 3.1354586466165415, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 57127 + }, + { + "epoch": 3.135518796992481, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 57128 + }, + { + "epoch": 3.1355789473684212, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 57129 + }, + { + "epoch": 3.135639097744361, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 57130 + }, + { + "epoch": 3.135699248120301, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 57131 + }, + { + "epoch": 3.1357593984962406, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 57132 + }, + { + "epoch": 3.1358195488721803, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 57133 + }, + { + "epoch": 3.1358796992481204, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.8374, + "step": 57134 + }, + { + "epoch": 3.13593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.933, + "step": 57135 + }, + { + "epoch": 3.136, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 57136 + }, + { + "epoch": 3.1360601503759398, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1071, + "step": 57137 + }, + { + "epoch": 3.13612030075188, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 57138 + }, + { + "epoch": 3.1361804511278195, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 57139 + }, + { + "epoch": 3.1362406015037596, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 57140 + }, + { + "epoch": 3.1363007518796993, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 57141 + }, + { + "epoch": 3.136360902255639, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 57142 + }, + { + "epoch": 3.136421052631579, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 57143 + }, + { + "epoch": 3.1364812030075186, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 57144 + }, + { + "epoch": 3.1365413533834587, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 57145 + }, + { + "epoch": 3.1366015037593984, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 57146 + }, + { + "epoch": 3.1366616541353385, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 57147 + }, + { + "epoch": 3.136721804511278, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 57148 + }, + { + "epoch": 3.1367819548872182, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 57149 + }, + { + "epoch": 3.136842105263158, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 57150 + }, + { + "epoch": 3.1369022556390975, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9282, + "step": 57151 + }, + { + "epoch": 3.1369624060150376, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9146, + "step": 57152 + }, + { + "epoch": 3.1370225563909773, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 57153 + }, + { + "epoch": 3.1370827067669174, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 57154 + }, + { + "epoch": 3.137142857142857, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 57155 + }, + { + "epoch": 3.137203007518797, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 57156 + }, + { + "epoch": 3.1372631578947368, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 57157 + }, + { + "epoch": 3.137323308270677, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.8819, + "step": 57158 + }, + { + "epoch": 3.1373834586466165, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 57159 + }, + { + "epoch": 3.1374436090225566, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 57160 + }, + { + "epoch": 3.1375037593984962, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 57161 + }, + { + "epoch": 3.137563909774436, + "grad_norm": 1.421875, + "learning_rate": 3e-05, + "loss": 1.0351, + "step": 57162 + }, + { + "epoch": 3.137624060150376, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9299, + "step": 57163 + }, + { + "epoch": 3.1376842105263156, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 57164 + }, + { + "epoch": 3.1377443609022557, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 57165 + }, + { + "epoch": 3.1378045112781954, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 57166 + }, + { + "epoch": 3.1378646616541355, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 57167 + }, + { + "epoch": 3.137924812030075, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 57168 + }, + { + "epoch": 3.137984962406015, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 57169 + }, + { + "epoch": 3.138045112781955, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 57170 + }, + { + "epoch": 3.138105263157895, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 57171 + }, + { + "epoch": 3.1381654135338346, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 57172 + }, + { + "epoch": 3.1382255639097743, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 57173 + }, + { + "epoch": 3.1382857142857143, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 57174 + }, + { + "epoch": 3.138345864661654, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 57175 + }, + { + "epoch": 3.138406015037594, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 57176 + }, + { + "epoch": 3.1384661654135337, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 57177 + }, + { + "epoch": 3.138526315789474, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 57178 + }, + { + "epoch": 3.1385864661654135, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 57179 + }, + { + "epoch": 3.1386466165413536, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 57180 + }, + { + "epoch": 3.1387067669172932, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 57181 + }, + { + "epoch": 3.138766917293233, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 57182 + }, + { + "epoch": 3.138827067669173, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 57183 + }, + { + "epoch": 3.1388872180451126, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 57184 + }, + { + "epoch": 3.1389473684210527, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0675, + "step": 57185 + }, + { + "epoch": 3.1390075187969924, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 57186 + }, + { + "epoch": 3.1390676691729325, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 57187 + }, + { + "epoch": 3.139127819548872, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 57188 + }, + { + "epoch": 3.139187969924812, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1887, + "step": 57189 + }, + { + "epoch": 3.139248120300752, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 57190 + }, + { + "epoch": 3.139308270676692, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 57191 + }, + { + "epoch": 3.1393684210526316, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 57192 + }, + { + "epoch": 3.1394285714285712, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 57193 + }, + { + "epoch": 3.1394887218045113, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9371, + "step": 57194 + }, + { + "epoch": 3.139548872180451, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 57195 + }, + { + "epoch": 3.139609022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9012, + "step": 57196 + }, + { + "epoch": 3.1396691729323307, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 57197 + }, + { + "epoch": 3.139729323308271, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 57198 + }, + { + "epoch": 3.1397894736842105, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 57199 + }, + { + "epoch": 3.1398496240601506, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 57200 + }, + { + "epoch": 3.13990977443609, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 57201 + }, + { + "epoch": 3.1399699248120303, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1334, + "step": 57202 + }, + { + "epoch": 3.14003007518797, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 57203 + }, + { + "epoch": 3.1400902255639096, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 57204 + }, + { + "epoch": 3.1401503759398497, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 57205 + }, + { + "epoch": 3.1402105263157893, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 57206 + }, + { + "epoch": 3.1402706766917294, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 57207 + }, + { + "epoch": 3.140330827067669, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 57208 + }, + { + "epoch": 3.140390977443609, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 57209 + }, + { + "epoch": 3.140451127819549, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 57210 + }, + { + "epoch": 3.140511278195489, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 57211 + }, + { + "epoch": 3.1405714285714286, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9155, + "step": 57212 + }, + { + "epoch": 3.140631578947368, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 57213 + }, + { + "epoch": 3.1406917293233083, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 57214 + }, + { + "epoch": 3.140751879699248, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 57215 + }, + { + "epoch": 3.140812030075188, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 57216 + }, + { + "epoch": 3.1408721804511277, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 57217 + }, + { + "epoch": 3.140932330827068, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 57218 + }, + { + "epoch": 3.1409924812030074, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 57219 + }, + { + "epoch": 3.1410526315789475, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 57220 + }, + { + "epoch": 3.141112781954887, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 57221 + }, + { + "epoch": 3.141172932330827, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 57222 + }, + { + "epoch": 3.141233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 57223 + }, + { + "epoch": 3.1412932330827066, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9236, + "step": 57224 + }, + { + "epoch": 3.1413533834586467, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 57225 + }, + { + "epoch": 3.1414135338345863, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 57226 + }, + { + "epoch": 3.1414736842105264, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 57227 + }, + { + "epoch": 3.141533834586466, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.827, + "step": 57228 + }, + { + "epoch": 3.141593984962406, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.922, + "step": 57229 + }, + { + "epoch": 3.141654135338346, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 57230 + }, + { + "epoch": 3.141714285714286, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 57231 + }, + { + "epoch": 3.1417744360902256, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9287, + "step": 57232 + }, + { + "epoch": 3.1418345864661656, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.8987, + "step": 57233 + }, + { + "epoch": 3.1418947368421053, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 57234 + }, + { + "epoch": 3.141954887218045, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 57235 + }, + { + "epoch": 3.142015037593985, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 57236 + }, + { + "epoch": 3.1420751879699247, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 57237 + }, + { + "epoch": 3.142135338345865, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 57238 + }, + { + "epoch": 3.1421954887218044, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 57239 + }, + { + "epoch": 3.1422556390977445, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 57240 + }, + { + "epoch": 3.142315789473684, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9161, + "step": 57241 + }, + { + "epoch": 3.1423759398496243, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 57242 + }, + { + "epoch": 3.142436090225564, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 57243 + }, + { + "epoch": 3.1424962406015036, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 57244 + }, + { + "epoch": 3.1425563909774437, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 57245 + }, + { + "epoch": 3.1426165413533833, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 57246 + }, + { + "epoch": 3.1426766917293234, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 57247 + }, + { + "epoch": 3.142736842105263, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 57248 + }, + { + "epoch": 3.142796992481203, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0737, + "step": 57249 + }, + { + "epoch": 3.142857142857143, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 57250 + }, + { + "epoch": 3.142917293233083, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9092, + "step": 57251 + }, + { + "epoch": 3.1429774436090225, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 57252 + }, + { + "epoch": 3.143037593984962, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9191, + "step": 57253 + }, + { + "epoch": 3.1430977443609023, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 57254 + }, + { + "epoch": 3.143157894736842, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 57255 + }, + { + "epoch": 3.143218045112782, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 57256 + }, + { + "epoch": 3.1432781954887217, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 57257 + }, + { + "epoch": 3.1433383458646618, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 57258 + }, + { + "epoch": 3.1433984962406014, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 57259 + }, + { + "epoch": 3.1434586466165415, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 57260 + }, + { + "epoch": 3.143518796992481, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 57261 + }, + { + "epoch": 3.1435789473684213, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 57262 + }, + { + "epoch": 3.143639097744361, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 57263 + }, + { + "epoch": 3.1436992481203005, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9394, + "step": 57264 + }, + { + "epoch": 3.1437593984962406, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 57265 + }, + { + "epoch": 3.1438195488721803, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.2149, + "step": 57266 + }, + { + "epoch": 3.1438796992481204, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 57267 + }, + { + "epoch": 3.14393984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9438, + "step": 57268 + }, + { + "epoch": 3.144, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 57269 + }, + { + "epoch": 3.1440601503759398, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 57270 + }, + { + "epoch": 3.14412030075188, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 57271 + }, + { + "epoch": 3.1441804511278195, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 57272 + }, + { + "epoch": 3.1442406015037596, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 57273 + }, + { + "epoch": 3.1443007518796993, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 57274 + }, + { + "epoch": 3.144360902255639, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1327, + "step": 57275 + }, + { + "epoch": 3.144421052631579, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.8985, + "step": 57276 + }, + { + "epoch": 3.1444812030075187, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 57277 + }, + { + "epoch": 3.1445413533834587, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 57278 + }, + { + "epoch": 3.1446015037593984, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 57279 + }, + { + "epoch": 3.1446616541353385, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9148, + "step": 57280 + }, + { + "epoch": 3.144721804511278, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 57281 + }, + { + "epoch": 3.1447819548872182, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.924, + "step": 57282 + }, + { + "epoch": 3.144842105263158, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 57283 + }, + { + "epoch": 3.1449022556390975, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 57284 + }, + { + "epoch": 3.1449624060150376, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 57285 + }, + { + "epoch": 3.1450225563909773, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 57286 + }, + { + "epoch": 3.1450827067669174, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 57287 + }, + { + "epoch": 3.145142857142857, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 57288 + }, + { + "epoch": 3.145203007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 57289 + }, + { + "epoch": 3.1452631578947368, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 57290 + }, + { + "epoch": 3.145323308270677, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 57291 + }, + { + "epoch": 3.1453834586466165, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.969, + "step": 57292 + }, + { + "epoch": 3.1454436090225566, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 57293 + }, + { + "epoch": 3.1455037593984962, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 57294 + }, + { + "epoch": 3.145563909774436, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 57295 + }, + { + "epoch": 3.145624060150376, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 57296 + }, + { + "epoch": 3.1456842105263156, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 57297 + }, + { + "epoch": 3.1457443609022557, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 57298 + }, + { + "epoch": 3.1458045112781954, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 57299 + }, + { + "epoch": 3.1458646616541355, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 57300 + }, + { + "epoch": 3.145924812030075, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 57301 + }, + { + "epoch": 3.145984962406015, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9261, + "step": 57302 + }, + { + "epoch": 3.146045112781955, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 57303 + }, + { + "epoch": 3.146105263157895, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 57304 + }, + { + "epoch": 3.1461654135338346, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1018, + "step": 57305 + }, + { + "epoch": 3.1462255639097743, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 57306 + }, + { + "epoch": 3.1462857142857144, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 57307 + }, + { + "epoch": 3.146345864661654, + "grad_norm": 1.40625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 57308 + }, + { + "epoch": 3.146406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 57309 + }, + { + "epoch": 3.1464661654135337, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 57310 + }, + { + "epoch": 3.146526315789474, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 57311 + }, + { + "epoch": 3.1465864661654135, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 57312 + }, + { + "epoch": 3.1466466165413536, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 57313 + }, + { + "epoch": 3.1467067669172932, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 57314 + }, + { + "epoch": 3.146766917293233, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9491, + "step": 57315 + }, + { + "epoch": 3.146827067669173, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 57316 + }, + { + "epoch": 3.1468872180451126, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 57317 + }, + { + "epoch": 3.1469473684210527, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 57318 + }, + { + "epoch": 3.1470075187969924, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 57319 + }, + { + "epoch": 3.1470676691729325, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 57320 + }, + { + "epoch": 3.147127819548872, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 57321 + }, + { + "epoch": 3.147187969924812, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.109, + "step": 57322 + }, + { + "epoch": 3.147248120300752, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 57323 + }, + { + "epoch": 3.1473082706766915, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9511, + "step": 57324 + }, + { + "epoch": 3.1473684210526316, + "grad_norm": 1.390625, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 57325 + }, + { + "epoch": 3.1474285714285712, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 57326 + }, + { + "epoch": 3.1474887218045113, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 57327 + }, + { + "epoch": 3.147548872180451, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 57328 + }, + { + "epoch": 3.147609022556391, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 57329 + }, + { + "epoch": 3.1476691729323307, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 57330 + }, + { + "epoch": 3.147729323308271, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 57331 + }, + { + "epoch": 3.1477894736842105, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 57332 + }, + { + "epoch": 3.1478496240601506, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 57333 + }, + { + "epoch": 3.14790977443609, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 57334 + }, + { + "epoch": 3.1479699248120303, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9172, + "step": 57335 + }, + { + "epoch": 3.14803007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 57336 + }, + { + "epoch": 3.1480902255639096, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 57337 + }, + { + "epoch": 3.1481503759398497, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 57338 + }, + { + "epoch": 3.1482105263157893, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 57339 + }, + { + "epoch": 3.1482706766917294, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 57340 + }, + { + "epoch": 3.148330827067669, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 57341 + }, + { + "epoch": 3.148390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 57342 + }, + { + "epoch": 3.148451127819549, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 57343 + }, + { + "epoch": 3.148511278195489, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 57344 + }, + { + "epoch": 3.1485714285714286, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 57345 + }, + { + "epoch": 3.1486315789473682, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 57346 + }, + { + "epoch": 3.1486917293233083, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 57347 + }, + { + "epoch": 3.148751879699248, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 57348 + }, + { + "epoch": 3.148812030075188, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0506, + "step": 57349 + }, + { + "epoch": 3.1488721804511277, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 57350 + }, + { + "epoch": 3.148932330827068, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 57351 + }, + { + "epoch": 3.1489924812030075, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 57352 + }, + { + "epoch": 3.1490526315789475, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9097, + "step": 57353 + }, + { + "epoch": 3.149112781954887, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 57354 + }, + { + "epoch": 3.149172932330827, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9408, + "step": 57355 + }, + { + "epoch": 3.149233082706767, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9378, + "step": 57356 + }, + { + "epoch": 3.1492932330827066, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 57357 + }, + { + "epoch": 3.1493533834586467, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 57358 + }, + { + "epoch": 3.1494135338345863, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 57359 + }, + { + "epoch": 3.1494736842105264, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 57360 + }, + { + "epoch": 3.149533834586466, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 57361 + }, + { + "epoch": 3.149593984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 57362 + }, + { + "epoch": 3.149654135338346, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 57363 + }, + { + "epoch": 3.149714285714286, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 57364 + }, + { + "epoch": 3.1497744360902256, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 57365 + }, + { + "epoch": 3.149834586466165, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 57366 + }, + { + "epoch": 3.1498947368421053, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9706, + "step": 57367 + }, + { + "epoch": 3.149954887218045, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 57368 + }, + { + "epoch": 3.150015037593985, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 57369 + }, + { + "epoch": 3.1500751879699247, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 57370 + }, + { + "epoch": 3.150135338345865, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 57371 + }, + { + "epoch": 3.1501954887218044, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 57372 + }, + { + "epoch": 3.1502556390977445, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9442, + "step": 57373 + }, + { + "epoch": 3.150315789473684, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 57374 + }, + { + "epoch": 3.1503759398496243, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 57375 + }, + { + "epoch": 3.150436090225564, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 57376 + }, + { + "epoch": 3.1504962406015036, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 57377 + }, + { + "epoch": 3.1505563909774437, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 57378 + }, + { + "epoch": 3.1506165413533833, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 57379 + }, + { + "epoch": 3.1506766917293234, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9041, + "step": 57380 + }, + { + "epoch": 3.150736842105263, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0825, + "step": 57381 + }, + { + "epoch": 3.150796992481203, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9511, + "step": 57382 + }, + { + "epoch": 3.150857142857143, + "grad_norm": 0.498046875, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 57383 + }, + { + "epoch": 3.150917293233083, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 57384 + }, + { + "epoch": 3.1509774436090225, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 57385 + }, + { + "epoch": 3.151037593984962, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 57386 + }, + { + "epoch": 3.1510977443609023, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 57387 + }, + { + "epoch": 3.151157894736842, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 57388 + }, + { + "epoch": 3.151218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 57389 + }, + { + "epoch": 3.1512781954887217, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 57390 + }, + { + "epoch": 3.1513383458646618, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 57391 + }, + { + "epoch": 3.1513984962406014, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 57392 + }, + { + "epoch": 3.1514586466165415, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 57393 + }, + { + "epoch": 3.151518796992481, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 57394 + }, + { + "epoch": 3.1515789473684213, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9193, + "step": 57395 + }, + { + "epoch": 3.151639097744361, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 57396 + }, + { + "epoch": 3.1516992481203006, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 57397 + }, + { + "epoch": 3.1517593984962406, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 57398 + }, + { + "epoch": 3.1518195488721803, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 57399 + }, + { + "epoch": 3.1518796992481204, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 57400 + }, + { + "epoch": 3.15193984962406, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 57401 + }, + { + "epoch": 3.152, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.939, + "step": 57402 + }, + { + "epoch": 3.15206015037594, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 57403 + }, + { + "epoch": 3.15212030075188, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 57404 + }, + { + "epoch": 3.1521804511278195, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9445, + "step": 57405 + }, + { + "epoch": 3.1522406015037596, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 57406 + }, + { + "epoch": 3.1523007518796993, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 57407 + }, + { + "epoch": 3.152360902255639, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 57408 + }, + { + "epoch": 3.152421052631579, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 57409 + }, + { + "epoch": 3.1524812030075187, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 57410 + }, + { + "epoch": 3.1525413533834588, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 57411 + }, + { + "epoch": 3.1526015037593984, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 57412 + }, + { + "epoch": 3.1526616541353385, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 57413 + }, + { + "epoch": 3.152721804511278, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 57414 + }, + { + "epoch": 3.1527819548872182, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 57415 + }, + { + "epoch": 3.152842105263158, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 57416 + }, + { + "epoch": 3.1529022556390975, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 57417 + }, + { + "epoch": 3.1529624060150376, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 57418 + }, + { + "epoch": 3.1530225563909773, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 57419 + }, + { + "epoch": 3.1530827067669174, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 57420 + }, + { + "epoch": 3.153142857142857, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 57421 + }, + { + "epoch": 3.153203007518797, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9537, + "step": 57422 + }, + { + "epoch": 3.1532631578947368, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 57423 + }, + { + "epoch": 3.153323308270677, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 57424 + }, + { + "epoch": 3.1533834586466165, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 57425 + }, + { + "epoch": 3.153443609022556, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0808, + "step": 57426 + }, + { + "epoch": 3.1535037593984963, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 57427 + }, + { + "epoch": 3.153563909774436, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 57428 + }, + { + "epoch": 3.153624060150376, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 57429 + }, + { + "epoch": 3.1536842105263156, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 57430 + }, + { + "epoch": 3.1537443609022557, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1242, + "step": 57431 + }, + { + "epoch": 3.1538045112781954, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 57432 + }, + { + "epoch": 3.1538646616541355, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1626, + "step": 57433 + }, + { + "epoch": 3.153924812030075, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9268, + "step": 57434 + }, + { + "epoch": 3.1539849624060152, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 57435 + }, + { + "epoch": 3.154045112781955, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 57436 + }, + { + "epoch": 3.154105263157895, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 57437 + }, + { + "epoch": 3.1541654135338346, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9212, + "step": 57438 + }, + { + "epoch": 3.1542255639097743, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 57439 + }, + { + "epoch": 3.1542857142857144, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 57440 + }, + { + "epoch": 3.154345864661654, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 57441 + }, + { + "epoch": 3.154406015037594, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 57442 + }, + { + "epoch": 3.1544661654135338, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 57443 + }, + { + "epoch": 3.154526315789474, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 57444 + }, + { + "epoch": 3.1545864661654135, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 57445 + }, + { + "epoch": 3.1546466165413536, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 57446 + }, + { + "epoch": 3.1547067669172932, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.2215, + "step": 57447 + }, + { + "epoch": 3.154766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 57448 + }, + { + "epoch": 3.154827067669173, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 57449 + }, + { + "epoch": 3.1548872180451126, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 57450 + }, + { + "epoch": 3.1549473684210527, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 57451 + }, + { + "epoch": 3.1550075187969924, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 57452 + }, + { + "epoch": 3.1550676691729325, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 57453 + }, + { + "epoch": 3.155127819548872, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 57454 + }, + { + "epoch": 3.155187969924812, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 57455 + }, + { + "epoch": 3.155248120300752, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 57456 + }, + { + "epoch": 3.1553082706766915, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 57457 + }, + { + "epoch": 3.1553684210526316, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 57458 + }, + { + "epoch": 3.1554285714285712, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 57459 + }, + { + "epoch": 3.1554887218045113, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 57460 + }, + { + "epoch": 3.155548872180451, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 57461 + }, + { + "epoch": 3.155609022556391, + "grad_norm": 0.48046875, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 57462 + }, + { + "epoch": 3.1556691729323307, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 57463 + }, + { + "epoch": 3.155729323308271, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 57464 + }, + { + "epoch": 3.1557894736842105, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 57465 + }, + { + "epoch": 3.1558496240601506, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 57466 + }, + { + "epoch": 3.15590977443609, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 57467 + }, + { + "epoch": 3.15596992481203, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9225, + "step": 57468 + }, + { + "epoch": 3.15603007518797, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.935, + "step": 57469 + }, + { + "epoch": 3.1560902255639096, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 57470 + }, + { + "epoch": 3.1561503759398497, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 57471 + }, + { + "epoch": 3.1562105263157894, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 57472 + }, + { + "epoch": 3.1562706766917294, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 57473 + }, + { + "epoch": 3.156330827067669, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0987, + "step": 57474 + }, + { + "epoch": 3.156390977443609, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 57475 + }, + { + "epoch": 3.156451127819549, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 57476 + }, + { + "epoch": 3.156511278195489, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9235, + "step": 57477 + }, + { + "epoch": 3.1565714285714286, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 57478 + }, + { + "epoch": 3.1566315789473682, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 57479 + }, + { + "epoch": 3.1566917293233083, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 57480 + }, + { + "epoch": 3.156751879699248, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 57481 + }, + { + "epoch": 3.156812030075188, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 57482 + }, + { + "epoch": 3.1568721804511277, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 57483 + }, + { + "epoch": 3.156932330827068, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0671, + "step": 57484 + }, + { + "epoch": 3.1569924812030075, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 57485 + }, + { + "epoch": 3.1570526315789476, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 57486 + }, + { + "epoch": 3.157112781954887, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9434, + "step": 57487 + }, + { + "epoch": 3.157172932330827, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 57488 + }, + { + "epoch": 3.157233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 57489 + }, + { + "epoch": 3.1572932330827066, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 57490 + }, + { + "epoch": 3.1573533834586467, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 57491 + }, + { + "epoch": 3.1574135338345863, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 57492 + }, + { + "epoch": 3.1574736842105264, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 57493 + }, + { + "epoch": 3.157533834586466, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9077, + "step": 57494 + }, + { + "epoch": 3.157593984962406, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 57495 + }, + { + "epoch": 3.157654135338346, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 57496 + }, + { + "epoch": 3.157714285714286, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 57497 + }, + { + "epoch": 3.1577744360902256, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 57498 + }, + { + "epoch": 3.157834586466165, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 57499 + }, + { + "epoch": 3.1578947368421053, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 57500 + }, + { + "epoch": 3.157954887218045, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 57501 + }, + { + "epoch": 3.158015037593985, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 57502 + }, + { + "epoch": 3.1580751879699247, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 57503 + }, + { + "epoch": 3.158135338345865, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 57504 + }, + { + "epoch": 3.1581954887218044, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.2082, + "step": 57505 + }, + { + "epoch": 3.1582556390977445, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 57506 + }, + { + "epoch": 3.158315789473684, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 57507 + }, + { + "epoch": 3.1583759398496243, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.067, + "step": 57508 + }, + { + "epoch": 3.158436090225564, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 57509 + }, + { + "epoch": 3.1584962406015036, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 57510 + }, + { + "epoch": 3.1585563909774437, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 57511 + }, + { + "epoch": 3.1586165413533833, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 57512 + }, + { + "epoch": 3.1586766917293234, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 57513 + }, + { + "epoch": 3.158736842105263, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 57514 + }, + { + "epoch": 3.158796992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 57515 + }, + { + "epoch": 3.158857142857143, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 57516 + }, + { + "epoch": 3.158917293233083, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 57517 + }, + { + "epoch": 3.1589774436090226, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 57518 + }, + { + "epoch": 3.159037593984962, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 57519 + }, + { + "epoch": 3.1590977443609023, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 57520 + }, + { + "epoch": 3.159157894736842, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 57521 + }, + { + "epoch": 3.159218045112782, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 57522 + }, + { + "epoch": 3.1592781954887217, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 57523 + }, + { + "epoch": 3.159338345864662, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 57524 + }, + { + "epoch": 3.1593984962406014, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0244, + "step": 57525 + }, + { + "epoch": 3.1594586466165415, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 57526 + }, + { + "epoch": 3.159518796992481, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 57527 + }, + { + "epoch": 3.159578947368421, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 57528 + }, + { + "epoch": 3.159639097744361, + "grad_norm": 0.4921875, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 57529 + }, + { + "epoch": 3.1596992481203006, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 57530 + }, + { + "epoch": 3.1597593984962407, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 57531 + }, + { + "epoch": 3.1598195488721803, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 57532 + }, + { + "epoch": 3.1598796992481204, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 57533 + }, + { + "epoch": 3.15993984962406, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 57534 + }, + { + "epoch": 3.16, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 57535 + }, + { + "epoch": 3.16006015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 57536 + }, + { + "epoch": 3.16012030075188, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9254, + "step": 57537 + }, + { + "epoch": 3.1601804511278195, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 57538 + }, + { + "epoch": 3.1602406015037596, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 57539 + }, + { + "epoch": 3.1603007518796993, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 57540 + }, + { + "epoch": 3.160360902255639, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 57541 + }, + { + "epoch": 3.160421052631579, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 57542 + }, + { + "epoch": 3.1604812030075187, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 57543 + }, + { + "epoch": 3.1605413533834588, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9304, + "step": 57544 + }, + { + "epoch": 3.1606015037593984, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9294, + "step": 57545 + }, + { + "epoch": 3.1606616541353385, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 57546 + }, + { + "epoch": 3.160721804511278, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 57547 + }, + { + "epoch": 3.1607819548872182, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9336, + "step": 57548 + }, + { + "epoch": 3.160842105263158, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 57549 + }, + { + "epoch": 3.1609022556390975, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9268, + "step": 57550 + }, + { + "epoch": 3.1609624060150376, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 57551 + }, + { + "epoch": 3.1610225563909773, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 57552 + }, + { + "epoch": 3.1610827067669174, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 57553 + }, + { + "epoch": 3.161142857142857, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 57554 + }, + { + "epoch": 3.161203007518797, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 57555 + }, + { + "epoch": 3.1612631578947368, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 57556 + }, + { + "epoch": 3.161323308270677, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 57557 + }, + { + "epoch": 3.1613834586466165, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9162, + "step": 57558 + }, + { + "epoch": 3.161443609022556, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.8958, + "step": 57559 + }, + { + "epoch": 3.1615037593984963, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9357, + "step": 57560 + }, + { + "epoch": 3.161563909774436, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 57561 + }, + { + "epoch": 3.161624060150376, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 57562 + }, + { + "epoch": 3.1616842105263157, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 57563 + }, + { + "epoch": 3.1617443609022557, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 57564 + }, + { + "epoch": 3.1618045112781954, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 57565 + }, + { + "epoch": 3.1618646616541355, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 57566 + }, + { + "epoch": 3.161924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 57567 + }, + { + "epoch": 3.1619849624060152, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 57568 + }, + { + "epoch": 3.162045112781955, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 57569 + }, + { + "epoch": 3.1621052631578945, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 57570 + }, + { + "epoch": 3.1621654135338346, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 57571 + }, + { + "epoch": 3.1622255639097743, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9537, + "step": 57572 + }, + { + "epoch": 3.1622857142857144, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 57573 + }, + { + "epoch": 3.162345864661654, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0969, + "step": 57574 + }, + { + "epoch": 3.162406015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 57575 + }, + { + "epoch": 3.1624661654135338, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 57576 + }, + { + "epoch": 3.162526315789474, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 57577 + }, + { + "epoch": 3.1625864661654135, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 57578 + }, + { + "epoch": 3.1626466165413536, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 57579 + }, + { + "epoch": 3.1627067669172932, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0725, + "step": 57580 + }, + { + "epoch": 3.162766917293233, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 57581 + }, + { + "epoch": 3.162827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 57582 + }, + { + "epoch": 3.1628872180451126, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 57583 + }, + { + "epoch": 3.1629473684210527, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 57584 + }, + { + "epoch": 3.1630075187969924, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 57585 + }, + { + "epoch": 3.1630676691729325, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 57586 + }, + { + "epoch": 3.163127819548872, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 57587 + }, + { + "epoch": 3.163187969924812, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 57588 + }, + { + "epoch": 3.163248120300752, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9389, + "step": 57589 + }, + { + "epoch": 3.1633082706766915, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 57590 + }, + { + "epoch": 3.1633684210526316, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 57591 + }, + { + "epoch": 3.1634285714285713, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 57592 + }, + { + "epoch": 3.1634887218045113, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 57593 + }, + { + "epoch": 3.163548872180451, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 57594 + }, + { + "epoch": 3.163609022556391, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 57595 + }, + { + "epoch": 3.1636691729323307, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9219, + "step": 57596 + }, + { + "epoch": 3.163729323308271, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 57597 + }, + { + "epoch": 3.1637894736842105, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 57598 + }, + { + "epoch": 3.1638496240601506, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 57599 + }, + { + "epoch": 3.1639097744360902, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1683, + "step": 57600 + }, + { + "epoch": 3.16396992481203, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 57601 + }, + { + "epoch": 3.16403007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 57602 + }, + { + "epoch": 3.1640902255639096, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 57603 + }, + { + "epoch": 3.1641503759398497, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9251, + "step": 57604 + }, + { + "epoch": 3.1642105263157894, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 57605 + }, + { + "epoch": 3.1642706766917295, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 57606 + }, + { + "epoch": 3.164330827067669, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 57607 + }, + { + "epoch": 3.164390977443609, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 57608 + }, + { + "epoch": 3.164451127819549, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 57609 + }, + { + "epoch": 3.164511278195489, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1705, + "step": 57610 + }, + { + "epoch": 3.1645714285714286, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 57611 + }, + { + "epoch": 3.1646315789473682, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 57612 + }, + { + "epoch": 3.1646917293233083, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 57613 + }, + { + "epoch": 3.164751879699248, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 57614 + }, + { + "epoch": 3.164812030075188, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 57615 + }, + { + "epoch": 3.1648721804511277, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 57616 + }, + { + "epoch": 3.164932330827068, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 57617 + }, + { + "epoch": 3.1649924812030075, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 57618 + }, + { + "epoch": 3.1650526315789476, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 57619 + }, + { + "epoch": 3.165112781954887, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 57620 + }, + { + "epoch": 3.165172932330827, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 57621 + }, + { + "epoch": 3.165233082706767, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 57622 + }, + { + "epoch": 3.1652932330827066, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 57623 + }, + { + "epoch": 3.1653533834586467, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 57624 + }, + { + "epoch": 3.1654135338345863, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 57625 + }, + { + "epoch": 3.1654736842105264, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 57626 + }, + { + "epoch": 3.165533834586466, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 57627 + }, + { + "epoch": 3.165593984962406, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 57628 + }, + { + "epoch": 3.165654135338346, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 57629 + }, + { + "epoch": 3.1657142857142855, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 57630 + }, + { + "epoch": 3.1657744360902256, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 57631 + }, + { + "epoch": 3.1658345864661652, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 57632 + }, + { + "epoch": 3.1658947368421053, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 57633 + }, + { + "epoch": 3.165954887218045, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 57634 + }, + { + "epoch": 3.166015037593985, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 57635 + }, + { + "epoch": 3.1660751879699247, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 57636 + }, + { + "epoch": 3.166135338345865, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 57637 + }, + { + "epoch": 3.1661954887218045, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 57638 + }, + { + "epoch": 3.1662556390977445, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 57639 + }, + { + "epoch": 3.166315789473684, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9348, + "step": 57640 + }, + { + "epoch": 3.1663759398496243, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 57641 + }, + { + "epoch": 3.166436090225564, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 57642 + }, + { + "epoch": 3.1664962406015036, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 57643 + }, + { + "epoch": 3.1665563909774437, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 57644 + }, + { + "epoch": 3.1666165413533833, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 57645 + }, + { + "epoch": 3.1666766917293234, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 57646 + }, + { + "epoch": 3.166736842105263, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 57647 + }, + { + "epoch": 3.166796992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 57648 + }, + { + "epoch": 3.166857142857143, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 57649 + }, + { + "epoch": 3.166917293233083, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 57650 + }, + { + "epoch": 3.1669774436090226, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 57651 + }, + { + "epoch": 3.167037593984962, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9191, + "step": 57652 + }, + { + "epoch": 3.1670977443609023, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.908, + "step": 57653 + }, + { + "epoch": 3.167157894736842, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9247, + "step": 57654 + }, + { + "epoch": 3.167218045112782, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9369, + "step": 57655 + }, + { + "epoch": 3.1672781954887217, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 57656 + }, + { + "epoch": 3.167338345864662, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 57657 + }, + { + "epoch": 3.1673984962406014, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 57658 + }, + { + "epoch": 3.1674586466165415, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 57659 + }, + { + "epoch": 3.167518796992481, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 57660 + }, + { + "epoch": 3.167578947368421, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 57661 + }, + { + "epoch": 3.167639097744361, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 57662 + }, + { + "epoch": 3.1676992481203006, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 57663 + }, + { + "epoch": 3.1677593984962407, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9255, + "step": 57664 + }, + { + "epoch": 3.1678195488721803, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 57665 + }, + { + "epoch": 3.1678796992481204, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 57666 + }, + { + "epoch": 3.16793984962406, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9256, + "step": 57667 + }, + { + "epoch": 3.168, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 57668 + }, + { + "epoch": 3.16806015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 57669 + }, + { + "epoch": 3.16812030075188, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 57670 + }, + { + "epoch": 3.1681804511278195, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 57671 + }, + { + "epoch": 3.168240601503759, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0821, + "step": 57672 + }, + { + "epoch": 3.1683007518796993, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 57673 + }, + { + "epoch": 3.168360902255639, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 57674 + }, + { + "epoch": 3.168421052631579, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.941, + "step": 57675 + }, + { + "epoch": 3.1684812030075187, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 57676 + }, + { + "epoch": 3.1685413533834588, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 57677 + }, + { + "epoch": 3.1686015037593984, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 57678 + }, + { + "epoch": 3.1686616541353385, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 57679 + }, + { + "epoch": 3.168721804511278, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 57680 + }, + { + "epoch": 3.1687819548872183, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 57681 + }, + { + "epoch": 3.168842105263158, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0004, + "step": 57682 + }, + { + "epoch": 3.1689022556390976, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 57683 + }, + { + "epoch": 3.1689624060150376, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1375, + "step": 57684 + }, + { + "epoch": 3.1690225563909773, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 57685 + }, + { + "epoch": 3.1690827067669174, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 57686 + }, + { + "epoch": 3.169142857142857, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 57687 + }, + { + "epoch": 3.169203007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 57688 + }, + { + "epoch": 3.169263157894737, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 57689 + }, + { + "epoch": 3.169323308270677, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 57690 + }, + { + "epoch": 3.1693834586466165, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1027, + "step": 57691 + }, + { + "epoch": 3.169443609022556, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 57692 + }, + { + "epoch": 3.1695037593984963, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 57693 + }, + { + "epoch": 3.169563909774436, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 57694 + }, + { + "epoch": 3.169624060150376, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 57695 + }, + { + "epoch": 3.1696842105263157, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0724, + "step": 57696 + }, + { + "epoch": 3.1697443609022558, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 57697 + }, + { + "epoch": 3.1698045112781954, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 57698 + }, + { + "epoch": 3.1698646616541355, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 57699 + }, + { + "epoch": 3.169924812030075, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 57700 + }, + { + "epoch": 3.1699849624060152, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 57701 + }, + { + "epoch": 3.170045112781955, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 57702 + }, + { + "epoch": 3.1701052631578945, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9146, + "step": 57703 + }, + { + "epoch": 3.1701654135338346, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 57704 + }, + { + "epoch": 3.1702255639097743, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1089, + "step": 57705 + }, + { + "epoch": 3.1702857142857144, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 57706 + }, + { + "epoch": 3.170345864661654, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 57707 + }, + { + "epoch": 3.170406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 57708 + }, + { + "epoch": 3.1704661654135338, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 57709 + }, + { + "epoch": 3.170526315789474, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 57710 + }, + { + "epoch": 3.1705864661654135, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 57711 + }, + { + "epoch": 3.1706466165413536, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 57712 + }, + { + "epoch": 3.1707067669172933, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 57713 + }, + { + "epoch": 3.170766917293233, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 57714 + }, + { + "epoch": 3.170827067669173, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 57715 + }, + { + "epoch": 3.1708872180451126, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9273, + "step": 57716 + }, + { + "epoch": 3.1709473684210527, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 57717 + }, + { + "epoch": 3.1710075187969924, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 57718 + }, + { + "epoch": 3.1710676691729325, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 57719 + }, + { + "epoch": 3.171127819548872, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 57720 + }, + { + "epoch": 3.1711879699248122, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 57721 + }, + { + "epoch": 3.171248120300752, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.03, + "step": 57722 + }, + { + "epoch": 3.1713082706766915, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 57723 + }, + { + "epoch": 3.1713684210526316, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 57724 + }, + { + "epoch": 3.1714285714285713, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 57725 + }, + { + "epoch": 3.1714887218045114, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 57726 + }, + { + "epoch": 3.171548872180451, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 57727 + }, + { + "epoch": 3.171609022556391, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 57728 + }, + { + "epoch": 3.1716691729323307, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 57729 + }, + { + "epoch": 3.171729323308271, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9177, + "step": 57730 + }, + { + "epoch": 3.1717894736842105, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 57731 + }, + { + "epoch": 3.17184962406015, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 57732 + }, + { + "epoch": 3.1719097744360902, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 57733 + }, + { + "epoch": 3.17196992481203, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 57734 + }, + { + "epoch": 3.17203007518797, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.129, + "step": 57735 + }, + { + "epoch": 3.1720902255639096, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 57736 + }, + { + "epoch": 3.1721503759398497, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 57737 + }, + { + "epoch": 3.1722105263157894, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 57738 + }, + { + "epoch": 3.1722706766917295, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 57739 + }, + { + "epoch": 3.172330827067669, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9169, + "step": 57740 + }, + { + "epoch": 3.172390977443609, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 57741 + }, + { + "epoch": 3.172451127819549, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 57742 + }, + { + "epoch": 3.172511278195489, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 57743 + }, + { + "epoch": 3.1725714285714286, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 57744 + }, + { + "epoch": 3.1726315789473682, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 57745 + }, + { + "epoch": 3.1726917293233083, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 57746 + }, + { + "epoch": 3.172751879699248, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 57747 + }, + { + "epoch": 3.172812030075188, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 57748 + }, + { + "epoch": 3.1728721804511277, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1709, + "step": 57749 + }, + { + "epoch": 3.172932330827068, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 57750 + }, + { + "epoch": 3.1729924812030075, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0844, + "step": 57751 + }, + { + "epoch": 3.1730526315789476, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9495, + "step": 57752 + }, + { + "epoch": 3.173112781954887, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 57753 + }, + { + "epoch": 3.173172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 57754 + }, + { + "epoch": 3.173233082706767, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 57755 + }, + { + "epoch": 3.1732932330827066, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 57756 + }, + { + "epoch": 3.1733533834586467, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 57757 + }, + { + "epoch": 3.1734135338345864, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9177, + "step": 57758 + }, + { + "epoch": 3.1734736842105264, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 57759 + }, + { + "epoch": 3.173533834586466, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 57760 + }, + { + "epoch": 3.173593984962406, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 57761 + }, + { + "epoch": 3.173654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 57762 + }, + { + "epoch": 3.1737142857142855, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.933, + "step": 57763 + }, + { + "epoch": 3.1737744360902256, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0356, + "step": 57764 + }, + { + "epoch": 3.1738345864661652, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9287, + "step": 57765 + }, + { + "epoch": 3.1738947368421053, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 57766 + }, + { + "epoch": 3.173954887218045, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9287, + "step": 57767 + }, + { + "epoch": 3.174015037593985, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 57768 + }, + { + "epoch": 3.1740751879699247, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 57769 + }, + { + "epoch": 3.174135338345865, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 57770 + }, + { + "epoch": 3.1741954887218045, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 57771 + }, + { + "epoch": 3.1742556390977446, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 57772 + }, + { + "epoch": 3.174315789473684, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 57773 + }, + { + "epoch": 3.1743759398496243, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 57774 + }, + { + "epoch": 3.174436090225564, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 57775 + }, + { + "epoch": 3.1744962406015036, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 57776 + }, + { + "epoch": 3.1745563909774437, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 57777 + }, + { + "epoch": 3.1746165413533833, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.926, + "step": 57778 + }, + { + "epoch": 3.1746766917293234, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 57779 + }, + { + "epoch": 3.174736842105263, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 57780 + }, + { + "epoch": 3.174796992481203, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 57781 + }, + { + "epoch": 3.174857142857143, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 57782 + }, + { + "epoch": 3.174917293233083, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 57783 + }, + { + "epoch": 3.1749774436090226, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 57784 + }, + { + "epoch": 3.175037593984962, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 57785 + }, + { + "epoch": 3.1750977443609023, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 57786 + }, + { + "epoch": 3.175157894736842, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 57787 + }, + { + "epoch": 3.175218045112782, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 57788 + }, + { + "epoch": 3.1752781954887217, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9416, + "step": 57789 + }, + { + "epoch": 3.175338345864662, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 57790 + }, + { + "epoch": 3.1753984962406014, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.092, + "step": 57791 + }, + { + "epoch": 3.1754586466165415, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 57792 + }, + { + "epoch": 3.175518796992481, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 57793 + }, + { + "epoch": 3.175578947368421, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 57794 + }, + { + "epoch": 3.175639097744361, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 57795 + }, + { + "epoch": 3.1756992481203006, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 57796 + }, + { + "epoch": 3.1757593984962407, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 57797 + }, + { + "epoch": 3.1758195488721803, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 57798 + }, + { + "epoch": 3.1758796992481204, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 57799 + }, + { + "epoch": 3.17593984962406, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 57800 + }, + { + "epoch": 3.176, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 57801 + }, + { + "epoch": 3.17606015037594, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 57802 + }, + { + "epoch": 3.17612030075188, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 57803 + }, + { + "epoch": 3.1761804511278195, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9504, + "step": 57804 + }, + { + "epoch": 3.176240601503759, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 57805 + }, + { + "epoch": 3.1763007518796993, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 57806 + }, + { + "epoch": 3.176360902255639, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 57807 + }, + { + "epoch": 3.176421052631579, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 57808 + }, + { + "epoch": 3.1764812030075187, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 57809 + }, + { + "epoch": 3.1765413533834588, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 57810 + }, + { + "epoch": 3.1766015037593984, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 57811 + }, + { + "epoch": 3.1766616541353385, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 57812 + }, + { + "epoch": 3.176721804511278, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 57813 + }, + { + "epoch": 3.1767819548872183, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 57814 + }, + { + "epoch": 3.176842105263158, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 57815 + }, + { + "epoch": 3.1769022556390976, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 57816 + }, + { + "epoch": 3.1769624060150377, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 57817 + }, + { + "epoch": 3.1770225563909773, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 57818 + }, + { + "epoch": 3.1770827067669174, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 57819 + }, + { + "epoch": 3.177142857142857, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 57820 + }, + { + "epoch": 3.177203007518797, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 57821 + }, + { + "epoch": 3.177263157894737, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9219, + "step": 57822 + }, + { + "epoch": 3.177323308270677, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 57823 + }, + { + "epoch": 3.1773834586466165, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 57824 + }, + { + "epoch": 3.177443609022556, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 57825 + }, + { + "epoch": 3.1775037593984963, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 57826 + }, + { + "epoch": 3.177563909774436, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 57827 + }, + { + "epoch": 3.177624060150376, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 57828 + }, + { + "epoch": 3.1776842105263157, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 57829 + }, + { + "epoch": 3.1777443609022558, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 57830 + }, + { + "epoch": 3.1778045112781954, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 57831 + }, + { + "epoch": 3.1778646616541355, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9336, + "step": 57832 + }, + { + "epoch": 3.177924812030075, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 57833 + }, + { + "epoch": 3.177984962406015, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 57834 + }, + { + "epoch": 3.178045112781955, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 57835 + }, + { + "epoch": 3.1781052631578945, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 57836 + }, + { + "epoch": 3.1781654135338346, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9364, + "step": 57837 + }, + { + "epoch": 3.1782255639097743, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 57838 + }, + { + "epoch": 3.1782857142857144, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 57839 + }, + { + "epoch": 3.178345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1091, + "step": 57840 + }, + { + "epoch": 3.178406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 57841 + }, + { + "epoch": 3.1784661654135338, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 57842 + }, + { + "epoch": 3.178526315789474, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 57843 + }, + { + "epoch": 3.1785864661654135, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 57844 + }, + { + "epoch": 3.1786466165413536, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9215, + "step": 57845 + }, + { + "epoch": 3.1787067669172933, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 57846 + }, + { + "epoch": 3.178766917293233, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 57847 + }, + { + "epoch": 3.178827067669173, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 57848 + }, + { + "epoch": 3.1788872180451127, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 57849 + }, + { + "epoch": 3.1789473684210527, + "grad_norm": 1.75, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 57850 + }, + { + "epoch": 3.1790075187969924, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 57851 + }, + { + "epoch": 3.1790676691729325, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 57852 + }, + { + "epoch": 3.179127819548872, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 57853 + }, + { + "epoch": 3.1791879699248122, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 57854 + }, + { + "epoch": 3.179248120300752, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 57855 + }, + { + "epoch": 3.1793082706766915, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 57856 + }, + { + "epoch": 3.1793684210526316, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 57857 + }, + { + "epoch": 3.1794285714285713, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 57858 + }, + { + "epoch": 3.1794887218045114, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 57859 + }, + { + "epoch": 3.179548872180451, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 57860 + }, + { + "epoch": 3.179609022556391, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1326, + "step": 57861 + }, + { + "epoch": 3.1796691729323308, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 57862 + }, + { + "epoch": 3.179729323308271, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9414, + "step": 57863 + }, + { + "epoch": 3.1797894736842105, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 57864 + }, + { + "epoch": 3.17984962406015, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.078, + "step": 57865 + }, + { + "epoch": 3.1799097744360902, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 57866 + }, + { + "epoch": 3.17996992481203, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 57867 + }, + { + "epoch": 3.18003007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 57868 + }, + { + "epoch": 3.1800902255639096, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 57869 + }, + { + "epoch": 3.1801503759398497, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 57870 + }, + { + "epoch": 3.1802105263157894, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 57871 + }, + { + "epoch": 3.1802706766917295, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1611, + "step": 57872 + }, + { + "epoch": 3.180330827067669, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 57873 + }, + { + "epoch": 3.180390977443609, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 57874 + }, + { + "epoch": 3.180451127819549, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 57875 + }, + { + "epoch": 3.180511278195489, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 57876 + }, + { + "epoch": 3.1805714285714286, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 57877 + }, + { + "epoch": 3.1806315789473683, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 57878 + }, + { + "epoch": 3.1806917293233083, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 57879 + }, + { + "epoch": 3.180751879699248, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.8675, + "step": 57880 + }, + { + "epoch": 3.180812030075188, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 57881 + }, + { + "epoch": 3.1808721804511277, + "grad_norm": 1.3671875, + "learning_rate": 3e-05, + "loss": 1.1323, + "step": 57882 + }, + { + "epoch": 3.180932330827068, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 57883 + }, + { + "epoch": 3.1809924812030075, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 57884 + }, + { + "epoch": 3.1810526315789476, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 57885 + }, + { + "epoch": 3.1811127819548872, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 57886 + }, + { + "epoch": 3.181172932330827, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 57887 + }, + { + "epoch": 3.181233082706767, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 57888 + }, + { + "epoch": 3.1812932330827066, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 57889 + }, + { + "epoch": 3.1813533834586467, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 57890 + }, + { + "epoch": 3.1814135338345864, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.8976, + "step": 57891 + }, + { + "epoch": 3.1814736842105265, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.8983, + "step": 57892 + }, + { + "epoch": 3.181533834586466, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 57893 + }, + { + "epoch": 3.181593984962406, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1499, + "step": 57894 + }, + { + "epoch": 3.181654135338346, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.1549, + "step": 57895 + }, + { + "epoch": 3.1817142857142855, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9402, + "step": 57896 + }, + { + "epoch": 3.1817744360902256, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 57897 + }, + { + "epoch": 3.1818345864661652, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 57898 + }, + { + "epoch": 3.1818947368421053, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 57899 + }, + { + "epoch": 3.181954887218045, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 57900 + }, + { + "epoch": 3.182015037593985, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.2254, + "step": 57901 + }, + { + "epoch": 3.1820751879699247, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 57902 + }, + { + "epoch": 3.182135338345865, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 57903 + }, + { + "epoch": 3.1821954887218045, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 57904 + }, + { + "epoch": 3.1822556390977446, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9203, + "step": 57905 + }, + { + "epoch": 3.182315789473684, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9388, + "step": 57906 + }, + { + "epoch": 3.182375939849624, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.8886, + "step": 57907 + }, + { + "epoch": 3.182436090225564, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 57908 + }, + { + "epoch": 3.1824962406015036, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 57909 + }, + { + "epoch": 3.1825563909774437, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 57910 + }, + { + "epoch": 3.1826165413533833, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 57911 + }, + { + "epoch": 3.1826766917293234, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 57912 + }, + { + "epoch": 3.182736842105263, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9169, + "step": 57913 + }, + { + "epoch": 3.182796992481203, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 57914 + }, + { + "epoch": 3.182857142857143, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 57915 + }, + { + "epoch": 3.182917293233083, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 57916 + }, + { + "epoch": 3.1829774436090226, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 57917 + }, + { + "epoch": 3.183037593984962, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 57918 + }, + { + "epoch": 3.1830977443609023, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9377, + "step": 57919 + }, + { + "epoch": 3.183157894736842, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 57920 + }, + { + "epoch": 3.183218045112782, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 57921 + }, + { + "epoch": 3.1832781954887217, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9269, + "step": 57922 + }, + { + "epoch": 3.183338345864662, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 57923 + }, + { + "epoch": 3.1833984962406014, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 57924 + }, + { + "epoch": 3.1834586466165415, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 57925 + }, + { + "epoch": 3.183518796992481, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 57926 + }, + { + "epoch": 3.183578947368421, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 57927 + }, + { + "epoch": 3.183639097744361, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 57928 + }, + { + "epoch": 3.1836992481203006, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 57929 + }, + { + "epoch": 3.1837593984962407, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9292, + "step": 57930 + }, + { + "epoch": 3.1838195488721803, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 57931 + }, + { + "epoch": 3.1838796992481204, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 57932 + }, + { + "epoch": 3.18393984962406, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 57933 + }, + { + "epoch": 3.184, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 57934 + }, + { + "epoch": 3.18406015037594, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 57935 + }, + { + "epoch": 3.18412030075188, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 57936 + }, + { + "epoch": 3.1841804511278196, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 57937 + }, + { + "epoch": 3.184240601503759, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 57938 + }, + { + "epoch": 3.1843007518796993, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.915, + "step": 57939 + }, + { + "epoch": 3.184360902255639, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 57940 + }, + { + "epoch": 3.184421052631579, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 57941 + }, + { + "epoch": 3.1844812030075187, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 57942 + }, + { + "epoch": 3.184541353383459, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9202, + "step": 57943 + }, + { + "epoch": 3.1846015037593984, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 57944 + }, + { + "epoch": 3.1846616541353385, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1148, + "step": 57945 + }, + { + "epoch": 3.184721804511278, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 57946 + }, + { + "epoch": 3.1847819548872183, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 57947 + }, + { + "epoch": 3.184842105263158, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 57948 + }, + { + "epoch": 3.1849022556390976, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 57949 + }, + { + "epoch": 3.1849624060150377, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 57950 + }, + { + "epoch": 3.1850225563909773, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 57951 + }, + { + "epoch": 3.1850827067669174, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9331, + "step": 57952 + }, + { + "epoch": 3.185142857142857, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 57953 + }, + { + "epoch": 3.185203007518797, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 57954 + }, + { + "epoch": 3.185263157894737, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 57955 + }, + { + "epoch": 3.185323308270677, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 57956 + }, + { + "epoch": 3.1853834586466165, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 57957 + }, + { + "epoch": 3.185443609022556, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 57958 + }, + { + "epoch": 3.1855037593984963, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 57959 + }, + { + "epoch": 3.185563909774436, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 57960 + }, + { + "epoch": 3.185624060150376, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 57961 + }, + { + "epoch": 3.1856842105263157, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 57962 + }, + { + "epoch": 3.1857443609022558, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 57963 + }, + { + "epoch": 3.1858045112781954, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 57964 + }, + { + "epoch": 3.1858646616541355, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0616, + "step": 57965 + }, + { + "epoch": 3.185924812030075, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9303, + "step": 57966 + }, + { + "epoch": 3.185984962406015, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 57967 + }, + { + "epoch": 3.186045112781955, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 57968 + }, + { + "epoch": 3.1861052631578946, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 57969 + }, + { + "epoch": 3.1861654135338346, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.8424, + "step": 57970 + }, + { + "epoch": 3.1862255639097743, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 57971 + }, + { + "epoch": 3.1862857142857144, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0516, + "step": 57972 + }, + { + "epoch": 3.186345864661654, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0772, + "step": 57973 + }, + { + "epoch": 3.186406015037594, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 57974 + }, + { + "epoch": 3.186466165413534, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9549, + "step": 57975 + }, + { + "epoch": 3.186526315789474, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 57976 + }, + { + "epoch": 3.1865864661654135, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 57977 + }, + { + "epoch": 3.1866466165413536, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 57978 + }, + { + "epoch": 3.1867067669172933, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 57979 + }, + { + "epoch": 3.186766917293233, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 57980 + }, + { + "epoch": 3.186827067669173, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 57981 + }, + { + "epoch": 3.1868872180451127, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 57982 + }, + { + "epoch": 3.1869473684210528, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 57983 + }, + { + "epoch": 3.1870075187969924, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 57984 + }, + { + "epoch": 3.1870676691729325, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 57985 + }, + { + "epoch": 3.187127819548872, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 57986 + }, + { + "epoch": 3.1871879699248122, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 57987 + }, + { + "epoch": 3.187248120300752, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 57988 + }, + { + "epoch": 3.1873082706766915, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 57989 + }, + { + "epoch": 3.1873684210526316, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 57990 + }, + { + "epoch": 3.1874285714285713, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 57991 + }, + { + "epoch": 3.1874887218045114, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 57992 + }, + { + "epoch": 3.187548872180451, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 0.9352, + "step": 57993 + }, + { + "epoch": 3.187609022556391, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 57994 + }, + { + "epoch": 3.1876691729323308, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 57995 + }, + { + "epoch": 3.187729323308271, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0436, + "step": 57996 + }, + { + "epoch": 3.1877894736842105, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 57997 + }, + { + "epoch": 3.18784962406015, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 57998 + }, + { + "epoch": 3.1879097744360902, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 57999 + }, + { + "epoch": 3.18796992481203, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 58000 + }, + { + "epoch": 3.18803007518797, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0937, + "step": 58001 + }, + { + "epoch": 3.1880902255639096, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 58002 + }, + { + "epoch": 3.1881503759398497, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 58003 + }, + { + "epoch": 3.1882105263157894, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 58004 + }, + { + "epoch": 3.1882706766917295, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 58005 + }, + { + "epoch": 3.188330827067669, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 58006 + }, + { + "epoch": 3.188390977443609, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 58007 + }, + { + "epoch": 3.188451127819549, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 58008 + }, + { + "epoch": 3.1885112781954885, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 58009 + }, + { + "epoch": 3.1885714285714286, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1527, + "step": 58010 + }, + { + "epoch": 3.1886315789473683, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 58011 + }, + { + "epoch": 3.1886917293233084, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 58012 + }, + { + "epoch": 3.188751879699248, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 58013 + }, + { + "epoch": 3.188812030075188, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 58014 + }, + { + "epoch": 3.1888721804511277, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 58015 + }, + { + "epoch": 3.188932330827068, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 58016 + }, + { + "epoch": 3.1889924812030075, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 58017 + }, + { + "epoch": 3.1890526315789476, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 58018 + }, + { + "epoch": 3.1891127819548872, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 58019 + }, + { + "epoch": 3.189172932330827, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 58020 + }, + { + "epoch": 3.189233082706767, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 58021 + }, + { + "epoch": 3.1892932330827066, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 58022 + }, + { + "epoch": 3.1893533834586467, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 58023 + }, + { + "epoch": 3.1894135338345864, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 58024 + }, + { + "epoch": 3.1894736842105265, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 58025 + }, + { + "epoch": 3.189533834586466, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 58026 + }, + { + "epoch": 3.189593984962406, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0884, + "step": 58027 + }, + { + "epoch": 3.189654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 58028 + }, + { + "epoch": 3.1897142857142855, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 58029 + }, + { + "epoch": 3.1897744360902256, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.1164, + "step": 58030 + }, + { + "epoch": 3.1898345864661652, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 58031 + }, + { + "epoch": 3.1898947368421053, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9444, + "step": 58032 + }, + { + "epoch": 3.189954887218045, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 58033 + }, + { + "epoch": 3.190015037593985, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 58034 + }, + { + "epoch": 3.1900751879699247, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.897, + "step": 58035 + }, + { + "epoch": 3.190135338345865, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 58036 + }, + { + "epoch": 3.1901954887218045, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9231, + "step": 58037 + }, + { + "epoch": 3.1902556390977446, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 58038 + }, + { + "epoch": 3.190315789473684, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 58039 + }, + { + "epoch": 3.190375939849624, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 58040 + }, + { + "epoch": 3.190436090225564, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0207, + "step": 58041 + }, + { + "epoch": 3.1904962406015036, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 58042 + }, + { + "epoch": 3.1905563909774437, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 58043 + }, + { + "epoch": 3.1906165413533834, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 58044 + }, + { + "epoch": 3.1906766917293234, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 58045 + }, + { + "epoch": 3.190736842105263, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.2866, + "step": 58046 + }, + { + "epoch": 3.190796992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 58047 + }, + { + "epoch": 3.190857142857143, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 58048 + }, + { + "epoch": 3.190917293233083, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 58049 + }, + { + "epoch": 3.1909774436090226, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 58050 + }, + { + "epoch": 3.1910375939849622, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 58051 + }, + { + "epoch": 3.1910977443609023, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 58052 + }, + { + "epoch": 3.191157894736842, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 58053 + }, + { + "epoch": 3.191218045112782, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 58054 + }, + { + "epoch": 3.1912781954887217, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 58055 + }, + { + "epoch": 3.191338345864662, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 58056 + }, + { + "epoch": 3.1913984962406015, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.103, + "step": 58057 + }, + { + "epoch": 3.1914586466165416, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 58058 + }, + { + "epoch": 3.191518796992481, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.2887, + "step": 58059 + }, + { + "epoch": 3.191578947368421, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 58060 + }, + { + "epoch": 3.191639097744361, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 58061 + }, + { + "epoch": 3.1916992481203006, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 58062 + }, + { + "epoch": 3.1917593984962407, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 58063 + }, + { + "epoch": 3.1918195488721803, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 58064 + }, + { + "epoch": 3.1918796992481204, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 58065 + }, + { + "epoch": 3.19193984962406, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 58066 + }, + { + "epoch": 3.192, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 58067 + }, + { + "epoch": 3.19206015037594, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 58068 + }, + { + "epoch": 3.1921203007518795, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 58069 + }, + { + "epoch": 3.1921804511278196, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 58070 + }, + { + "epoch": 3.192240601503759, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 58071 + }, + { + "epoch": 3.1923007518796993, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 58072 + }, + { + "epoch": 3.192360902255639, + "grad_norm": 1.875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 58073 + }, + { + "epoch": 3.192421052631579, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9495, + "step": 58074 + }, + { + "epoch": 3.1924812030075187, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 58075 + }, + { + "epoch": 3.192541353383459, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 58076 + }, + { + "epoch": 3.1926015037593984, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.939, + "step": 58077 + }, + { + "epoch": 3.1926616541353385, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 58078 + }, + { + "epoch": 3.192721804511278, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.8956, + "step": 58079 + }, + { + "epoch": 3.1927819548872183, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0807, + "step": 58080 + }, + { + "epoch": 3.192842105263158, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 58081 + }, + { + "epoch": 3.1929022556390976, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 58082 + }, + { + "epoch": 3.1929624060150377, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0318, + "step": 58083 + }, + { + "epoch": 3.1930225563909773, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 58084 + }, + { + "epoch": 3.1930827067669174, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 58085 + }, + { + "epoch": 3.193142857142857, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1684, + "step": 58086 + }, + { + "epoch": 3.193203007518797, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9329, + "step": 58087 + }, + { + "epoch": 3.193263157894737, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 58088 + }, + { + "epoch": 3.193323308270677, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 58089 + }, + { + "epoch": 3.1933834586466165, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9531, + "step": 58090 + }, + { + "epoch": 3.193443609022556, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9266, + "step": 58091 + }, + { + "epoch": 3.1935037593984963, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 58092 + }, + { + "epoch": 3.193563909774436, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 58093 + }, + { + "epoch": 3.193624060150376, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1473, + "step": 58094 + }, + { + "epoch": 3.1936842105263157, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 58095 + }, + { + "epoch": 3.1937443609022558, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 58096 + }, + { + "epoch": 3.1938045112781954, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 58097 + }, + { + "epoch": 3.1938646616541355, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 58098 + }, + { + "epoch": 3.193924812030075, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9209, + "step": 58099 + }, + { + "epoch": 3.193984962406015, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 58100 + }, + { + "epoch": 3.194045112781955, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 58101 + }, + { + "epoch": 3.1941052631578946, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 58102 + }, + { + "epoch": 3.1941654135338347, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 58103 + }, + { + "epoch": 3.1942255639097743, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 58104 + }, + { + "epoch": 3.1942857142857144, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 58105 + }, + { + "epoch": 3.194345864661654, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 58106 + }, + { + "epoch": 3.194406015037594, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 58107 + }, + { + "epoch": 3.194466165413534, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9495, + "step": 58108 + }, + { + "epoch": 3.194526315789474, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 58109 + }, + { + "epoch": 3.1945864661654135, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 58110 + }, + { + "epoch": 3.194646616541353, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 58111 + }, + { + "epoch": 3.1947067669172933, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 58112 + }, + { + "epoch": 3.194766917293233, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 58113 + }, + { + "epoch": 3.194827067669173, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 58114 + }, + { + "epoch": 3.1948872180451127, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 58115 + }, + { + "epoch": 3.1949473684210528, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 58116 + }, + { + "epoch": 3.1950075187969924, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 58117 + }, + { + "epoch": 3.1950676691729325, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 58118 + }, + { + "epoch": 3.195127819548872, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9016, + "step": 58119 + }, + { + "epoch": 3.1951879699248122, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 58120 + }, + { + "epoch": 3.195248120300752, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 58121 + }, + { + "epoch": 3.1953082706766915, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 58122 + }, + { + "epoch": 3.1953684210526316, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 58123 + }, + { + "epoch": 3.1954285714285713, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 58124 + }, + { + "epoch": 3.1954887218045114, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 58125 + }, + { + "epoch": 3.195548872180451, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 58126 + }, + { + "epoch": 3.195609022556391, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 58127 + }, + { + "epoch": 3.1956691729323308, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.8976, + "step": 58128 + }, + { + "epoch": 3.195729323308271, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9481, + "step": 58129 + }, + { + "epoch": 3.1957894736842105, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 58130 + }, + { + "epoch": 3.19584962406015, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 58131 + }, + { + "epoch": 3.1959097744360903, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 58132 + }, + { + "epoch": 3.19596992481203, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 58133 + }, + { + "epoch": 3.19603007518797, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 58134 + }, + { + "epoch": 3.1960902255639096, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 58135 + }, + { + "epoch": 3.1961503759398497, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 58136 + }, + { + "epoch": 3.1962105263157894, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 58137 + }, + { + "epoch": 3.1962706766917295, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 58138 + }, + { + "epoch": 3.196330827067669, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 58139 + }, + { + "epoch": 3.1963909774436092, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 58140 + }, + { + "epoch": 3.196451127819549, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 58141 + }, + { + "epoch": 3.1965112781954885, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 58142 + }, + { + "epoch": 3.1965714285714286, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 58143 + }, + { + "epoch": 3.1966315789473683, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 58144 + }, + { + "epoch": 3.1966917293233084, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 58145 + }, + { + "epoch": 3.196751879699248, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 58146 + }, + { + "epoch": 3.196812030075188, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 58147 + }, + { + "epoch": 3.1968721804511278, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 58148 + }, + { + "epoch": 3.196932330827068, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 58149 + }, + { + "epoch": 3.1969924812030075, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 58150 + }, + { + "epoch": 3.1970526315789476, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 58151 + }, + { + "epoch": 3.1971127819548872, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1555, + "step": 58152 + }, + { + "epoch": 3.197172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 58153 + }, + { + "epoch": 3.197233082706767, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 58154 + }, + { + "epoch": 3.1972932330827066, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 58155 + }, + { + "epoch": 3.1973533834586467, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 58156 + }, + { + "epoch": 3.1974135338345864, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 58157 + }, + { + "epoch": 3.1974736842105265, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 58158 + }, + { + "epoch": 3.197533834586466, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 58159 + }, + { + "epoch": 3.197593984962406, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 58160 + }, + { + "epoch": 3.197654135338346, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9259, + "step": 58161 + }, + { + "epoch": 3.1977142857142855, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 58162 + }, + { + "epoch": 3.1977744360902256, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 58163 + }, + { + "epoch": 3.1978345864661653, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9212, + "step": 58164 + }, + { + "epoch": 3.1978947368421053, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 58165 + }, + { + "epoch": 3.197954887218045, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 58166 + }, + { + "epoch": 3.198015037593985, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 58167 + }, + { + "epoch": 3.1980751879699247, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 58168 + }, + { + "epoch": 3.198135338345865, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 58169 + }, + { + "epoch": 3.1981954887218045, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 58170 + }, + { + "epoch": 3.198255639097744, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9292, + "step": 58171 + }, + { + "epoch": 3.1983157894736842, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 58172 + }, + { + "epoch": 3.198375939849624, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 58173 + }, + { + "epoch": 3.198436090225564, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 58174 + }, + { + "epoch": 3.1984962406015036, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 58175 + }, + { + "epoch": 3.1985563909774437, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 58176 + }, + { + "epoch": 3.1986165413533834, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 58177 + }, + { + "epoch": 3.1986766917293235, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 58178 + }, + { + "epoch": 3.198736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 58179 + }, + { + "epoch": 3.198796992481203, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 58180 + }, + { + "epoch": 3.198857142857143, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 58181 + }, + { + "epoch": 3.198917293233083, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 58182 + }, + { + "epoch": 3.1989774436090226, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 58183 + }, + { + "epoch": 3.1990375939849622, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 58184 + }, + { + "epoch": 3.1990977443609023, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 58185 + }, + { + "epoch": 3.199157894736842, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 58186 + }, + { + "epoch": 3.199218045112782, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0421, + "step": 58187 + }, + { + "epoch": 3.1992781954887217, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 58188 + }, + { + "epoch": 3.199338345864662, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 58189 + }, + { + "epoch": 3.1993984962406015, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 58190 + }, + { + "epoch": 3.1994586466165416, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 58191 + }, + { + "epoch": 3.199518796992481, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 58192 + }, + { + "epoch": 3.199578947368421, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 58193 + }, + { + "epoch": 3.199639097744361, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 58194 + }, + { + "epoch": 3.1996992481203006, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 58195 + }, + { + "epoch": 3.1997593984962407, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 58196 + }, + { + "epoch": 3.1998195488721803, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 58197 + }, + { + "epoch": 3.1998796992481204, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 58198 + }, + { + "epoch": 3.19993984962406, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 58199 + }, + { + "epoch": 3.2, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 58200 + }, + { + "epoch": 3.20006015037594, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1645, + "step": 58201 + }, + { + "epoch": 3.2001203007518795, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 58202 + }, + { + "epoch": 3.2001804511278196, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1367, + "step": 58203 + }, + { + "epoch": 3.200240601503759, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 58204 + }, + { + "epoch": 3.2003007518796993, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.8883, + "step": 58205 + }, + { + "epoch": 3.200360902255639, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 58206 + }, + { + "epoch": 3.200421052631579, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 58207 + }, + { + "epoch": 3.2004812030075187, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 58208 + }, + { + "epoch": 3.200541353383459, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 58209 + }, + { + "epoch": 3.2006015037593984, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 58210 + }, + { + "epoch": 3.2006616541353385, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 58211 + }, + { + "epoch": 3.200721804511278, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 58212 + }, + { + "epoch": 3.200781954887218, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 58213 + }, + { + "epoch": 3.200842105263158, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 58214 + }, + { + "epoch": 3.2009022556390976, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9384, + "step": 58215 + }, + { + "epoch": 3.2009624060150377, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9518, + "step": 58216 + }, + { + "epoch": 3.2010225563909773, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 58217 + }, + { + "epoch": 3.2010827067669174, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 58218 + }, + { + "epoch": 3.201142857142857, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 58219 + }, + { + "epoch": 3.201203007518797, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 58220 + }, + { + "epoch": 3.201263157894737, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 58221 + }, + { + "epoch": 3.201323308270677, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9169, + "step": 58222 + }, + { + "epoch": 3.2013834586466166, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 58223 + }, + { + "epoch": 3.201443609022556, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 58224 + }, + { + "epoch": 3.2015037593984963, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 58225 + }, + { + "epoch": 3.201563909774436, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 58226 + }, + { + "epoch": 3.201624060150376, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 58227 + }, + { + "epoch": 3.2016842105263157, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 58228 + }, + { + "epoch": 3.201744360902256, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 58229 + }, + { + "epoch": 3.2018045112781954, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1086, + "step": 58230 + }, + { + "epoch": 3.2018646616541355, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 58231 + }, + { + "epoch": 3.201924812030075, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 58232 + }, + { + "epoch": 3.201984962406015, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 58233 + }, + { + "epoch": 3.202045112781955, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 58234 + }, + { + "epoch": 3.2021052631578946, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 58235 + }, + { + "epoch": 3.2021654135338347, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 58236 + }, + { + "epoch": 3.2022255639097743, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 58237 + }, + { + "epoch": 3.2022857142857144, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 58238 + }, + { + "epoch": 3.202345864661654, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9261, + "step": 58239 + }, + { + "epoch": 3.202406015037594, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 58240 + }, + { + "epoch": 3.202466165413534, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 58241 + }, + { + "epoch": 3.202526315789474, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 58242 + }, + { + "epoch": 3.2025864661654135, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 58243 + }, + { + "epoch": 3.202646616541353, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1225, + "step": 58244 + }, + { + "epoch": 3.2027067669172933, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9388, + "step": 58245 + }, + { + "epoch": 3.202766917293233, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 58246 + }, + { + "epoch": 3.202827067669173, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 58247 + }, + { + "epoch": 3.2028872180451127, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9431, + "step": 58248 + }, + { + "epoch": 3.2029473684210528, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 58249 + }, + { + "epoch": 3.2030075187969924, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.8303, + "step": 58250 + }, + { + "epoch": 3.2030676691729325, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.177, + "step": 58251 + }, + { + "epoch": 3.203127819548872, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 58252 + }, + { + "epoch": 3.2031879699248123, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 58253 + }, + { + "epoch": 3.203248120300752, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 58254 + }, + { + "epoch": 3.2033082706766915, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 58255 + }, + { + "epoch": 3.2033684210526316, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 58256 + }, + { + "epoch": 3.2034285714285713, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 58257 + }, + { + "epoch": 3.2034887218045114, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1928, + "step": 58258 + }, + { + "epoch": 3.203548872180451, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 58259 + }, + { + "epoch": 3.203609022556391, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 58260 + }, + { + "epoch": 3.2036691729323308, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9451, + "step": 58261 + }, + { + "epoch": 3.203729323308271, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 58262 + }, + { + "epoch": 3.2037894736842105, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 58263 + }, + { + "epoch": 3.20384962406015, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0789, + "step": 58264 + }, + { + "epoch": 3.2039097744360903, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 58265 + }, + { + "epoch": 3.20396992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 58266 + }, + { + "epoch": 3.20403007518797, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1452, + "step": 58267 + }, + { + "epoch": 3.2040902255639097, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9225, + "step": 58268 + }, + { + "epoch": 3.2041503759398497, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 58269 + }, + { + "epoch": 3.2042105263157894, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 58270 + }, + { + "epoch": 3.2042706766917295, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 58271 + }, + { + "epoch": 3.204330827067669, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 58272 + }, + { + "epoch": 3.204390977443609, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 58273 + }, + { + "epoch": 3.204451127819549, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 58274 + }, + { + "epoch": 3.2045112781954885, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 58275 + }, + { + "epoch": 3.2045714285714286, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 58276 + }, + { + "epoch": 3.2046315789473683, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 58277 + }, + { + "epoch": 3.2046917293233084, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 58278 + }, + { + "epoch": 3.204751879699248, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 58279 + }, + { + "epoch": 3.204812030075188, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0147, + "step": 58280 + }, + { + "epoch": 3.2048721804511278, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9088, + "step": 58281 + }, + { + "epoch": 3.204932330827068, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 58282 + }, + { + "epoch": 3.2049924812030075, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 58283 + }, + { + "epoch": 3.2050526315789476, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1818, + "step": 58284 + }, + { + "epoch": 3.2051127819548872, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 58285 + }, + { + "epoch": 3.205172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 58286 + }, + { + "epoch": 3.205233082706767, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 58287 + }, + { + "epoch": 3.2052932330827066, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 58288 + }, + { + "epoch": 3.2053533834586467, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 58289 + }, + { + "epoch": 3.2054135338345864, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 58290 + }, + { + "epoch": 3.2054736842105265, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 58291 + }, + { + "epoch": 3.205533834586466, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 58292 + }, + { + "epoch": 3.205593984962406, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9225, + "step": 58293 + }, + { + "epoch": 3.205654135338346, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 58294 + }, + { + "epoch": 3.2057142857142855, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9227, + "step": 58295 + }, + { + "epoch": 3.2057744360902256, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 58296 + }, + { + "epoch": 3.2058345864661653, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 58297 + }, + { + "epoch": 3.2058947368421054, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 58298 + }, + { + "epoch": 3.205954887218045, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 58299 + }, + { + "epoch": 3.206015037593985, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 58300 + }, + { + "epoch": 3.2060751879699247, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0441, + "step": 58301 + }, + { + "epoch": 3.206135338345865, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.925, + "step": 58302 + }, + { + "epoch": 3.2061954887218045, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 58303 + }, + { + "epoch": 3.206255639097744, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 58304 + }, + { + "epoch": 3.2063157894736842, + "grad_norm": 1.8359375, + "learning_rate": 3e-05, + "loss": 0.9162, + "step": 58305 + }, + { + "epoch": 3.206375939849624, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 58306 + }, + { + "epoch": 3.206436090225564, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 58307 + }, + { + "epoch": 3.2064962406015036, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 58308 + }, + { + "epoch": 3.2065563909774437, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 58309 + }, + { + "epoch": 3.2066165413533834, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 58310 + }, + { + "epoch": 3.2066766917293235, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 58311 + }, + { + "epoch": 3.206736842105263, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 58312 + }, + { + "epoch": 3.206796992481203, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 58313 + }, + { + "epoch": 3.206857142857143, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 58314 + }, + { + "epoch": 3.2069172932330825, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 58315 + }, + { + "epoch": 3.2069774436090226, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 58316 + }, + { + "epoch": 3.2070375939849622, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 58317 + }, + { + "epoch": 3.2070977443609023, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1, + "step": 58318 + }, + { + "epoch": 3.207157894736842, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 58319 + }, + { + "epoch": 3.207218045112782, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 58320 + }, + { + "epoch": 3.2072781954887217, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 58321 + }, + { + "epoch": 3.207338345864662, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1886, + "step": 58322 + }, + { + "epoch": 3.2073984962406015, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 58323 + }, + { + "epoch": 3.2074586466165416, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 58324 + }, + { + "epoch": 3.207518796992481, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 58325 + }, + { + "epoch": 3.207578947368421, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 58326 + }, + { + "epoch": 3.207639097744361, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 58327 + }, + { + "epoch": 3.2076992481203006, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 58328 + }, + { + "epoch": 3.2077593984962407, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 58329 + }, + { + "epoch": 3.2078195488721803, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 58330 + }, + { + "epoch": 3.2078796992481204, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 58331 + }, + { + "epoch": 3.20793984962406, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 58332 + }, + { + "epoch": 3.208, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 58333 + }, + { + "epoch": 3.20806015037594, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 58334 + }, + { + "epoch": 3.2081203007518795, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 58335 + }, + { + "epoch": 3.2081804511278196, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 58336 + }, + { + "epoch": 3.2082406015037592, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 58337 + }, + { + "epoch": 3.2083007518796993, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 58338 + }, + { + "epoch": 3.208360902255639, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 58339 + }, + { + "epoch": 3.208421052631579, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 58340 + }, + { + "epoch": 3.2084812030075187, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 58341 + }, + { + "epoch": 3.208541353383459, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 58342 + }, + { + "epoch": 3.2086015037593985, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 58343 + }, + { + "epoch": 3.2086616541353385, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 58344 + }, + { + "epoch": 3.208721804511278, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 58345 + }, + { + "epoch": 3.208781954887218, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 58346 + }, + { + "epoch": 3.208842105263158, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9085, + "step": 58347 + }, + { + "epoch": 3.2089022556390976, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 58348 + }, + { + "epoch": 3.2089624060150377, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 58349 + }, + { + "epoch": 3.2090225563909773, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9092, + "step": 58350 + }, + { + "epoch": 3.2090827067669174, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 58351 + }, + { + "epoch": 3.209142857142857, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 58352 + }, + { + "epoch": 3.209203007518797, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 58353 + }, + { + "epoch": 3.209263157894737, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.5293, + "step": 58354 + }, + { + "epoch": 3.209323308270677, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 58355 + }, + { + "epoch": 3.2093834586466166, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.8828, + "step": 58356 + }, + { + "epoch": 3.209443609022556, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 58357 + }, + { + "epoch": 3.2095037593984963, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 58358 + }, + { + "epoch": 3.209563909774436, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 58359 + }, + { + "epoch": 3.209624060150376, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 58360 + }, + { + "epoch": 3.2096842105263157, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9472, + "step": 58361 + }, + { + "epoch": 3.209744360902256, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 58362 + }, + { + "epoch": 3.2098045112781954, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 58363 + }, + { + "epoch": 3.2098646616541355, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 58364 + }, + { + "epoch": 3.209924812030075, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9038, + "step": 58365 + }, + { + "epoch": 3.209984962406015, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 58366 + }, + { + "epoch": 3.210045112781955, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 58367 + }, + { + "epoch": 3.2101052631578946, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 58368 + }, + { + "epoch": 3.2101654135338347, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 58369 + }, + { + "epoch": 3.2102255639097743, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 58370 + }, + { + "epoch": 3.2102857142857144, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9292, + "step": 58371 + }, + { + "epoch": 3.210345864661654, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.8803, + "step": 58372 + }, + { + "epoch": 3.210406015037594, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 58373 + }, + { + "epoch": 3.210466165413534, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9066, + "step": 58374 + }, + { + "epoch": 3.2105263157894735, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 58375 + }, + { + "epoch": 3.2105864661654135, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1897, + "step": 58376 + }, + { + "epoch": 3.210646616541353, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 58377 + }, + { + "epoch": 3.2107067669172933, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 58378 + }, + { + "epoch": 3.210766917293233, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 58379 + }, + { + "epoch": 3.210827067669173, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 58380 + }, + { + "epoch": 3.2108872180451127, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 58381 + }, + { + "epoch": 3.2109473684210528, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 58382 + }, + { + "epoch": 3.2110075187969924, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 58383 + }, + { + "epoch": 3.2110676691729325, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 58384 + }, + { + "epoch": 3.211127819548872, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 58385 + }, + { + "epoch": 3.2111879699248123, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 58386 + }, + { + "epoch": 3.211248120300752, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 58387 + }, + { + "epoch": 3.2113082706766916, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 58388 + }, + { + "epoch": 3.2113684210526316, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0462, + "step": 58389 + }, + { + "epoch": 3.2114285714285713, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 58390 + }, + { + "epoch": 3.2114887218045114, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 58391 + }, + { + "epoch": 3.211548872180451, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9256, + "step": 58392 + }, + { + "epoch": 3.211609022556391, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 58393 + }, + { + "epoch": 3.211669172932331, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 58394 + }, + { + "epoch": 3.211729323308271, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 58395 + }, + { + "epoch": 3.2117894736842105, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.976, + "step": 58396 + }, + { + "epoch": 3.21184962406015, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 58397 + }, + { + "epoch": 3.2119097744360903, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9489, + "step": 58398 + }, + { + "epoch": 3.21196992481203, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 58399 + }, + { + "epoch": 3.21203007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 58400 + }, + { + "epoch": 3.2120902255639097, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1197, + "step": 58401 + }, + { + "epoch": 3.2121503759398498, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 58402 + }, + { + "epoch": 3.2122105263157894, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 58403 + }, + { + "epoch": 3.2122706766917295, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0928, + "step": 58404 + }, + { + "epoch": 3.212330827067669, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 58405 + }, + { + "epoch": 3.212390977443609, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1673, + "step": 58406 + }, + { + "epoch": 3.212451127819549, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 58407 + }, + { + "epoch": 3.2125112781954885, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 58408 + }, + { + "epoch": 3.2125714285714286, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 58409 + }, + { + "epoch": 3.2126315789473683, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1984, + "step": 58410 + }, + { + "epoch": 3.2126917293233084, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9075, + "step": 58411 + }, + { + "epoch": 3.212751879699248, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 58412 + }, + { + "epoch": 3.212812030075188, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 58413 + }, + { + "epoch": 3.2128721804511278, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 58414 + }, + { + "epoch": 3.212932330827068, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 58415 + }, + { + "epoch": 3.2129924812030075, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9329, + "step": 58416 + }, + { + "epoch": 3.213052631578947, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 58417 + }, + { + "epoch": 3.2131127819548873, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 58418 + }, + { + "epoch": 3.213172932330827, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 58419 + }, + { + "epoch": 3.213233082706767, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 58420 + }, + { + "epoch": 3.2132932330827066, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 58421 + }, + { + "epoch": 3.2133533834586467, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 58422 + }, + { + "epoch": 3.2134135338345864, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 58423 + }, + { + "epoch": 3.2134736842105265, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 58424 + }, + { + "epoch": 3.213533834586466, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 58425 + }, + { + "epoch": 3.2135939849624062, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 58426 + }, + { + "epoch": 3.213654135338346, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9148, + "step": 58427 + }, + { + "epoch": 3.2137142857142855, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 58428 + }, + { + "epoch": 3.2137744360902256, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 58429 + }, + { + "epoch": 3.2138345864661653, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0878, + "step": 58430 + }, + { + "epoch": 3.2138947368421054, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 58431 + }, + { + "epoch": 3.213954887218045, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 58432 + }, + { + "epoch": 3.214015037593985, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 58433 + }, + { + "epoch": 3.2140751879699248, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 58434 + }, + { + "epoch": 3.214135338345865, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 58435 + }, + { + "epoch": 3.2141954887218045, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 58436 + }, + { + "epoch": 3.214255639097744, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 58437 + }, + { + "epoch": 3.2143157894736842, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 58438 + }, + { + "epoch": 3.214375939849624, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.941, + "step": 58439 + }, + { + "epoch": 3.214436090225564, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.181, + "step": 58440 + }, + { + "epoch": 3.2144962406015036, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 58441 + }, + { + "epoch": 3.2145563909774437, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 58442 + }, + { + "epoch": 3.2146165413533834, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9128, + "step": 58443 + }, + { + "epoch": 3.2146766917293235, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 58444 + }, + { + "epoch": 3.214736842105263, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 58445 + }, + { + "epoch": 3.214796992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9177, + "step": 58446 + }, + { + "epoch": 3.214857142857143, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 58447 + }, + { + "epoch": 3.2149172932330825, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 58448 + }, + { + "epoch": 3.2149774436090226, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 58449 + }, + { + "epoch": 3.2150375939849622, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 58450 + }, + { + "epoch": 3.2150977443609023, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 58451 + }, + { + "epoch": 3.215157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 58452 + }, + { + "epoch": 3.215218045112782, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 58453 + }, + { + "epoch": 3.2152781954887217, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 58454 + }, + { + "epoch": 3.215338345864662, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9396, + "step": 58455 + }, + { + "epoch": 3.2153984962406015, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 58456 + }, + { + "epoch": 3.2154586466165416, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9357, + "step": 58457 + }, + { + "epoch": 3.215518796992481, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 58458 + }, + { + "epoch": 3.215578947368421, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 58459 + }, + { + "epoch": 3.215639097744361, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 58460 + }, + { + "epoch": 3.2156992481203006, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 58461 + }, + { + "epoch": 3.2157593984962407, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 58462 + }, + { + "epoch": 3.2158195488721804, + "grad_norm": 1.7578125, + "learning_rate": 3e-05, + "loss": 1.1189, + "step": 58463 + }, + { + "epoch": 3.2158796992481204, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 58464 + }, + { + "epoch": 3.21593984962406, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 58465 + }, + { + "epoch": 3.216, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 58466 + }, + { + "epoch": 3.21606015037594, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 58467 + }, + { + "epoch": 3.2161203007518795, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 58468 + }, + { + "epoch": 3.2161804511278196, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 58469 + }, + { + "epoch": 3.2162406015037592, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 58470 + }, + { + "epoch": 3.2163007518796993, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9429, + "step": 58471 + }, + { + "epoch": 3.216360902255639, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 58472 + }, + { + "epoch": 3.216421052631579, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0888, + "step": 58473 + }, + { + "epoch": 3.2164812030075187, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 58474 + }, + { + "epoch": 3.216541353383459, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 58475 + }, + { + "epoch": 3.2166015037593985, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 58476 + }, + { + "epoch": 3.216661654135338, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 58477 + }, + { + "epoch": 3.216721804511278, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1087, + "step": 58478 + }, + { + "epoch": 3.216781954887218, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 58479 + }, + { + "epoch": 3.216842105263158, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 58480 + }, + { + "epoch": 3.2169022556390976, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 58481 + }, + { + "epoch": 3.2169624060150377, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 58482 + }, + { + "epoch": 3.2170225563909773, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9481, + "step": 58483 + }, + { + "epoch": 3.2170827067669174, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 58484 + }, + { + "epoch": 3.217142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 58485 + }, + { + "epoch": 3.217203007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 58486 + }, + { + "epoch": 3.217263157894737, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 58487 + }, + { + "epoch": 3.217323308270677, + "grad_norm": 1.5234375, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 58488 + }, + { + "epoch": 3.2173834586466166, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 58489 + }, + { + "epoch": 3.217443609022556, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 58490 + }, + { + "epoch": 3.2175037593984963, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 58491 + }, + { + "epoch": 3.217563909774436, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 58492 + }, + { + "epoch": 3.217624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 58493 + }, + { + "epoch": 3.2176842105263157, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 58494 + }, + { + "epoch": 3.217744360902256, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 58495 + }, + { + "epoch": 3.2178045112781954, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 58496 + }, + { + "epoch": 3.2178646616541355, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9444, + "step": 58497 + }, + { + "epoch": 3.217924812030075, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9313, + "step": 58498 + }, + { + "epoch": 3.217984962406015, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 58499 + }, + { + "epoch": 3.218045112781955, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 58500 + }, + { + "epoch": 3.2181052631578946, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 58501 + }, + { + "epoch": 3.2181654135338347, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0634, + "step": 58502 + }, + { + "epoch": 3.2182255639097743, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 58503 + }, + { + "epoch": 3.2182857142857144, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9081, + "step": 58504 + }, + { + "epoch": 3.218345864661654, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 58505 + }, + { + "epoch": 3.218406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 58506 + }, + { + "epoch": 3.218466165413534, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 58507 + }, + { + "epoch": 3.2185263157894735, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 58508 + }, + { + "epoch": 3.2185864661654136, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0914, + "step": 58509 + }, + { + "epoch": 3.218646616541353, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 58510 + }, + { + "epoch": 3.2187067669172933, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 58511 + }, + { + "epoch": 3.218766917293233, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 58512 + }, + { + "epoch": 3.218827067669173, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 58513 + }, + { + "epoch": 3.2188872180451127, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0272, + "step": 58514 + }, + { + "epoch": 3.218947368421053, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 58515 + }, + { + "epoch": 3.2190075187969924, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 58516 + }, + { + "epoch": 3.2190676691729325, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 58517 + }, + { + "epoch": 3.219127819548872, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 58518 + }, + { + "epoch": 3.2191879699248123, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 58519 + }, + { + "epoch": 3.219248120300752, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 58520 + }, + { + "epoch": 3.2193082706766916, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 58521 + }, + { + "epoch": 3.2193684210526317, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 58522 + }, + { + "epoch": 3.2194285714285713, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 58523 + }, + { + "epoch": 3.2194887218045114, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 58524 + }, + { + "epoch": 3.219548872180451, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 58525 + }, + { + "epoch": 3.219609022556391, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 58526 + }, + { + "epoch": 3.219669172932331, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 58527 + }, + { + "epoch": 3.219729323308271, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 58528 + }, + { + "epoch": 3.2197894736842105, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 58529 + }, + { + "epoch": 3.21984962406015, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 58530 + }, + { + "epoch": 3.2199097744360903, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 58531 + }, + { + "epoch": 3.21996992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9074, + "step": 58532 + }, + { + "epoch": 3.22003007518797, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 58533 + }, + { + "epoch": 3.2200902255639097, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 58534 + }, + { + "epoch": 3.2201503759398498, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 58535 + }, + { + "epoch": 3.2202105263157894, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 58536 + }, + { + "epoch": 3.2202706766917295, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 58537 + }, + { + "epoch": 3.220330827067669, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 58538 + }, + { + "epoch": 3.220390977443609, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 58539 + }, + { + "epoch": 3.220451127819549, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 58540 + }, + { + "epoch": 3.2205112781954885, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1212, + "step": 58541 + }, + { + "epoch": 3.2205714285714286, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 58542 + }, + { + "epoch": 3.2206315789473683, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 58543 + }, + { + "epoch": 3.2206917293233084, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9079, + "step": 58544 + }, + { + "epoch": 3.220751879699248, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 58545 + }, + { + "epoch": 3.220812030075188, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 58546 + }, + { + "epoch": 3.2208721804511278, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 58547 + }, + { + "epoch": 3.220932330827068, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 58548 + }, + { + "epoch": 3.2209924812030075, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 58549 + }, + { + "epoch": 3.221052631578947, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 58550 + }, + { + "epoch": 3.2211127819548873, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 58551 + }, + { + "epoch": 3.221172932330827, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 58552 + }, + { + "epoch": 3.221233082706767, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 58553 + }, + { + "epoch": 3.2212932330827067, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 58554 + }, + { + "epoch": 3.2213533834586467, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 58555 + }, + { + "epoch": 3.2214135338345864, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9001, + "step": 58556 + }, + { + "epoch": 3.2214736842105265, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9495, + "step": 58557 + }, + { + "epoch": 3.221533834586466, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9184, + "step": 58558 + }, + { + "epoch": 3.2215939849624062, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 58559 + }, + { + "epoch": 3.221654135338346, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 58560 + }, + { + "epoch": 3.2217142857142855, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 58561 + }, + { + "epoch": 3.2217744360902256, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 58562 + }, + { + "epoch": 3.2218345864661653, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 58563 + }, + { + "epoch": 3.2218947368421054, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9307, + "step": 58564 + }, + { + "epoch": 3.221954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 58565 + }, + { + "epoch": 3.222015037593985, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 58566 + }, + { + "epoch": 3.2220751879699248, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 58567 + }, + { + "epoch": 3.222135338345865, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9364, + "step": 58568 + }, + { + "epoch": 3.2221954887218045, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 58569 + }, + { + "epoch": 3.222255639097744, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 58570 + }, + { + "epoch": 3.2223157894736842, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 58571 + }, + { + "epoch": 3.222375939849624, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 58572 + }, + { + "epoch": 3.222436090225564, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 58573 + }, + { + "epoch": 3.2224962406015036, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 58574 + }, + { + "epoch": 3.2225563909774437, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 58575 + }, + { + "epoch": 3.2226165413533834, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 58576 + }, + { + "epoch": 3.2226766917293235, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 58577 + }, + { + "epoch": 3.222736842105263, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 58578 + }, + { + "epoch": 3.222796992481203, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 58579 + }, + { + "epoch": 3.222857142857143, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 58580 + }, + { + "epoch": 3.2229172932330825, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 58581 + }, + { + "epoch": 3.2229774436090226, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 58582 + }, + { + "epoch": 3.2230375939849623, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 58583 + }, + { + "epoch": 3.2230977443609024, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 58584 + }, + { + "epoch": 3.223157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 58585 + }, + { + "epoch": 3.223218045112782, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 58586 + }, + { + "epoch": 3.2232781954887217, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 58587 + }, + { + "epoch": 3.223338345864662, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 58588 + }, + { + "epoch": 3.2233984962406015, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 58589 + }, + { + "epoch": 3.2234586466165416, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 58590 + }, + { + "epoch": 3.2235187969924812, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 58591 + }, + { + "epoch": 3.223578947368421, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 58592 + }, + { + "epoch": 3.223639097744361, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 58593 + }, + { + "epoch": 3.2236992481203006, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 58594 + }, + { + "epoch": 3.2237593984962407, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 58595 + }, + { + "epoch": 3.2238195488721804, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 58596 + }, + { + "epoch": 3.2238796992481205, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 58597 + }, + { + "epoch": 3.22393984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0768, + "step": 58598 + }, + { + "epoch": 3.224, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 58599 + }, + { + "epoch": 3.22406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.914, + "step": 58600 + }, + { + "epoch": 3.2241203007518795, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 58601 + }, + { + "epoch": 3.2241804511278196, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 58602 + }, + { + "epoch": 3.2242406015037592, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9226, + "step": 58603 + }, + { + "epoch": 3.2243007518796993, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 58604 + }, + { + "epoch": 3.224360902255639, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 58605 + }, + { + "epoch": 3.224421052631579, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 58606 + }, + { + "epoch": 3.2244812030075187, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 58607 + }, + { + "epoch": 3.224541353383459, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9221, + "step": 58608 + }, + { + "epoch": 3.2246015037593985, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 58609 + }, + { + "epoch": 3.224661654135338, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 58610 + }, + { + "epoch": 3.224721804511278, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 58611 + }, + { + "epoch": 3.224781954887218, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 58612 + }, + { + "epoch": 3.224842105263158, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0959, + "step": 58613 + }, + { + "epoch": 3.2249022556390976, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 58614 + }, + { + "epoch": 3.2249624060150377, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 58615 + }, + { + "epoch": 3.2250225563909773, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 58616 + }, + { + "epoch": 3.2250827067669174, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 58617 + }, + { + "epoch": 3.225142857142857, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 58618 + }, + { + "epoch": 3.225203007518797, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 58619 + }, + { + "epoch": 3.225263157894737, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 58620 + }, + { + "epoch": 3.225323308270677, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 58621 + }, + { + "epoch": 3.2253834586466166, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 58622 + }, + { + "epoch": 3.2254436090225562, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 58623 + }, + { + "epoch": 3.2255037593984963, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.8949, + "step": 58624 + }, + { + "epoch": 3.225563909774436, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 58625 + }, + { + "epoch": 3.225624060150376, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 58626 + }, + { + "epoch": 3.2256842105263157, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 58627 + }, + { + "epoch": 3.225744360902256, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9325, + "step": 58628 + }, + { + "epoch": 3.2258045112781955, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 58629 + }, + { + "epoch": 3.2258646616541355, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 58630 + }, + { + "epoch": 3.225924812030075, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 58631 + }, + { + "epoch": 3.225984962406015, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 58632 + }, + { + "epoch": 3.226045112781955, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 58633 + }, + { + "epoch": 3.2261052631578946, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 58634 + }, + { + "epoch": 3.2261654135338347, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.1588, + "step": 58635 + }, + { + "epoch": 3.2262255639097743, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 58636 + }, + { + "epoch": 3.2262857142857144, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1035, + "step": 58637 + }, + { + "epoch": 3.226345864661654, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0542, + "step": 58638 + }, + { + "epoch": 3.226406015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 58639 + }, + { + "epoch": 3.226466165413534, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 58640 + }, + { + "epoch": 3.2265263157894735, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.98, + "step": 58641 + }, + { + "epoch": 3.2265864661654136, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 58642 + }, + { + "epoch": 3.226646616541353, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 58643 + }, + { + "epoch": 3.2267067669172933, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 58644 + }, + { + "epoch": 3.226766917293233, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 58645 + }, + { + "epoch": 3.226827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9427, + "step": 58646 + }, + { + "epoch": 3.2268872180451127, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 58647 + }, + { + "epoch": 3.226947368421053, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9216, + "step": 58648 + }, + { + "epoch": 3.2270075187969924, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 58649 + }, + { + "epoch": 3.2270676691729325, + "grad_norm": 0.484375, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 58650 + }, + { + "epoch": 3.227127819548872, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 58651 + }, + { + "epoch": 3.227187969924812, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 58652 + }, + { + "epoch": 3.227248120300752, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 58653 + }, + { + "epoch": 3.2273082706766916, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9482, + "step": 58654 + }, + { + "epoch": 3.2273684210526317, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 58655 + }, + { + "epoch": 3.2274285714285713, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 58656 + }, + { + "epoch": 3.2274887218045114, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 58657 + }, + { + "epoch": 3.227548872180451, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 58658 + }, + { + "epoch": 3.227609022556391, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 58659 + }, + { + "epoch": 3.227669172932331, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 58660 + }, + { + "epoch": 3.227729323308271, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 58661 + }, + { + "epoch": 3.2277894736842105, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 58662 + }, + { + "epoch": 3.22784962406015, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 58663 + }, + { + "epoch": 3.2279097744360903, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 58664 + }, + { + "epoch": 3.22796992481203, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 58665 + }, + { + "epoch": 3.22803007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 58666 + }, + { + "epoch": 3.2280902255639097, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 58667 + }, + { + "epoch": 3.2281503759398498, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 58668 + }, + { + "epoch": 3.2282105263157894, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9411, + "step": 58669 + }, + { + "epoch": 3.2282706766917295, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 58670 + }, + { + "epoch": 3.228330827067669, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9228, + "step": 58671 + }, + { + "epoch": 3.228390977443609, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 58672 + }, + { + "epoch": 3.228451127819549, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 58673 + }, + { + "epoch": 3.2285112781954886, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 58674 + }, + { + "epoch": 3.2285714285714286, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9081, + "step": 58675 + }, + { + "epoch": 3.2286315789473683, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 58676 + }, + { + "epoch": 3.2286917293233084, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 58677 + }, + { + "epoch": 3.228751879699248, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 58678 + }, + { + "epoch": 3.228812030075188, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 58679 + }, + { + "epoch": 3.228872180451128, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 58680 + }, + { + "epoch": 3.228932330827068, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 58681 + }, + { + "epoch": 3.2289924812030075, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 58682 + }, + { + "epoch": 3.229052631578947, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 58683 + }, + { + "epoch": 3.2291127819548873, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.8865, + "step": 58684 + }, + { + "epoch": 3.229172932330827, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9149, + "step": 58685 + }, + { + "epoch": 3.229233082706767, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 58686 + }, + { + "epoch": 3.2292932330827067, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 58687 + }, + { + "epoch": 3.2293533834586468, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 58688 + }, + { + "epoch": 3.2294135338345864, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9443, + "step": 58689 + }, + { + "epoch": 3.2294736842105265, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 58690 + }, + { + "epoch": 3.229533834586466, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 58691 + }, + { + "epoch": 3.2295939849624062, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.2788, + "step": 58692 + }, + { + "epoch": 3.229654135338346, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 58693 + }, + { + "epoch": 3.2297142857142855, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 58694 + }, + { + "epoch": 3.2297744360902256, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 58695 + }, + { + "epoch": 3.2298345864661653, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9459, + "step": 58696 + }, + { + "epoch": 3.2298947368421054, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.8983, + "step": 58697 + }, + { + "epoch": 3.229954887218045, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 58698 + }, + { + "epoch": 3.230015037593985, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 58699 + }, + { + "epoch": 3.2300751879699248, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 58700 + }, + { + "epoch": 3.230135338345865, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 58701 + }, + { + "epoch": 3.2301954887218045, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 58702 + }, + { + "epoch": 3.230255639097744, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 58703 + }, + { + "epoch": 3.2303157894736843, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9086, + "step": 58704 + }, + { + "epoch": 3.230375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 58705 + }, + { + "epoch": 3.230436090225564, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 58706 + }, + { + "epoch": 3.2304962406015036, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9374, + "step": 58707 + }, + { + "epoch": 3.2305563909774437, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 58708 + }, + { + "epoch": 3.2306165413533834, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 58709 + }, + { + "epoch": 3.2306766917293235, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 58710 + }, + { + "epoch": 3.230736842105263, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9264, + "step": 58711 + }, + { + "epoch": 3.230796992481203, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 58712 + }, + { + "epoch": 3.230857142857143, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 58713 + }, + { + "epoch": 3.2309172932330825, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9619, + "step": 58714 + }, + { + "epoch": 3.2309774436090226, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 58715 + }, + { + "epoch": 3.2310375939849623, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 58716 + }, + { + "epoch": 3.2310977443609024, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9412, + "step": 58717 + }, + { + "epoch": 3.231157894736842, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 58718 + }, + { + "epoch": 3.231218045112782, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 58719 + }, + { + "epoch": 3.2312781954887217, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 58720 + }, + { + "epoch": 3.231338345864662, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 58721 + }, + { + "epoch": 3.2313984962406015, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 58722 + }, + { + "epoch": 3.2314586466165416, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 58723 + }, + { + "epoch": 3.2315187969924812, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 58724 + }, + { + "epoch": 3.231578947368421, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 58725 + }, + { + "epoch": 3.231639097744361, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 58726 + }, + { + "epoch": 3.2316992481203006, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 58727 + }, + { + "epoch": 3.2317593984962407, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 58728 + }, + { + "epoch": 3.2318195488721804, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 58729 + }, + { + "epoch": 3.2318796992481205, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 58730 + }, + { + "epoch": 3.23193984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 58731 + }, + { + "epoch": 3.232, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 58732 + }, + { + "epoch": 3.23206015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0565, + "step": 58733 + }, + { + "epoch": 3.2321203007518795, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 58734 + }, + { + "epoch": 3.2321804511278196, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 58735 + }, + { + "epoch": 3.2322406015037592, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 58736 + }, + { + "epoch": 3.2323007518796993, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 58737 + }, + { + "epoch": 3.232360902255639, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0688, + "step": 58738 + }, + { + "epoch": 3.232421052631579, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 58739 + }, + { + "epoch": 3.2324812030075187, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.127, + "step": 58740 + }, + { + "epoch": 3.232541353383459, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 58741 + }, + { + "epoch": 3.2326015037593985, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.12, + "step": 58742 + }, + { + "epoch": 3.232661654135338, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 58743 + }, + { + "epoch": 3.232721804511278, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 58744 + }, + { + "epoch": 3.232781954887218, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 58745 + }, + { + "epoch": 3.232842105263158, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.921, + "step": 58746 + }, + { + "epoch": 3.2329022556390976, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 58747 + }, + { + "epoch": 3.2329624060150377, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 58748 + }, + { + "epoch": 3.2330225563909774, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 58749 + }, + { + "epoch": 3.2330827067669174, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 58750 + }, + { + "epoch": 3.233142857142857, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 58751 + }, + { + "epoch": 3.233203007518797, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 58752 + }, + { + "epoch": 3.233263157894737, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 58753 + }, + { + "epoch": 3.2333233082706765, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 58754 + }, + { + "epoch": 3.2333834586466166, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 58755 + }, + { + "epoch": 3.2334436090225562, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.159, + "step": 58756 + }, + { + "epoch": 3.2335037593984963, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 58757 + }, + { + "epoch": 3.233563909774436, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9342, + "step": 58758 + }, + { + "epoch": 3.233624060150376, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 58759 + }, + { + "epoch": 3.2336842105263157, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.079, + "step": 58760 + }, + { + "epoch": 3.233744360902256, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 58761 + }, + { + "epoch": 3.2338045112781955, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 58762 + }, + { + "epoch": 3.2338646616541356, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 58763 + }, + { + "epoch": 3.233924812030075, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 58764 + }, + { + "epoch": 3.233984962406015, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1506, + "step": 58765 + }, + { + "epoch": 3.234045112781955, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 58766 + }, + { + "epoch": 3.2341052631578946, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 58767 + }, + { + "epoch": 3.2341654135338347, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9194, + "step": 58768 + }, + { + "epoch": 3.2342255639097743, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 58769 + }, + { + "epoch": 3.2342857142857144, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 58770 + }, + { + "epoch": 3.234345864661654, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 58771 + }, + { + "epoch": 3.234406015037594, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 58772 + }, + { + "epoch": 3.234466165413534, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9812, + "step": 58773 + }, + { + "epoch": 3.2345263157894735, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 58774 + }, + { + "epoch": 3.2345864661654136, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9761, + "step": 58775 + }, + { + "epoch": 3.234646616541353, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 58776 + }, + { + "epoch": 3.2347067669172933, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 58777 + }, + { + "epoch": 3.234766917293233, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 58778 + }, + { + "epoch": 3.234827067669173, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 58779 + }, + { + "epoch": 3.2348872180451127, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 58780 + }, + { + "epoch": 3.234947368421053, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 58781 + }, + { + "epoch": 3.2350075187969924, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 58782 + }, + { + "epoch": 3.2350676691729325, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 58783 + }, + { + "epoch": 3.235127819548872, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 58784 + }, + { + "epoch": 3.235187969924812, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 58785 + }, + { + "epoch": 3.235248120300752, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 58786 + }, + { + "epoch": 3.2353082706766916, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 58787 + }, + { + "epoch": 3.2353684210526317, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 58788 + }, + { + "epoch": 3.2354285714285713, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 58789 + }, + { + "epoch": 3.2354887218045114, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0802, + "step": 58790 + }, + { + "epoch": 3.235548872180451, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 58791 + }, + { + "epoch": 3.235609022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 58792 + }, + { + "epoch": 3.235669172932331, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 58793 + }, + { + "epoch": 3.235729323308271, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9206, + "step": 58794 + }, + { + "epoch": 3.2357894736842105, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 58795 + }, + { + "epoch": 3.23584962406015, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 58796 + }, + { + "epoch": 3.2359097744360903, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 58797 + }, + { + "epoch": 3.23596992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 58798 + }, + { + "epoch": 3.23603007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 58799 + }, + { + "epoch": 3.2360902255639097, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 58800 + }, + { + "epoch": 3.2361503759398498, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 58801 + }, + { + "epoch": 3.2362105263157894, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 58802 + }, + { + "epoch": 3.2362706766917295, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 58803 + }, + { + "epoch": 3.236330827067669, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 58804 + }, + { + "epoch": 3.236390977443609, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 58805 + }, + { + "epoch": 3.236451127819549, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 58806 + }, + { + "epoch": 3.2365112781954886, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 58807 + }, + { + "epoch": 3.2365714285714287, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 58808 + }, + { + "epoch": 3.2366315789473683, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 58809 + }, + { + "epoch": 3.2366917293233084, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 58810 + }, + { + "epoch": 3.236751879699248, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 58811 + }, + { + "epoch": 3.236812030075188, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 58812 + }, + { + "epoch": 3.236872180451128, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 58813 + }, + { + "epoch": 3.2369323308270674, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 58814 + }, + { + "epoch": 3.2369924812030075, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0557, + "step": 58815 + }, + { + "epoch": 3.237052631578947, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9384, + "step": 58816 + }, + { + "epoch": 3.2371127819548873, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 58817 + }, + { + "epoch": 3.237172932330827, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 58818 + }, + { + "epoch": 3.237233082706767, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 58819 + }, + { + "epoch": 3.2372932330827067, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9495, + "step": 58820 + }, + { + "epoch": 3.2373533834586468, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 58821 + }, + { + "epoch": 3.2374135338345864, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 58822 + }, + { + "epoch": 3.2374736842105265, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0437, + "step": 58823 + }, + { + "epoch": 3.237533834586466, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0756, + "step": 58824 + }, + { + "epoch": 3.2375939849624062, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 58825 + }, + { + "epoch": 3.237654135338346, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9421, + "step": 58826 + }, + { + "epoch": 3.2377142857142855, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9345, + "step": 58827 + }, + { + "epoch": 3.2377744360902256, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 58828 + }, + { + "epoch": 3.2378345864661653, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1096, + "step": 58829 + }, + { + "epoch": 3.2378947368421054, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9067, + "step": 58830 + }, + { + "epoch": 3.237954887218045, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 58831 + }, + { + "epoch": 3.238015037593985, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9199, + "step": 58832 + }, + { + "epoch": 3.2380751879699248, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0692, + "step": 58833 + }, + { + "epoch": 3.238135338345865, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 58834 + }, + { + "epoch": 3.2381954887218045, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9148, + "step": 58835 + }, + { + "epoch": 3.238255639097744, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 58836 + }, + { + "epoch": 3.2383157894736843, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 58837 + }, + { + "epoch": 3.238375939849624, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9553, + "step": 58838 + }, + { + "epoch": 3.238436090225564, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 58839 + }, + { + "epoch": 3.2384962406015037, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 58840 + }, + { + "epoch": 3.2385563909774437, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 58841 + }, + { + "epoch": 3.2386165413533834, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 58842 + }, + { + "epoch": 3.2386766917293235, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 58843 + }, + { + "epoch": 3.238736842105263, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1511, + "step": 58844 + }, + { + "epoch": 3.238796992481203, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 58845 + }, + { + "epoch": 3.238857142857143, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 58846 + }, + { + "epoch": 3.2389172932330825, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.1143, + "step": 58847 + }, + { + "epoch": 3.2389774436090226, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 58848 + }, + { + "epoch": 3.2390375939849623, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9429, + "step": 58849 + }, + { + "epoch": 3.2390977443609024, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 58850 + }, + { + "epoch": 3.239157894736842, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 58851 + }, + { + "epoch": 3.239218045112782, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 58852 + }, + { + "epoch": 3.2392781954887218, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 58853 + }, + { + "epoch": 3.239338345864662, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9141, + "step": 58854 + }, + { + "epoch": 3.2393984962406015, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 58855 + }, + { + "epoch": 3.239458646616541, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 58856 + }, + { + "epoch": 3.2395187969924812, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.8663, + "step": 58857 + }, + { + "epoch": 3.239578947368421, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 58858 + }, + { + "epoch": 3.239639097744361, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9202, + "step": 58859 + }, + { + "epoch": 3.2396992481203006, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 58860 + }, + { + "epoch": 3.2397593984962407, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9326, + "step": 58861 + }, + { + "epoch": 3.2398195488721804, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9306, + "step": 58862 + }, + { + "epoch": 3.2398796992481205, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9431, + "step": 58863 + }, + { + "epoch": 3.23993984962406, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 58864 + }, + { + "epoch": 3.24, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 58865 + }, + { + "epoch": 3.24006015037594, + "grad_norm": 1.515625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 58866 + }, + { + "epoch": 3.2401203007518795, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 58867 + }, + { + "epoch": 3.2401804511278196, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 58868 + }, + { + "epoch": 3.2402406015037593, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9382, + "step": 58869 + }, + { + "epoch": 3.2403007518796993, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9315, + "step": 58870 + }, + { + "epoch": 3.240360902255639, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 58871 + }, + { + "epoch": 3.240421052631579, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 58872 + }, + { + "epoch": 3.2404812030075187, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 58873 + }, + { + "epoch": 3.240541353383459, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 58874 + }, + { + "epoch": 3.2406015037593985, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 58875 + }, + { + "epoch": 3.240661654135338, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 58876 + }, + { + "epoch": 3.2407218045112782, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 58877 + }, + { + "epoch": 3.240781954887218, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 58878 + }, + { + "epoch": 3.240842105263158, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 58879 + }, + { + "epoch": 3.2409022556390976, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9245, + "step": 58880 + }, + { + "epoch": 3.2409624060150377, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 58881 + }, + { + "epoch": 3.2410225563909774, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0289, + "step": 58882 + }, + { + "epoch": 3.2410827067669175, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 58883 + }, + { + "epoch": 3.241142857142857, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 58884 + }, + { + "epoch": 3.241203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 58885 + }, + { + "epoch": 3.241263157894737, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 58886 + }, + { + "epoch": 3.2413233082706765, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 58887 + }, + { + "epoch": 3.2413834586466166, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.1967, + "step": 58888 + }, + { + "epoch": 3.2414436090225562, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0714, + "step": 58889 + }, + { + "epoch": 3.2415037593984963, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 58890 + }, + { + "epoch": 3.241563909774436, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 58891 + }, + { + "epoch": 3.241624060150376, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 58892 + }, + { + "epoch": 3.2416842105263157, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 58893 + }, + { + "epoch": 3.241744360902256, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 58894 + }, + { + "epoch": 3.2418045112781955, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 58895 + }, + { + "epoch": 3.2418646616541356, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 58896 + }, + { + "epoch": 3.241924812030075, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 58897 + }, + { + "epoch": 3.241984962406015, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.073, + "step": 58898 + }, + { + "epoch": 3.242045112781955, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1557, + "step": 58899 + }, + { + "epoch": 3.2421052631578946, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 58900 + }, + { + "epoch": 3.2421654135338347, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 58901 + }, + { + "epoch": 3.2422255639097743, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 58902 + }, + { + "epoch": 3.2422857142857144, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 58903 + }, + { + "epoch": 3.242345864661654, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 58904 + }, + { + "epoch": 3.242406015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9194, + "step": 58905 + }, + { + "epoch": 3.242466165413534, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 58906 + }, + { + "epoch": 3.2425263157894735, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 58907 + }, + { + "epoch": 3.2425864661654136, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 58908 + }, + { + "epoch": 3.242646616541353, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 58909 + }, + { + "epoch": 3.2427067669172933, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 58910 + }, + { + "epoch": 3.242766917293233, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 58911 + }, + { + "epoch": 3.242827067669173, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 58912 + }, + { + "epoch": 3.2428872180451127, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9254, + "step": 58913 + }, + { + "epoch": 3.242947368421053, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 58914 + }, + { + "epoch": 3.2430075187969924, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9518, + "step": 58915 + }, + { + "epoch": 3.243067669172932, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0433, + "step": 58916 + }, + { + "epoch": 3.243127819548872, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.052, + "step": 58917 + }, + { + "epoch": 3.243187969924812, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0837, + "step": 58918 + }, + { + "epoch": 3.243248120300752, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 58919 + }, + { + "epoch": 3.2433082706766916, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 58920 + }, + { + "epoch": 3.2433684210526317, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 58921 + }, + { + "epoch": 3.2434285714285713, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9376, + "step": 58922 + }, + { + "epoch": 3.2434887218045114, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 58923 + }, + { + "epoch": 3.243548872180451, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9343, + "step": 58924 + }, + { + "epoch": 3.243609022556391, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9232, + "step": 58925 + }, + { + "epoch": 3.243669172932331, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 58926 + }, + { + "epoch": 3.243729323308271, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 58927 + }, + { + "epoch": 3.2437894736842106, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 58928 + }, + { + "epoch": 3.24384962406015, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 58929 + }, + { + "epoch": 3.2439097744360903, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 58930 + }, + { + "epoch": 3.24396992481203, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 58931 + }, + { + "epoch": 3.24403007518797, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 58932 + }, + { + "epoch": 3.2440902255639097, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 58933 + }, + { + "epoch": 3.24415037593985, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1828, + "step": 58934 + }, + { + "epoch": 3.2442105263157894, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9206, + "step": 58935 + }, + { + "epoch": 3.2442706766917295, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 58936 + }, + { + "epoch": 3.244330827067669, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 58937 + }, + { + "epoch": 3.244390977443609, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 58938 + }, + { + "epoch": 3.244451127819549, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 58939 + }, + { + "epoch": 3.2445112781954886, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 58940 + }, + { + "epoch": 3.2445714285714287, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 58941 + }, + { + "epoch": 3.2446315789473683, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 58942 + }, + { + "epoch": 3.2446917293233084, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 58943 + }, + { + "epoch": 3.244751879699248, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 58944 + }, + { + "epoch": 3.244812030075188, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 58945 + }, + { + "epoch": 3.244872180451128, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9147, + "step": 58946 + }, + { + "epoch": 3.2449323308270674, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 58947 + }, + { + "epoch": 3.2449924812030075, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 58948 + }, + { + "epoch": 3.245052631578947, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 58949 + }, + { + "epoch": 3.2451127819548873, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 58950 + }, + { + "epoch": 3.245172932330827, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.8598, + "step": 58951 + }, + { + "epoch": 3.245233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 58952 + }, + { + "epoch": 3.2452932330827067, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 58953 + }, + { + "epoch": 3.2453533834586468, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 58954 + }, + { + "epoch": 3.2454135338345864, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 58955 + }, + { + "epoch": 3.2454736842105265, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9298, + "step": 58956 + }, + { + "epoch": 3.245533834586466, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 58957 + }, + { + "epoch": 3.245593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0932, + "step": 58958 + }, + { + "epoch": 3.245654135338346, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 58959 + }, + { + "epoch": 3.2457142857142856, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 58960 + }, + { + "epoch": 3.2457744360902256, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 58961 + }, + { + "epoch": 3.2458345864661653, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 58962 + }, + { + "epoch": 3.2458947368421054, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 58963 + }, + { + "epoch": 3.245954887218045, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 58964 + }, + { + "epoch": 3.246015037593985, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1533, + "step": 58965 + }, + { + "epoch": 3.246075187969925, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 58966 + }, + { + "epoch": 3.246135338345865, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 58967 + }, + { + "epoch": 3.2461954887218045, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 58968 + }, + { + "epoch": 3.246255639097744, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 58969 + }, + { + "epoch": 3.2463157894736843, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 58970 + }, + { + "epoch": 3.246375939849624, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 58971 + }, + { + "epoch": 3.246436090225564, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 58972 + }, + { + "epoch": 3.2464962406015037, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 58973 + }, + { + "epoch": 3.2465563909774438, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 58974 + }, + { + "epoch": 3.2466165413533834, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 58975 + }, + { + "epoch": 3.2466766917293235, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 58976 + }, + { + "epoch": 3.246736842105263, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 58977 + }, + { + "epoch": 3.246796992481203, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 58978 + }, + { + "epoch": 3.246857142857143, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 58979 + }, + { + "epoch": 3.2469172932330825, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 58980 + }, + { + "epoch": 3.2469774436090226, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1132, + "step": 58981 + }, + { + "epoch": 3.2470375939849623, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 58982 + }, + { + "epoch": 3.2470977443609024, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 58983 + }, + { + "epoch": 3.247157894736842, + "grad_norm": 1.578125, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 58984 + }, + { + "epoch": 3.247218045112782, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 58985 + }, + { + "epoch": 3.2472781954887218, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 58986 + }, + { + "epoch": 3.247338345864662, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 58987 + }, + { + "epoch": 3.2473984962406015, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 58988 + }, + { + "epoch": 3.247458646616541, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 58989 + }, + { + "epoch": 3.2475187969924812, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 58990 + }, + { + "epoch": 3.247578947368421, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 58991 + }, + { + "epoch": 3.247639097744361, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9029, + "step": 58992 + }, + { + "epoch": 3.2476992481203006, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 58993 + }, + { + "epoch": 3.2477593984962407, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 58994 + }, + { + "epoch": 3.2478195488721804, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 58995 + }, + { + "epoch": 3.2478796992481205, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9149, + "step": 58996 + }, + { + "epoch": 3.24793984962406, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 58997 + }, + { + "epoch": 3.248, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 58998 + }, + { + "epoch": 3.24806015037594, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 58999 + }, + { + "epoch": 3.2481203007518795, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 59000 + }, + { + "epoch": 3.2481804511278196, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 59001 + }, + { + "epoch": 3.2482406015037593, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 59002 + }, + { + "epoch": 3.2483007518796994, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 59003 + }, + { + "epoch": 3.248360902255639, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 59004 + }, + { + "epoch": 3.248421052631579, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 59005 + }, + { + "epoch": 3.2484812030075187, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 59006 + }, + { + "epoch": 3.248541353383459, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 59007 + }, + { + "epoch": 3.2486015037593985, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0752, + "step": 59008 + }, + { + "epoch": 3.248661654135338, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 59009 + }, + { + "epoch": 3.2487218045112782, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 59010 + }, + { + "epoch": 3.248781954887218, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 59011 + }, + { + "epoch": 3.248842105263158, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 59012 + }, + { + "epoch": 3.2489022556390976, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 59013 + }, + { + "epoch": 3.2489624060150377, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 59014 + }, + { + "epoch": 3.2490225563909774, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1532, + "step": 59015 + }, + { + "epoch": 3.2490827067669175, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9162, + "step": 59016 + }, + { + "epoch": 3.249142857142857, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.1455, + "step": 59017 + }, + { + "epoch": 3.2492030075187968, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 59018 + }, + { + "epoch": 3.249263157894737, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 59019 + }, + { + "epoch": 3.2493233082706765, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 59020 + }, + { + "epoch": 3.2493834586466166, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 59021 + }, + { + "epoch": 3.2494436090225562, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 59022 + }, + { + "epoch": 3.2495037593984963, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 59023 + }, + { + "epoch": 3.249563909774436, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 59024 + }, + { + "epoch": 3.249624060150376, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9315, + "step": 59025 + }, + { + "epoch": 3.2496842105263157, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 59026 + }, + { + "epoch": 3.249744360902256, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0991, + "step": 59027 + }, + { + "epoch": 3.2498045112781955, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 59028 + }, + { + "epoch": 3.2498646616541356, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 59029 + }, + { + "epoch": 3.249924812030075, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 59030 + }, + { + "epoch": 3.249984962406015, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 59031 + }, + { + "epoch": 3.250045112781955, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 59032 + }, + { + "epoch": 3.2501052631578946, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 59033 + }, + { + "epoch": 3.2501654135338347, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 59034 + }, + { + "epoch": 3.2502255639097744, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 59035 + }, + { + "epoch": 3.2502857142857144, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 59036 + }, + { + "epoch": 3.250345864661654, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 59037 + }, + { + "epoch": 3.250406015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 59038 + }, + { + "epoch": 3.250466165413534, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9237, + "step": 59039 + }, + { + "epoch": 3.2505263157894735, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 59040 + }, + { + "epoch": 3.2505864661654136, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0945, + "step": 59041 + }, + { + "epoch": 3.2506466165413532, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 59042 + }, + { + "epoch": 3.2507067669172933, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9331, + "step": 59043 + }, + { + "epoch": 3.250766917293233, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 59044 + }, + { + "epoch": 3.250827067669173, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.114, + "step": 59045 + }, + { + "epoch": 3.2508872180451127, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 59046 + }, + { + "epoch": 3.250947368421053, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.929, + "step": 59047 + }, + { + "epoch": 3.2510075187969925, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 59048 + }, + { + "epoch": 3.251067669172932, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 59049 + }, + { + "epoch": 3.251127819548872, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 59050 + }, + { + "epoch": 3.251187969924812, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 59051 + }, + { + "epoch": 3.251248120300752, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1152, + "step": 59052 + }, + { + "epoch": 3.2513082706766916, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 59053 + }, + { + "epoch": 3.2513684210526317, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 59054 + }, + { + "epoch": 3.2514285714285713, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0663, + "step": 59055 + }, + { + "epoch": 3.2514887218045114, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 59056 + }, + { + "epoch": 3.251548872180451, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 59057 + }, + { + "epoch": 3.251609022556391, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 59058 + }, + { + "epoch": 3.251669172932331, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 59059 + }, + { + "epoch": 3.251729323308271, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0438, + "step": 59060 + }, + { + "epoch": 3.2517894736842106, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 59061 + }, + { + "epoch": 3.25184962406015, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 59062 + }, + { + "epoch": 3.2519097744360903, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 59063 + }, + { + "epoch": 3.25196992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 59064 + }, + { + "epoch": 3.25203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 59065 + }, + { + "epoch": 3.2520902255639097, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 59066 + }, + { + "epoch": 3.25215037593985, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 59067 + }, + { + "epoch": 3.2522105263157894, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 59068 + }, + { + "epoch": 3.2522706766917295, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 59069 + }, + { + "epoch": 3.252330827067669, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 59070 + }, + { + "epoch": 3.252390977443609, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 59071 + }, + { + "epoch": 3.252451127819549, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 59072 + }, + { + "epoch": 3.2525112781954886, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 59073 + }, + { + "epoch": 3.2525714285714287, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9434, + "step": 59074 + }, + { + "epoch": 3.2526315789473683, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 59075 + }, + { + "epoch": 3.2526917293233084, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 59076 + }, + { + "epoch": 3.252751879699248, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9356, + "step": 59077 + }, + { + "epoch": 3.252812030075188, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9383, + "step": 59078 + }, + { + "epoch": 3.252872180451128, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0727, + "step": 59079 + }, + { + "epoch": 3.2529323308270675, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.932, + "step": 59080 + }, + { + "epoch": 3.2529924812030075, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 59081 + }, + { + "epoch": 3.253052631578947, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.912, + "step": 59082 + }, + { + "epoch": 3.2531127819548873, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 59083 + }, + { + "epoch": 3.253172932330827, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 59084 + }, + { + "epoch": 3.253233082706767, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 59085 + }, + { + "epoch": 3.2532932330827067, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 59086 + }, + { + "epoch": 3.2533533834586468, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 59087 + }, + { + "epoch": 3.2534135338345864, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 59088 + }, + { + "epoch": 3.253473684210526, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1386, + "step": 59089 + }, + { + "epoch": 3.253533834586466, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9109, + "step": 59090 + }, + { + "epoch": 3.2535939849624063, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 59091 + }, + { + "epoch": 3.253654135338346, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 59092 + }, + { + "epoch": 3.2537142857142856, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 59093 + }, + { + "epoch": 3.2537744360902257, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 59094 + }, + { + "epoch": 3.2538345864661653, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 59095 + }, + { + "epoch": 3.2538947368421054, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9354, + "step": 59096 + }, + { + "epoch": 3.253954887218045, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.8688, + "step": 59097 + }, + { + "epoch": 3.254015037593985, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 59098 + }, + { + "epoch": 3.254075187969925, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0819, + "step": 59099 + }, + { + "epoch": 3.254135338345865, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 59100 + }, + { + "epoch": 3.2541954887218045, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 59101 + }, + { + "epoch": 3.254255639097744, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 59102 + }, + { + "epoch": 3.2543157894736843, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 59103 + }, + { + "epoch": 3.254375939849624, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 59104 + }, + { + "epoch": 3.254436090225564, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 59105 + }, + { + "epoch": 3.2544962406015037, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 59106 + }, + { + "epoch": 3.2545563909774438, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0943, + "step": 59107 + }, + { + "epoch": 3.2546165413533834, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 59108 + }, + { + "epoch": 3.2546766917293235, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9228, + "step": 59109 + }, + { + "epoch": 3.254736842105263, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 59110 + }, + { + "epoch": 3.254796992481203, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0342, + "step": 59111 + }, + { + "epoch": 3.254857142857143, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 59112 + }, + { + "epoch": 3.2549172932330825, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 59113 + }, + { + "epoch": 3.2549774436090226, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 59114 + }, + { + "epoch": 3.2550375939849623, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 59115 + }, + { + "epoch": 3.2550977443609024, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9586, + "step": 59116 + }, + { + "epoch": 3.255157894736842, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 59117 + }, + { + "epoch": 3.255218045112782, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9105, + "step": 59118 + }, + { + "epoch": 3.2552781954887218, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 59119 + }, + { + "epoch": 3.2553383458646614, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 59120 + }, + { + "epoch": 3.2553984962406015, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 59121 + }, + { + "epoch": 3.255458646616541, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 59122 + }, + { + "epoch": 3.2555187969924813, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 59123 + }, + { + "epoch": 3.255578947368421, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 59124 + }, + { + "epoch": 3.255639097744361, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.8704, + "step": 59125 + }, + { + "epoch": 3.2556992481203006, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 59126 + }, + { + "epoch": 3.2557593984962407, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 59127 + }, + { + "epoch": 3.2558195488721804, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 59128 + }, + { + "epoch": 3.2558796992481205, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 59129 + }, + { + "epoch": 3.25593984962406, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 59130 + }, + { + "epoch": 3.2560000000000002, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9271, + "step": 59131 + }, + { + "epoch": 3.25606015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.1362, + "step": 59132 + }, + { + "epoch": 3.2561203007518795, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9288, + "step": 59133 + }, + { + "epoch": 3.2561804511278196, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 59134 + }, + { + "epoch": 3.2562406015037593, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 59135 + }, + { + "epoch": 3.2563007518796994, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 59136 + }, + { + "epoch": 3.256360902255639, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 59137 + }, + { + "epoch": 3.256421052631579, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0988, + "step": 59138 + }, + { + "epoch": 3.2564812030075188, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 59139 + }, + { + "epoch": 3.256541353383459, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0527, + "step": 59140 + }, + { + "epoch": 3.2566015037593985, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 59141 + }, + { + "epoch": 3.256661654135338, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 59142 + }, + { + "epoch": 3.2567218045112782, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 59143 + }, + { + "epoch": 3.256781954887218, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1451, + "step": 59144 + }, + { + "epoch": 3.256842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 59145 + }, + { + "epoch": 3.2569022556390976, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 59146 + }, + { + "epoch": 3.2569624060150377, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 59147 + }, + { + "epoch": 3.2570225563909774, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 59148 + }, + { + "epoch": 3.2570827067669175, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 59149 + }, + { + "epoch": 3.257142857142857, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 59150 + }, + { + "epoch": 3.2572030075187968, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 59151 + }, + { + "epoch": 3.257263157894737, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.922, + "step": 59152 + }, + { + "epoch": 3.2573233082706765, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 59153 + }, + { + "epoch": 3.2573834586466166, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9062, + "step": 59154 + }, + { + "epoch": 3.2574436090225563, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 59155 + }, + { + "epoch": 3.2575037593984963, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 59156 + }, + { + "epoch": 3.257563909774436, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 59157 + }, + { + "epoch": 3.257624060150376, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 59158 + }, + { + "epoch": 3.2576842105263157, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 59159 + }, + { + "epoch": 3.257744360902256, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 59160 + }, + { + "epoch": 3.2578045112781955, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0505, + "step": 59161 + }, + { + "epoch": 3.2578646616541356, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 59162 + }, + { + "epoch": 3.2579248120300752, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 0.9239, + "step": 59163 + }, + { + "epoch": 3.257984962406015, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 59164 + }, + { + "epoch": 3.258045112781955, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 59165 + }, + { + "epoch": 3.2581052631578946, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 59166 + }, + { + "epoch": 3.2581654135338347, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 59167 + }, + { + "epoch": 3.2582255639097744, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 59168 + }, + { + "epoch": 3.2582857142857145, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 59169 + }, + { + "epoch": 3.258345864661654, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 59170 + }, + { + "epoch": 3.258406015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 59171 + }, + { + "epoch": 3.258466165413534, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 59172 + }, + { + "epoch": 3.2585263157894735, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 59173 + }, + { + "epoch": 3.2585864661654136, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 59174 + }, + { + "epoch": 3.2586466165413532, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.915, + "step": 59175 + }, + { + "epoch": 3.2587067669172933, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 59176 + }, + { + "epoch": 3.258766917293233, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.931, + "step": 59177 + }, + { + "epoch": 3.258827067669173, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.8839, + "step": 59178 + }, + { + "epoch": 3.2588872180451127, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 59179 + }, + { + "epoch": 3.258947368421053, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 59180 + }, + { + "epoch": 3.2590075187969925, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.8898, + "step": 59181 + }, + { + "epoch": 3.259067669172932, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 59182 + }, + { + "epoch": 3.259127819548872, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 59183 + }, + { + "epoch": 3.259187969924812, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 59184 + }, + { + "epoch": 3.259248120300752, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 59185 + }, + { + "epoch": 3.2593082706766916, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 59186 + }, + { + "epoch": 3.2593684210526317, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 59187 + }, + { + "epoch": 3.2594285714285713, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 59188 + }, + { + "epoch": 3.2594887218045114, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9267, + "step": 59189 + }, + { + "epoch": 3.259548872180451, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 59190 + }, + { + "epoch": 3.2596090225563907, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 59191 + }, + { + "epoch": 3.259669172932331, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 59192 + }, + { + "epoch": 3.259729323308271, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 59193 + }, + { + "epoch": 3.2597894736842106, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 59194 + }, + { + "epoch": 3.25984962406015, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 59195 + }, + { + "epoch": 3.2599097744360903, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 59196 + }, + { + "epoch": 3.25996992481203, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 59197 + }, + { + "epoch": 3.26003007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 59198 + }, + { + "epoch": 3.2600902255639097, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 59199 + }, + { + "epoch": 3.26015037593985, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0018, + "step": 59200 + }, + { + "epoch": 3.2602105263157894, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 59201 + }, + { + "epoch": 3.2602706766917295, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 59202 + }, + { + "epoch": 3.260330827067669, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 59203 + }, + { + "epoch": 3.260390977443609, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 59204 + }, + { + "epoch": 3.260451127819549, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9226, + "step": 59205 + }, + { + "epoch": 3.2605112781954886, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 59206 + }, + { + "epoch": 3.2605714285714287, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 59207 + }, + { + "epoch": 3.2606315789473683, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 59208 + }, + { + "epoch": 3.2606917293233084, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1301, + "step": 59209 + }, + { + "epoch": 3.260751879699248, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.997, + "step": 59210 + }, + { + "epoch": 3.260812030075188, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 59211 + }, + { + "epoch": 3.260872180451128, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 59212 + }, + { + "epoch": 3.2609323308270675, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 59213 + }, + { + "epoch": 3.2609924812030076, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 59214 + }, + { + "epoch": 3.261052631578947, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 59215 + }, + { + "epoch": 3.2611127819548873, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 59216 + }, + { + "epoch": 3.261172932330827, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 59217 + }, + { + "epoch": 3.261233082706767, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 59218 + }, + { + "epoch": 3.2612932330827067, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 59219 + }, + { + "epoch": 3.261353383458647, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 59220 + }, + { + "epoch": 3.2614135338345864, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 59221 + }, + { + "epoch": 3.261473684210526, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 59222 + }, + { + "epoch": 3.261533834586466, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 59223 + }, + { + "epoch": 3.261593984962406, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 59224 + }, + { + "epoch": 3.261654135338346, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 59225 + }, + { + "epoch": 3.2617142857142856, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 59226 + }, + { + "epoch": 3.2617744360902257, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9311, + "step": 59227 + }, + { + "epoch": 3.2618345864661653, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 59228 + }, + { + "epoch": 3.2618947368421054, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1318, + "step": 59229 + }, + { + "epoch": 3.261954887218045, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 59230 + }, + { + "epoch": 3.262015037593985, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 59231 + }, + { + "epoch": 3.262075187969925, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 59232 + }, + { + "epoch": 3.262135338345865, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 59233 + }, + { + "epoch": 3.2621954887218045, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 59234 + }, + { + "epoch": 3.262255639097744, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 59235 + }, + { + "epoch": 3.2623157894736843, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 59236 + }, + { + "epoch": 3.262375939849624, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 59237 + }, + { + "epoch": 3.262436090225564, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.8963, + "step": 59238 + }, + { + "epoch": 3.2624962406015037, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 59239 + }, + { + "epoch": 3.2625563909774438, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 59240 + }, + { + "epoch": 3.2626165413533834, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 59241 + }, + { + "epoch": 3.2626766917293235, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 59242 + }, + { + "epoch": 3.262736842105263, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 59243 + }, + { + "epoch": 3.262796992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 59244 + }, + { + "epoch": 3.262857142857143, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9072, + "step": 59245 + }, + { + "epoch": 3.2629172932330825, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9344, + "step": 59246 + }, + { + "epoch": 3.2629774436090226, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9456, + "step": 59247 + }, + { + "epoch": 3.2630375939849623, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 59248 + }, + { + "epoch": 3.2630977443609024, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 59249 + }, + { + "epoch": 3.263157894736842, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9005, + "step": 59250 + }, + { + "epoch": 3.263218045112782, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 59251 + }, + { + "epoch": 3.2632781954887218, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 59252 + }, + { + "epoch": 3.2633383458646614, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0619, + "step": 59253 + }, + { + "epoch": 3.2633984962406015, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 59254 + }, + { + "epoch": 3.263458646616541, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 59255 + }, + { + "epoch": 3.2635187969924813, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 59256 + }, + { + "epoch": 3.263578947368421, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0857, + "step": 59257 + }, + { + "epoch": 3.263639097744361, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 59258 + }, + { + "epoch": 3.2636992481203007, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 59259 + }, + { + "epoch": 3.2637593984962407, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 59260 + }, + { + "epoch": 3.2638195488721804, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 59261 + }, + { + "epoch": 3.2638796992481205, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 59262 + }, + { + "epoch": 3.26393984962406, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 59263 + }, + { + "epoch": 3.2640000000000002, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 59264 + }, + { + "epoch": 3.26406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 59265 + }, + { + "epoch": 3.2641203007518795, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 59266 + }, + { + "epoch": 3.2641804511278196, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 59267 + }, + { + "epoch": 3.2642406015037593, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9417, + "step": 59268 + }, + { + "epoch": 3.2643007518796994, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 59269 + }, + { + "epoch": 3.264360902255639, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 59270 + }, + { + "epoch": 3.264421052631579, + "grad_norm": 1.3203125, + "learning_rate": 3e-05, + "loss": 1.184, + "step": 59271 + }, + { + "epoch": 3.2644812030075188, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 59272 + }, + { + "epoch": 3.264541353383459, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9306, + "step": 59273 + }, + { + "epoch": 3.2646015037593985, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 59274 + }, + { + "epoch": 3.264661654135338, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 59275 + }, + { + "epoch": 3.2647218045112782, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 59276 + }, + { + "epoch": 3.264781954887218, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 59277 + }, + { + "epoch": 3.264842105263158, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 59278 + }, + { + "epoch": 3.2649022556390976, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.8857, + "step": 59279 + }, + { + "epoch": 3.2649624060150377, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 59280 + }, + { + "epoch": 3.2650225563909774, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 59281 + }, + { + "epoch": 3.2650827067669175, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9151, + "step": 59282 + }, + { + "epoch": 3.265142857142857, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0461, + "step": 59283 + }, + { + "epoch": 3.2652030075187968, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 59284 + }, + { + "epoch": 3.265263157894737, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.912, + "step": 59285 + }, + { + "epoch": 3.2653233082706765, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9252, + "step": 59286 + }, + { + "epoch": 3.2653834586466166, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 59287 + }, + { + "epoch": 3.2654436090225563, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 59288 + }, + { + "epoch": 3.2655037593984964, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9311, + "step": 59289 + }, + { + "epoch": 3.265563909774436, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 59290 + }, + { + "epoch": 3.265624060150376, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 59291 + }, + { + "epoch": 3.2656842105263157, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9373, + "step": 59292 + }, + { + "epoch": 3.2657443609022554, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.8964, + "step": 59293 + }, + { + "epoch": 3.2658045112781955, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 59294 + }, + { + "epoch": 3.2658646616541356, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 59295 + }, + { + "epoch": 3.2659248120300752, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 59296 + }, + { + "epoch": 3.265984962406015, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0354, + "step": 59297 + }, + { + "epoch": 3.266045112781955, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.931, + "step": 59298 + }, + { + "epoch": 3.2661052631578946, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0803, + "step": 59299 + }, + { + "epoch": 3.2661654135338347, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 59300 + }, + { + "epoch": 3.2662255639097744, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 59301 + }, + { + "epoch": 3.2662857142857145, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 59302 + }, + { + "epoch": 3.266345864661654, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 59303 + }, + { + "epoch": 3.266406015037594, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 59304 + }, + { + "epoch": 3.266466165413534, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 59305 + }, + { + "epoch": 3.2665263157894735, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9211, + "step": 59306 + }, + { + "epoch": 3.2665864661654136, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 59307 + }, + { + "epoch": 3.2666466165413532, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.088, + "step": 59308 + }, + { + "epoch": 3.2667067669172933, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 59309 + }, + { + "epoch": 3.266766917293233, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 59310 + }, + { + "epoch": 3.266827067669173, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9439, + "step": 59311 + }, + { + "epoch": 3.2668872180451127, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 59312 + }, + { + "epoch": 3.266947368421053, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 59313 + }, + { + "epoch": 3.2670075187969925, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 59314 + }, + { + "epoch": 3.267067669172932, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.932, + "step": 59315 + }, + { + "epoch": 3.267127819548872, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 59316 + }, + { + "epoch": 3.267187969924812, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 59317 + }, + { + "epoch": 3.267248120300752, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1503, + "step": 59318 + }, + { + "epoch": 3.2673082706766916, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 59319 + }, + { + "epoch": 3.2673684210526317, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 59320 + }, + { + "epoch": 3.2674285714285713, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 59321 + }, + { + "epoch": 3.2674887218045114, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1733, + "step": 59322 + }, + { + "epoch": 3.267548872180451, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 59323 + }, + { + "epoch": 3.2676090225563907, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 59324 + }, + { + "epoch": 3.267669172932331, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 59325 + }, + { + "epoch": 3.2677293233082705, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 59326 + }, + { + "epoch": 3.2677894736842106, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 59327 + }, + { + "epoch": 3.2678496240601502, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 59328 + }, + { + "epoch": 3.2679097744360903, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 59329 + }, + { + "epoch": 3.26796992481203, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 59330 + }, + { + "epoch": 3.26803007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 59331 + }, + { + "epoch": 3.2680902255639097, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 59332 + }, + { + "epoch": 3.26815037593985, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 59333 + }, + { + "epoch": 3.2682105263157895, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0324, + "step": 59334 + }, + { + "epoch": 3.2682706766917295, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 59335 + }, + { + "epoch": 3.268330827067669, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1774, + "step": 59336 + }, + { + "epoch": 3.268390977443609, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 59337 + }, + { + "epoch": 3.268451127819549, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 59338 + }, + { + "epoch": 3.2685112781954886, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1137, + "step": 59339 + }, + { + "epoch": 3.2685714285714287, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 59340 + }, + { + "epoch": 3.2686315789473683, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 59341 + }, + { + "epoch": 3.2686917293233084, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 59342 + }, + { + "epoch": 3.268751879699248, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 59343 + }, + { + "epoch": 3.268812030075188, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 59344 + }, + { + "epoch": 3.268872180451128, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 59345 + }, + { + "epoch": 3.2689323308270675, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 59346 + }, + { + "epoch": 3.2689924812030076, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0775, + "step": 59347 + }, + { + "epoch": 3.269052631578947, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 59348 + }, + { + "epoch": 3.2691127819548873, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 59349 + }, + { + "epoch": 3.269172932330827, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.1193, + "step": 59350 + }, + { + "epoch": 3.269233082706767, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 59351 + }, + { + "epoch": 3.2692932330827067, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 59352 + }, + { + "epoch": 3.269353383458647, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 59353 + }, + { + "epoch": 3.2694135338345864, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 59354 + }, + { + "epoch": 3.269473684210526, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 59355 + }, + { + "epoch": 3.269533834586466, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 59356 + }, + { + "epoch": 3.269593984962406, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 59357 + }, + { + "epoch": 3.269654135338346, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 59358 + }, + { + "epoch": 3.2697142857142856, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 59359 + }, + { + "epoch": 3.2697744360902257, + "grad_norm": 0.494140625, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 59360 + }, + { + "epoch": 3.2698345864661653, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 59361 + }, + { + "epoch": 3.2698947368421054, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0605, + "step": 59362 + }, + { + "epoch": 3.269954887218045, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 59363 + }, + { + "epoch": 3.270015037593985, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 59364 + }, + { + "epoch": 3.270075187969925, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9358, + "step": 59365 + }, + { + "epoch": 3.270135338345865, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 59366 + }, + { + "epoch": 3.2701954887218045, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 59367 + }, + { + "epoch": 3.270255639097744, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 59368 + }, + { + "epoch": 3.2703157894736843, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 59369 + }, + { + "epoch": 3.270375939849624, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 59370 + }, + { + "epoch": 3.270436090225564, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 59371 + }, + { + "epoch": 3.2704962406015037, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 59372 + }, + { + "epoch": 3.2705563909774438, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 59373 + }, + { + "epoch": 3.2706165413533834, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9134, + "step": 59374 + }, + { + "epoch": 3.2706766917293235, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 59375 + }, + { + "epoch": 3.270736842105263, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 59376 + }, + { + "epoch": 3.270796992481203, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 59377 + }, + { + "epoch": 3.270857142857143, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 59378 + }, + { + "epoch": 3.2709172932330826, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 59379 + }, + { + "epoch": 3.2709774436090227, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 59380 + }, + { + "epoch": 3.2710375939849623, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 59381 + }, + { + "epoch": 3.2710977443609024, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0311, + "step": 59382 + }, + { + "epoch": 3.271157894736842, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 59383 + }, + { + "epoch": 3.271218045112782, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1697, + "step": 59384 + }, + { + "epoch": 3.271278195488722, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 59385 + }, + { + "epoch": 3.2713383458646614, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.089, + "step": 59386 + }, + { + "epoch": 3.2713984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 59387 + }, + { + "epoch": 3.271458646616541, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 59388 + }, + { + "epoch": 3.2715187969924813, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 59389 + }, + { + "epoch": 3.271578947368421, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 59390 + }, + { + "epoch": 3.271639097744361, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 59391 + }, + { + "epoch": 3.2716992481203007, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 59392 + }, + { + "epoch": 3.2717593984962408, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 59393 + }, + { + "epoch": 3.2718195488721804, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 59394 + }, + { + "epoch": 3.27187969924812, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9243, + "step": 59395 + }, + { + "epoch": 3.27193984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 59396 + }, + { + "epoch": 3.2720000000000002, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 59397 + }, + { + "epoch": 3.27206015037594, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 59398 + }, + { + "epoch": 3.2721203007518795, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.8951, + "step": 59399 + }, + { + "epoch": 3.2721804511278196, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 59400 + }, + { + "epoch": 3.2722406015037593, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 59401 + }, + { + "epoch": 3.2723007518796994, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 59402 + }, + { + "epoch": 3.272360902255639, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 59403 + }, + { + "epoch": 3.272421052631579, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 59404 + }, + { + "epoch": 3.2724812030075188, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 59405 + }, + { + "epoch": 3.272541353383459, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 59406 + }, + { + "epoch": 3.2726015037593985, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9416, + "step": 59407 + }, + { + "epoch": 3.272661654135338, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 59408 + }, + { + "epoch": 3.2727218045112783, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 59409 + }, + { + "epoch": 3.272781954887218, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 59410 + }, + { + "epoch": 3.272842105263158, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0358, + "step": 59411 + }, + { + "epoch": 3.2729022556390976, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0973, + "step": 59412 + }, + { + "epoch": 3.2729624060150377, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 59413 + }, + { + "epoch": 3.2730225563909774, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 59414 + }, + { + "epoch": 3.2730827067669175, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9278, + "step": 59415 + }, + { + "epoch": 3.273142857142857, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 59416 + }, + { + "epoch": 3.273203007518797, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.133, + "step": 59417 + }, + { + "epoch": 3.273263157894737, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 59418 + }, + { + "epoch": 3.2733233082706765, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.8955, + "step": 59419 + }, + { + "epoch": 3.2733834586466166, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9308, + "step": 59420 + }, + { + "epoch": 3.2734436090225563, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 59421 + }, + { + "epoch": 3.2735037593984964, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 59422 + }, + { + "epoch": 3.273563909774436, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 59423 + }, + { + "epoch": 3.273624060150376, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 59424 + }, + { + "epoch": 3.2736842105263158, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 59425 + }, + { + "epoch": 3.2737443609022554, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9364, + "step": 59426 + }, + { + "epoch": 3.2738045112781955, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 59427 + }, + { + "epoch": 3.273864661654135, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 59428 + }, + { + "epoch": 3.2739248120300752, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.8946, + "step": 59429 + }, + { + "epoch": 3.273984962406015, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 59430 + }, + { + "epoch": 3.274045112781955, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 59431 + }, + { + "epoch": 3.2741052631578946, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 59432 + }, + { + "epoch": 3.2741654135338347, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 59433 + }, + { + "epoch": 3.2742255639097744, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 59434 + }, + { + "epoch": 3.2742857142857145, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 59435 + }, + { + "epoch": 3.274345864661654, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 59436 + }, + { + "epoch": 3.274406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0754, + "step": 59437 + }, + { + "epoch": 3.274466165413534, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 59438 + }, + { + "epoch": 3.2745263157894735, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9373, + "step": 59439 + }, + { + "epoch": 3.2745864661654136, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 59440 + }, + { + "epoch": 3.2746466165413533, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 59441 + }, + { + "epoch": 3.2747067669172933, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 59442 + }, + { + "epoch": 3.274766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9169, + "step": 59443 + }, + { + "epoch": 3.274827067669173, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.8903, + "step": 59444 + }, + { + "epoch": 3.2748872180451127, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1398, + "step": 59445 + }, + { + "epoch": 3.274947368421053, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 59446 + }, + { + "epoch": 3.2750075187969925, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 59447 + }, + { + "epoch": 3.275067669172932, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 59448 + }, + { + "epoch": 3.275127819548872, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 59449 + }, + { + "epoch": 3.275187969924812, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9288, + "step": 59450 + }, + { + "epoch": 3.275248120300752, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 59451 + }, + { + "epoch": 3.2753082706766916, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 59452 + }, + { + "epoch": 3.2753684210526317, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 59453 + }, + { + "epoch": 3.2754285714285714, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 59454 + }, + { + "epoch": 3.2754887218045114, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 59455 + }, + { + "epoch": 3.275548872180451, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9311, + "step": 59456 + }, + { + "epoch": 3.2756090225563907, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9184, + "step": 59457 + }, + { + "epoch": 3.275669172932331, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 59458 + }, + { + "epoch": 3.2757293233082705, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9255, + "step": 59459 + }, + { + "epoch": 3.2757894736842106, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9412, + "step": 59460 + }, + { + "epoch": 3.2758496240601502, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 59461 + }, + { + "epoch": 3.2759097744360903, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 59462 + }, + { + "epoch": 3.27596992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9181, + "step": 59463 + }, + { + "epoch": 3.27603007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0496, + "step": 59464 + }, + { + "epoch": 3.2760902255639097, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 59465 + }, + { + "epoch": 3.27615037593985, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 59466 + }, + { + "epoch": 3.2762105263157895, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 59467 + }, + { + "epoch": 3.2762706766917296, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 59468 + }, + { + "epoch": 3.276330827067669, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 59469 + }, + { + "epoch": 3.276390977443609, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 59470 + }, + { + "epoch": 3.276451127819549, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 59471 + }, + { + "epoch": 3.2765112781954886, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9156, + "step": 59472 + }, + { + "epoch": 3.2765714285714287, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 59473 + }, + { + "epoch": 3.2766315789473683, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1277, + "step": 59474 + }, + { + "epoch": 3.2766917293233084, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 59475 + }, + { + "epoch": 3.276751879699248, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 59476 + }, + { + "epoch": 3.276812030075188, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 59477 + }, + { + "epoch": 3.276872180451128, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 59478 + }, + { + "epoch": 3.2769323308270675, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 59479 + }, + { + "epoch": 3.2769924812030076, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 59480 + }, + { + "epoch": 3.277052631578947, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 59481 + }, + { + "epoch": 3.2771127819548873, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 59482 + }, + { + "epoch": 3.277172932330827, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 59483 + }, + { + "epoch": 3.277233082706767, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 59484 + }, + { + "epoch": 3.2772932330827067, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1538, + "step": 59485 + }, + { + "epoch": 3.277353383458647, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.8714, + "step": 59486 + }, + { + "epoch": 3.2774135338345864, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 59487 + }, + { + "epoch": 3.277473684210526, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 59488 + }, + { + "epoch": 3.277533834586466, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 59489 + }, + { + "epoch": 3.277593984962406, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 59490 + }, + { + "epoch": 3.277654135338346, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 59491 + }, + { + "epoch": 3.2777142857142856, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 59492 + }, + { + "epoch": 3.2777744360902257, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 59493 + }, + { + "epoch": 3.2778345864661653, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 59494 + }, + { + "epoch": 3.2778947368421054, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 59495 + }, + { + "epoch": 3.277954887218045, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9317, + "step": 59496 + }, + { + "epoch": 3.2780150375939847, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 59497 + }, + { + "epoch": 3.278075187969925, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 59498 + }, + { + "epoch": 3.278135338345865, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0864, + "step": 59499 + }, + { + "epoch": 3.2781954887218046, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 59500 + }, + { + "epoch": 3.278255639097744, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 59501 + }, + { + "epoch": 3.2783157894736843, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 59502 + }, + { + "epoch": 3.278375939849624, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 59503 + }, + { + "epoch": 3.278436090225564, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 59504 + }, + { + "epoch": 3.2784962406015037, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 59505 + }, + { + "epoch": 3.278556390977444, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0769, + "step": 59506 + }, + { + "epoch": 3.2786165413533834, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1066, + "step": 59507 + }, + { + "epoch": 3.2786766917293235, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 59508 + }, + { + "epoch": 3.278736842105263, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 59509 + }, + { + "epoch": 3.278796992481203, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 59510 + }, + { + "epoch": 3.278857142857143, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 59511 + }, + { + "epoch": 3.2789172932330826, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9757, + "step": 59512 + }, + { + "epoch": 3.2789774436090227, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 59513 + }, + { + "epoch": 3.2790375939849623, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 59514 + }, + { + "epoch": 3.2790977443609024, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 59515 + }, + { + "epoch": 3.279157894736842, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9055, + "step": 59516 + }, + { + "epoch": 3.279218045112782, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0904, + "step": 59517 + }, + { + "epoch": 3.279278195488722, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 59518 + }, + { + "epoch": 3.2793383458646614, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 59519 + }, + { + "epoch": 3.2793984962406015, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 59520 + }, + { + "epoch": 3.279458646616541, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 59521 + }, + { + "epoch": 3.2795187969924813, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 59522 + }, + { + "epoch": 3.279578947368421, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 59523 + }, + { + "epoch": 3.279639097744361, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 59524 + }, + { + "epoch": 3.2796992481203007, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 59525 + }, + { + "epoch": 3.2797593984962408, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 59526 + }, + { + "epoch": 3.2798195488721804, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 59527 + }, + { + "epoch": 3.27987969924812, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 59528 + }, + { + "epoch": 3.27993984962406, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1304, + "step": 59529 + }, + { + "epoch": 3.2800000000000002, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 59530 + }, + { + "epoch": 3.28006015037594, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 59531 + }, + { + "epoch": 3.2801203007518795, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 59532 + }, + { + "epoch": 3.2801804511278196, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 59533 + }, + { + "epoch": 3.2802406015037593, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 59534 + }, + { + "epoch": 3.2803007518796994, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0875, + "step": 59535 + }, + { + "epoch": 3.280360902255639, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 59536 + }, + { + "epoch": 3.280421052631579, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 59537 + }, + { + "epoch": 3.2804812030075188, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 59538 + }, + { + "epoch": 3.280541353383459, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 59539 + }, + { + "epoch": 3.2806015037593985, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 59540 + }, + { + "epoch": 3.280661654135338, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 59541 + }, + { + "epoch": 3.2807218045112783, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 59542 + }, + { + "epoch": 3.280781954887218, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 59543 + }, + { + "epoch": 3.280842105263158, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 59544 + }, + { + "epoch": 3.2809022556390977, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 59545 + }, + { + "epoch": 3.2809624060150377, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9351, + "step": 59546 + }, + { + "epoch": 3.2810225563909774, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 59547 + }, + { + "epoch": 3.2810827067669175, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 59548 + }, + { + "epoch": 3.281142857142857, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.091, + "step": 59549 + }, + { + "epoch": 3.281203007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 59550 + }, + { + "epoch": 3.281263157894737, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 59551 + }, + { + "epoch": 3.2813233082706765, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1363, + "step": 59552 + }, + { + "epoch": 3.2813834586466166, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 59553 + }, + { + "epoch": 3.2814436090225563, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 59554 + }, + { + "epoch": 3.2815037593984964, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 59555 + }, + { + "epoch": 3.281563909774436, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 59556 + }, + { + "epoch": 3.281624060150376, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 59557 + }, + { + "epoch": 3.2816842105263158, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 59558 + }, + { + "epoch": 3.2817443609022554, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 59559 + }, + { + "epoch": 3.2818045112781955, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9543, + "step": 59560 + }, + { + "epoch": 3.281864661654135, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.937, + "step": 59561 + }, + { + "epoch": 3.2819248120300752, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 59562 + }, + { + "epoch": 3.281984962406015, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 59563 + }, + { + "epoch": 3.282045112781955, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 59564 + }, + { + "epoch": 3.2821052631578946, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 59565 + }, + { + "epoch": 3.2821654135338347, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 59566 + }, + { + "epoch": 3.2822255639097744, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 59567 + }, + { + "epoch": 3.2822857142857145, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 59568 + }, + { + "epoch": 3.282345864661654, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 59569 + }, + { + "epoch": 3.282406015037594, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 59570 + }, + { + "epoch": 3.282466165413534, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 59571 + }, + { + "epoch": 3.2825263157894735, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 59572 + }, + { + "epoch": 3.2825864661654136, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 59573 + }, + { + "epoch": 3.2826466165413533, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 59574 + }, + { + "epoch": 3.2827067669172934, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 59575 + }, + { + "epoch": 3.282766917293233, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 59576 + }, + { + "epoch": 3.282827067669173, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9224, + "step": 59577 + }, + { + "epoch": 3.2828872180451127, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 59578 + }, + { + "epoch": 3.282947368421053, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 59579 + }, + { + "epoch": 3.2830075187969925, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 59580 + }, + { + "epoch": 3.283067669172932, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 59581 + }, + { + "epoch": 3.2831278195488722, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 59582 + }, + { + "epoch": 3.283187969924812, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 59583 + }, + { + "epoch": 3.283248120300752, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 59584 + }, + { + "epoch": 3.2833082706766916, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.92, + "step": 59585 + }, + { + "epoch": 3.2833684210526317, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1708, + "step": 59586 + }, + { + "epoch": 3.2834285714285714, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 59587 + }, + { + "epoch": 3.2834887218045115, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.2096, + "step": 59588 + }, + { + "epoch": 3.283548872180451, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 59589 + }, + { + "epoch": 3.2836090225563908, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.91, + "step": 59590 + }, + { + "epoch": 3.283669172932331, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 59591 + }, + { + "epoch": 3.2837293233082705, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9088, + "step": 59592 + }, + { + "epoch": 3.2837894736842106, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 59593 + }, + { + "epoch": 3.2838496240601502, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.94, + "step": 59594 + }, + { + "epoch": 3.2839097744360903, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0678, + "step": 59595 + }, + { + "epoch": 3.28396992481203, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 59596 + }, + { + "epoch": 3.28403007518797, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 59597 + }, + { + "epoch": 3.2840902255639097, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 59598 + }, + { + "epoch": 3.2841503759398494, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 59599 + }, + { + "epoch": 3.2842105263157895, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 59600 + }, + { + "epoch": 3.2842706766917296, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 59601 + }, + { + "epoch": 3.284330827067669, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9542, + "step": 59602 + }, + { + "epoch": 3.284390977443609, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 59603 + }, + { + "epoch": 3.284451127819549, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 59604 + }, + { + "epoch": 3.2845112781954886, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 59605 + }, + { + "epoch": 3.2845714285714287, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 59606 + }, + { + "epoch": 3.2846315789473683, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 59607 + }, + { + "epoch": 3.2846917293233084, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 59608 + }, + { + "epoch": 3.284751879699248, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 59609 + }, + { + "epoch": 3.284812030075188, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 59610 + }, + { + "epoch": 3.284872180451128, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 59611 + }, + { + "epoch": 3.2849323308270675, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 59612 + }, + { + "epoch": 3.2849924812030076, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 59613 + }, + { + "epoch": 3.2850526315789472, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 59614 + }, + { + "epoch": 3.2851127819548873, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 59615 + }, + { + "epoch": 3.285172932330827, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 59616 + }, + { + "epoch": 3.285233082706767, + "grad_norm": 1.25, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 59617 + }, + { + "epoch": 3.2852932330827067, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 59618 + }, + { + "epoch": 3.285353383458647, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 59619 + }, + { + "epoch": 3.2854135338345865, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 59620 + }, + { + "epoch": 3.285473684210526, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 59621 + }, + { + "epoch": 3.285533834586466, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9709, + "step": 59622 + }, + { + "epoch": 3.285593984962406, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 59623 + }, + { + "epoch": 3.285654135338346, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9078, + "step": 59624 + }, + { + "epoch": 3.2857142857142856, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9529, + "step": 59625 + }, + { + "epoch": 3.2857744360902257, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 59626 + }, + { + "epoch": 3.2858345864661653, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 59627 + }, + { + "epoch": 3.2858947368421054, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 59628 + }, + { + "epoch": 3.285954887218045, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 59629 + }, + { + "epoch": 3.2860150375939847, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9288, + "step": 59630 + }, + { + "epoch": 3.286075187969925, + "grad_norm": 1.515625, + "learning_rate": 3e-05, + "loss": 1.0952, + "step": 59631 + }, + { + "epoch": 3.286135338345865, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 59632 + }, + { + "epoch": 3.2861954887218046, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 59633 + }, + { + "epoch": 3.286255639097744, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 59634 + }, + { + "epoch": 3.2863157894736843, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 59635 + }, + { + "epoch": 3.286375939849624, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 59636 + }, + { + "epoch": 3.286436090225564, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 59637 + }, + { + "epoch": 3.2864962406015037, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 59638 + }, + { + "epoch": 3.286556390977444, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 59639 + }, + { + "epoch": 3.2866165413533834, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 59640 + }, + { + "epoch": 3.2866766917293235, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.8939, + "step": 59641 + }, + { + "epoch": 3.286736842105263, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 59642 + }, + { + "epoch": 3.286796992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 59643 + }, + { + "epoch": 3.286857142857143, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 59644 + }, + { + "epoch": 3.2869172932330826, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 59645 + }, + { + "epoch": 3.2869774436090227, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9668, + "step": 59646 + }, + { + "epoch": 3.2870375939849623, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0994, + "step": 59647 + }, + { + "epoch": 3.2870977443609024, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9976, + "step": 59648 + }, + { + "epoch": 3.287157894736842, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 59649 + }, + { + "epoch": 3.287218045112782, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 59650 + }, + { + "epoch": 3.287278195488722, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 59651 + }, + { + "epoch": 3.2873383458646614, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 59652 + }, + { + "epoch": 3.2873984962406015, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 59653 + }, + { + "epoch": 3.287458646616541, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9698, + "step": 59654 + }, + { + "epoch": 3.2875187969924813, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 59655 + }, + { + "epoch": 3.287578947368421, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 59656 + }, + { + "epoch": 3.287639097744361, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 59657 + }, + { + "epoch": 3.2876992481203007, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9037, + "step": 59658 + }, + { + "epoch": 3.2877593984962408, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 59659 + }, + { + "epoch": 3.2878195488721804, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 59660 + }, + { + "epoch": 3.28787969924812, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 59661 + }, + { + "epoch": 3.28793984962406, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.89, + "step": 59662 + }, + { + "epoch": 3.288, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 59663 + }, + { + "epoch": 3.28806015037594, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 59664 + }, + { + "epoch": 3.2881203007518796, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 59665 + }, + { + "epoch": 3.2881804511278196, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 59666 + }, + { + "epoch": 3.2882406015037593, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 59667 + }, + { + "epoch": 3.2883007518796994, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 59668 + }, + { + "epoch": 3.288360902255639, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 59669 + }, + { + "epoch": 3.288421052631579, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 59670 + }, + { + "epoch": 3.288481203007519, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 59671 + }, + { + "epoch": 3.288541353383459, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 59672 + }, + { + "epoch": 3.2886015037593985, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 59673 + }, + { + "epoch": 3.288661654135338, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.1442, + "step": 59674 + }, + { + "epoch": 3.2887218045112783, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 59675 + }, + { + "epoch": 3.288781954887218, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9383, + "step": 59676 + }, + { + "epoch": 3.288842105263158, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 59677 + }, + { + "epoch": 3.2889022556390977, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 59678 + }, + { + "epoch": 3.2889624060150378, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.8597, + "step": 59679 + }, + { + "epoch": 3.2890225563909774, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 59680 + }, + { + "epoch": 3.2890827067669175, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 59681 + }, + { + "epoch": 3.289142857142857, + "grad_norm": 1.625, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 59682 + }, + { + "epoch": 3.289203007518797, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 59683 + }, + { + "epoch": 3.289263157894737, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 59684 + }, + { + "epoch": 3.2893233082706765, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1591, + "step": 59685 + }, + { + "epoch": 3.2893834586466166, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 59686 + }, + { + "epoch": 3.2894436090225563, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9226, + "step": 59687 + }, + { + "epoch": 3.2895037593984964, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 59688 + }, + { + "epoch": 3.289563909774436, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0585, + "step": 59689 + }, + { + "epoch": 3.289624060150376, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9559, + "step": 59690 + }, + { + "epoch": 3.2896842105263158, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9269, + "step": 59691 + }, + { + "epoch": 3.2897443609022554, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 59692 + }, + { + "epoch": 3.2898045112781955, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 59693 + }, + { + "epoch": 3.289864661654135, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 59694 + }, + { + "epoch": 3.2899248120300753, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.1046, + "step": 59695 + }, + { + "epoch": 3.289984962406015, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 59696 + }, + { + "epoch": 3.290045112781955, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 59697 + }, + { + "epoch": 3.2901052631578946, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 59698 + }, + { + "epoch": 3.2901654135338347, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 59699 + }, + { + "epoch": 3.2902255639097744, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9267, + "step": 59700 + }, + { + "epoch": 3.290285714285714, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 59701 + }, + { + "epoch": 3.290345864661654, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 59702 + }, + { + "epoch": 3.2904060150375942, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 59703 + }, + { + "epoch": 3.290466165413534, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 59704 + }, + { + "epoch": 3.2905263157894735, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 59705 + }, + { + "epoch": 3.2905864661654136, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1577, + "step": 59706 + }, + { + "epoch": 3.2906466165413533, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 59707 + }, + { + "epoch": 3.2907067669172934, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 59708 + }, + { + "epoch": 3.290766917293233, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9271, + "step": 59709 + }, + { + "epoch": 3.290827067669173, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 59710 + }, + { + "epoch": 3.2908872180451127, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 59711 + }, + { + "epoch": 3.290947368421053, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 59712 + }, + { + "epoch": 3.2910075187969925, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 59713 + }, + { + "epoch": 3.291067669172932, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 59714 + }, + { + "epoch": 3.2911278195488722, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 59715 + }, + { + "epoch": 3.291187969924812, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 59716 + }, + { + "epoch": 3.291248120300752, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9744, + "step": 59717 + }, + { + "epoch": 3.2913082706766916, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 59718 + }, + { + "epoch": 3.2913684210526317, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 59719 + }, + { + "epoch": 3.2914285714285714, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 59720 + }, + { + "epoch": 3.2914887218045115, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 59721 + }, + { + "epoch": 3.291548872180451, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 59722 + }, + { + "epoch": 3.2916090225563908, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 59723 + }, + { + "epoch": 3.291669172932331, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 59724 + }, + { + "epoch": 3.2917293233082705, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 59725 + }, + { + "epoch": 3.2917894736842106, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9263, + "step": 59726 + }, + { + "epoch": 3.2918496240601502, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 59727 + }, + { + "epoch": 3.2919097744360903, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9225, + "step": 59728 + }, + { + "epoch": 3.29196992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0951, + "step": 59729 + }, + { + "epoch": 3.29203007518797, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 59730 + }, + { + "epoch": 3.2920902255639097, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 59731 + }, + { + "epoch": 3.2921503759398494, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.8937, + "step": 59732 + }, + { + "epoch": 3.2922105263157895, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 59733 + }, + { + "epoch": 3.2922706766917296, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 59734 + }, + { + "epoch": 3.292330827067669, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 59735 + }, + { + "epoch": 3.292390977443609, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0782, + "step": 59736 + }, + { + "epoch": 3.292451127819549, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 59737 + }, + { + "epoch": 3.2925112781954886, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 59738 + }, + { + "epoch": 3.2925714285714287, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 59739 + }, + { + "epoch": 3.2926315789473684, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 59740 + }, + { + "epoch": 3.2926917293233084, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 59741 + }, + { + "epoch": 3.292751879699248, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9274, + "step": 59742 + }, + { + "epoch": 3.292812030075188, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 59743 + }, + { + "epoch": 3.292872180451128, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 59744 + }, + { + "epoch": 3.2929323308270675, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 59745 + }, + { + "epoch": 3.2929924812030076, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 59746 + }, + { + "epoch": 3.2930526315789472, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0116, + "step": 59747 + }, + { + "epoch": 3.2931127819548873, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 59748 + }, + { + "epoch": 3.293172932330827, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1302, + "step": 59749 + }, + { + "epoch": 3.293233082706767, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 59750 + }, + { + "epoch": 3.2932932330827067, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0534, + "step": 59751 + }, + { + "epoch": 3.293353383458647, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 59752 + }, + { + "epoch": 3.2934135338345865, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 59753 + }, + { + "epoch": 3.293473684210526, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 59754 + }, + { + "epoch": 3.293533834586466, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 59755 + }, + { + "epoch": 3.293593984962406, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 59756 + }, + { + "epoch": 3.293654135338346, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 59757 + }, + { + "epoch": 3.2937142857142856, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 59758 + }, + { + "epoch": 3.2937744360902257, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.941, + "step": 59759 + }, + { + "epoch": 3.2938345864661653, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 59760 + }, + { + "epoch": 3.2938947368421054, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 59761 + }, + { + "epoch": 3.293954887218045, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9196, + "step": 59762 + }, + { + "epoch": 3.2940150375939847, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 59763 + }, + { + "epoch": 3.294075187969925, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 59764 + }, + { + "epoch": 3.2941353383458645, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 59765 + }, + { + "epoch": 3.2941954887218046, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 59766 + }, + { + "epoch": 3.294255639097744, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1145, + "step": 59767 + }, + { + "epoch": 3.2943157894736843, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.956, + "step": 59768 + }, + { + "epoch": 3.294375939849624, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 59769 + }, + { + "epoch": 3.294436090225564, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 59770 + }, + { + "epoch": 3.2944962406015037, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 59771 + }, + { + "epoch": 3.294556390977444, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 59772 + }, + { + "epoch": 3.2946165413533834, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.222, + "step": 59773 + }, + { + "epoch": 3.2946766917293235, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 59774 + }, + { + "epoch": 3.294736842105263, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9091, + "step": 59775 + }, + { + "epoch": 3.294796992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9169, + "step": 59776 + }, + { + "epoch": 3.294857142857143, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0633, + "step": 59777 + }, + { + "epoch": 3.2949172932330826, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 59778 + }, + { + "epoch": 3.2949774436090227, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 59779 + }, + { + "epoch": 3.2950375939849623, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 59780 + }, + { + "epoch": 3.2950977443609024, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 59781 + }, + { + "epoch": 3.295157894736842, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 59782 + }, + { + "epoch": 3.295218045112782, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0655, + "step": 59783 + }, + { + "epoch": 3.295278195488722, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 59784 + }, + { + "epoch": 3.2953383458646615, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 59785 + }, + { + "epoch": 3.2953984962406015, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 59786 + }, + { + "epoch": 3.295458646616541, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9383, + "step": 59787 + }, + { + "epoch": 3.2955187969924813, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 59788 + }, + { + "epoch": 3.295578947368421, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 59789 + }, + { + "epoch": 3.295639097744361, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 59790 + }, + { + "epoch": 3.2956992481203007, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 59791 + }, + { + "epoch": 3.2957593984962408, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 59792 + }, + { + "epoch": 3.2958195488721804, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 59793 + }, + { + "epoch": 3.29587969924812, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 59794 + }, + { + "epoch": 3.29593984962406, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 59795 + }, + { + "epoch": 3.296, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 59796 + }, + { + "epoch": 3.29606015037594, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9473, + "step": 59797 + }, + { + "epoch": 3.2961203007518796, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9411, + "step": 59798 + }, + { + "epoch": 3.2961804511278197, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1335, + "step": 59799 + }, + { + "epoch": 3.2962406015037593, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 59800 + }, + { + "epoch": 3.2963007518796994, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 59801 + }, + { + "epoch": 3.296360902255639, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9247, + "step": 59802 + }, + { + "epoch": 3.296421052631579, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 59803 + }, + { + "epoch": 3.296481203007519, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 59804 + }, + { + "epoch": 3.296541353383459, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 59805 + }, + { + "epoch": 3.2966015037593985, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9144, + "step": 59806 + }, + { + "epoch": 3.296661654135338, + "grad_norm": 1.421875, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 59807 + }, + { + "epoch": 3.2967218045112783, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 59808 + }, + { + "epoch": 3.296781954887218, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9339, + "step": 59809 + }, + { + "epoch": 3.296842105263158, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 59810 + }, + { + "epoch": 3.2969022556390977, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 59811 + }, + { + "epoch": 3.2969624060150378, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9173, + "step": 59812 + }, + { + "epoch": 3.2970225563909774, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 59813 + }, + { + "epoch": 3.2970827067669175, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 59814 + }, + { + "epoch": 3.297142857142857, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 59815 + }, + { + "epoch": 3.297203007518797, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 59816 + }, + { + "epoch": 3.297263157894737, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.8481, + "step": 59817 + }, + { + "epoch": 3.2973233082706765, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 59818 + }, + { + "epoch": 3.2973834586466166, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 59819 + }, + { + "epoch": 3.2974436090225563, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 59820 + }, + { + "epoch": 3.2975037593984964, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 59821 + }, + { + "epoch": 3.297563909774436, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 59822 + }, + { + "epoch": 3.297624060150376, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 59823 + }, + { + "epoch": 3.2976842105263158, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 59824 + }, + { + "epoch": 3.2977443609022554, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 59825 + }, + { + "epoch": 3.2978045112781955, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 59826 + }, + { + "epoch": 3.297864661654135, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 59827 + }, + { + "epoch": 3.2979248120300753, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 59828 + }, + { + "epoch": 3.297984962406015, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 59829 + }, + { + "epoch": 3.298045112781955, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 59830 + }, + { + "epoch": 3.2981052631578947, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 59831 + }, + { + "epoch": 3.2981654135338347, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 59832 + }, + { + "epoch": 3.2982255639097744, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9931, + "step": 59833 + }, + { + "epoch": 3.298285714285714, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9842, + "step": 59834 + }, + { + "epoch": 3.298345864661654, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 59835 + }, + { + "epoch": 3.2984060150375942, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 59836 + }, + { + "epoch": 3.298466165413534, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 59837 + }, + { + "epoch": 3.2985263157894735, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 59838 + }, + { + "epoch": 3.2985864661654136, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 59839 + }, + { + "epoch": 3.2986466165413533, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9294, + "step": 59840 + }, + { + "epoch": 3.2987067669172934, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 59841 + }, + { + "epoch": 3.298766917293233, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 59842 + }, + { + "epoch": 3.298827067669173, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 59843 + }, + { + "epoch": 3.2988872180451128, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 59844 + }, + { + "epoch": 3.298947368421053, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 59845 + }, + { + "epoch": 3.2990075187969925, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 59846 + }, + { + "epoch": 3.299067669172932, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 59847 + }, + { + "epoch": 3.2991278195488722, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 59848 + }, + { + "epoch": 3.299187969924812, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 59849 + }, + { + "epoch": 3.299248120300752, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 59850 + }, + { + "epoch": 3.2993082706766916, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 59851 + }, + { + "epoch": 3.2993684210526317, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 59852 + }, + { + "epoch": 3.2994285714285714, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 59853 + }, + { + "epoch": 3.2994887218045115, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 59854 + }, + { + "epoch": 3.299548872180451, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 59855 + }, + { + "epoch": 3.2996090225563908, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 59856 + }, + { + "epoch": 3.299669172932331, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 59857 + }, + { + "epoch": 3.2997293233082705, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0712, + "step": 59858 + }, + { + "epoch": 3.2997894736842106, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 59859 + }, + { + "epoch": 3.2998496240601503, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 59860 + }, + { + "epoch": 3.2999097744360903, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 59861 + }, + { + "epoch": 3.29996992481203, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 59862 + }, + { + "epoch": 3.30003007518797, + "grad_norm": 1.2421875, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 59863 + }, + { + "epoch": 3.3000902255639097, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 59864 + }, + { + "epoch": 3.3001503759398494, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 59865 + }, + { + "epoch": 3.3002105263157895, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 59866 + }, + { + "epoch": 3.300270676691729, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 59867 + }, + { + "epoch": 3.3003308270676692, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 59868 + }, + { + "epoch": 3.300390977443609, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9647, + "step": 59869 + }, + { + "epoch": 3.300451127819549, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 59870 + }, + { + "epoch": 3.3005112781954886, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 59871 + }, + { + "epoch": 3.3005714285714287, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 59872 + }, + { + "epoch": 3.3006315789473684, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0642, + "step": 59873 + }, + { + "epoch": 3.3006917293233085, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0603, + "step": 59874 + }, + { + "epoch": 3.300751879699248, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0523, + "step": 59875 + }, + { + "epoch": 3.300812030075188, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 59876 + }, + { + "epoch": 3.300872180451128, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 59877 + }, + { + "epoch": 3.3009323308270675, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 59878 + }, + { + "epoch": 3.3009924812030076, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 59879 + }, + { + "epoch": 3.3010526315789472, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 59880 + }, + { + "epoch": 3.3011127819548873, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 59881 + }, + { + "epoch": 3.301172932330827, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9243, + "step": 59882 + }, + { + "epoch": 3.301233082706767, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 59883 + }, + { + "epoch": 3.3012932330827067, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 59884 + }, + { + "epoch": 3.301353383458647, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 59885 + }, + { + "epoch": 3.3014135338345865, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 59886 + }, + { + "epoch": 3.301473684210526, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 59887 + }, + { + "epoch": 3.301533834586466, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 59888 + }, + { + "epoch": 3.301593984962406, + "grad_norm": 1.2265625, + "learning_rate": 3e-05, + "loss": 1.0575, + "step": 59889 + }, + { + "epoch": 3.301654135338346, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 59890 + }, + { + "epoch": 3.3017142857142856, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 59891 + }, + { + "epoch": 3.3017744360902257, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 59892 + }, + { + "epoch": 3.3018345864661653, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9119, + "step": 59893 + }, + { + "epoch": 3.3018947368421054, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 59894 + }, + { + "epoch": 3.301954887218045, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0547, + "step": 59895 + }, + { + "epoch": 3.3020150375939847, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 59896 + }, + { + "epoch": 3.302075187969925, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 59897 + }, + { + "epoch": 3.3021353383458645, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 59898 + }, + { + "epoch": 3.3021954887218046, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 59899 + }, + { + "epoch": 3.302255639097744, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 59900 + }, + { + "epoch": 3.3023157894736843, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9224, + "step": 59901 + }, + { + "epoch": 3.302375939849624, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 59902 + }, + { + "epoch": 3.302436090225564, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.8854, + "step": 59903 + }, + { + "epoch": 3.3024962406015037, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 59904 + }, + { + "epoch": 3.302556390977444, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.039, + "step": 59905 + }, + { + "epoch": 3.3026165413533835, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 59906 + }, + { + "epoch": 3.3026766917293235, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 59907 + }, + { + "epoch": 3.302736842105263, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 59908 + }, + { + "epoch": 3.302796992481203, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 59909 + }, + { + "epoch": 3.302857142857143, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 59910 + }, + { + "epoch": 3.3029172932330826, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 59911 + }, + { + "epoch": 3.3029774436090227, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 59912 + }, + { + "epoch": 3.3030375939849623, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 59913 + }, + { + "epoch": 3.3030977443609024, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 59914 + }, + { + "epoch": 3.303157894736842, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 59915 + }, + { + "epoch": 3.303218045112782, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0788, + "step": 59916 + }, + { + "epoch": 3.303278195488722, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1735, + "step": 59917 + }, + { + "epoch": 3.3033383458646615, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 59918 + }, + { + "epoch": 3.3033984962406016, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 59919 + }, + { + "epoch": 3.303458646616541, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0934, + "step": 59920 + }, + { + "epoch": 3.3035187969924813, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0317, + "step": 59921 + }, + { + "epoch": 3.303578947368421, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 59922 + }, + { + "epoch": 3.303639097744361, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 59923 + }, + { + "epoch": 3.3036992481203007, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 59924 + }, + { + "epoch": 3.303759398496241, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1826, + "step": 59925 + }, + { + "epoch": 3.3038195488721804, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9188, + "step": 59926 + }, + { + "epoch": 3.30387969924812, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 59927 + }, + { + "epoch": 3.30393984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 59928 + }, + { + "epoch": 3.304, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 59929 + }, + { + "epoch": 3.30406015037594, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9239, + "step": 59930 + }, + { + "epoch": 3.3041203007518796, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 59931 + }, + { + "epoch": 3.3041804511278197, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 59932 + }, + { + "epoch": 3.3042406015037593, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 59933 + }, + { + "epoch": 3.3043007518796994, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 59934 + }, + { + "epoch": 3.304360902255639, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 59935 + }, + { + "epoch": 3.3044210526315787, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 59936 + }, + { + "epoch": 3.304481203007519, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9174, + "step": 59937 + }, + { + "epoch": 3.304541353383459, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 59938 + }, + { + "epoch": 3.3046015037593985, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1158, + "step": 59939 + }, + { + "epoch": 3.304661654135338, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 59940 + }, + { + "epoch": 3.3047218045112783, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 59941 + }, + { + "epoch": 3.304781954887218, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 59942 + }, + { + "epoch": 3.304842105263158, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 59943 + }, + { + "epoch": 3.3049022556390977, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 59944 + }, + { + "epoch": 3.3049624060150378, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 59945 + }, + { + "epoch": 3.3050225563909774, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 59946 + }, + { + "epoch": 3.3050827067669175, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 59947 + }, + { + "epoch": 3.305142857142857, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 59948 + }, + { + "epoch": 3.305203007518797, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9242, + "step": 59949 + }, + { + "epoch": 3.305263157894737, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 59950 + }, + { + "epoch": 3.3053233082706766, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 59951 + }, + { + "epoch": 3.3053834586466166, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 59952 + }, + { + "epoch": 3.3054436090225563, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 59953 + }, + { + "epoch": 3.3055037593984964, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 59954 + }, + { + "epoch": 3.305563909774436, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 59955 + }, + { + "epoch": 3.305624060150376, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9088, + "step": 59956 + }, + { + "epoch": 3.305684210526316, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 59957 + }, + { + "epoch": 3.3057443609022554, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1352, + "step": 59958 + }, + { + "epoch": 3.3058045112781955, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9372, + "step": 59959 + }, + { + "epoch": 3.305864661654135, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 59960 + }, + { + "epoch": 3.3059248120300753, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 59961 + }, + { + "epoch": 3.305984962406015, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.911, + "step": 59962 + }, + { + "epoch": 3.306045112781955, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 59963 + }, + { + "epoch": 3.3061052631578947, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 59964 + }, + { + "epoch": 3.3061654135338348, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 59965 + }, + { + "epoch": 3.3062255639097744, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 59966 + }, + { + "epoch": 3.306285714285714, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 59967 + }, + { + "epoch": 3.306345864661654, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 59968 + }, + { + "epoch": 3.306406015037594, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 59969 + }, + { + "epoch": 3.306466165413534, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 59970 + }, + { + "epoch": 3.3065263157894735, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9332, + "step": 59971 + }, + { + "epoch": 3.3065864661654136, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 59972 + }, + { + "epoch": 3.3066466165413533, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 59973 + }, + { + "epoch": 3.3067067669172934, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 59974 + }, + { + "epoch": 3.306766917293233, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 59975 + }, + { + "epoch": 3.306827067669173, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 59976 + }, + { + "epoch": 3.3068872180451128, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 59977 + }, + { + "epoch": 3.306947368421053, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 59978 + }, + { + "epoch": 3.3070075187969925, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.2943, + "step": 59979 + }, + { + "epoch": 3.307067669172932, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.3327, + "step": 59980 + }, + { + "epoch": 3.3071278195488722, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 59981 + }, + { + "epoch": 3.307187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 59982 + }, + { + "epoch": 3.307248120300752, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 59983 + }, + { + "epoch": 3.3073082706766916, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 59984 + }, + { + "epoch": 3.3073684210526317, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 59985 + }, + { + "epoch": 3.3074285714285714, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9908, + "step": 59986 + }, + { + "epoch": 3.3074887218045115, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 59987 + }, + { + "epoch": 3.307548872180451, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0387, + "step": 59988 + }, + { + "epoch": 3.3076090225563908, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9199, + "step": 59989 + }, + { + "epoch": 3.307669172932331, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0252, + "step": 59990 + }, + { + "epoch": 3.3077293233082705, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 59991 + }, + { + "epoch": 3.3077894736842106, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 59992 + }, + { + "epoch": 3.3078496240601503, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 59993 + }, + { + "epoch": 3.3079097744360904, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 59994 + }, + { + "epoch": 3.30796992481203, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 59995 + }, + { + "epoch": 3.30803007518797, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 59996 + }, + { + "epoch": 3.3080902255639097, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 59997 + }, + { + "epoch": 3.3081503759398494, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 59998 + }, + { + "epoch": 3.3082105263157895, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 59999 + }, + { + "epoch": 3.308270676691729, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 60000 + }, + { + "epoch": 3.3083308270676692, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 60001 + }, + { + "epoch": 3.308390977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 60002 + }, + { + "epoch": 3.308451127819549, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 60003 + }, + { + "epoch": 3.3085112781954886, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 60004 + }, + { + "epoch": 3.3085714285714287, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 60005 + }, + { + "epoch": 3.3086315789473684, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9286, + "step": 60006 + }, + { + "epoch": 3.3086917293233085, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 60007 + }, + { + "epoch": 3.308751879699248, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 60008 + }, + { + "epoch": 3.308812030075188, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 60009 + }, + { + "epoch": 3.308872180451128, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.936, + "step": 60010 + }, + { + "epoch": 3.3089323308270675, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 60011 + }, + { + "epoch": 3.3089924812030076, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 60012 + }, + { + "epoch": 3.3090526315789472, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 60013 + }, + { + "epoch": 3.3091127819548873, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 60014 + }, + { + "epoch": 3.309172932330827, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1575, + "step": 60015 + }, + { + "epoch": 3.309233082706767, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1838, + "step": 60016 + }, + { + "epoch": 3.3092932330827067, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 60017 + }, + { + "epoch": 3.309353383458647, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 60018 + }, + { + "epoch": 3.3094135338345865, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 60019 + }, + { + "epoch": 3.309473684210526, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1163, + "step": 60020 + }, + { + "epoch": 3.309533834586466, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 60021 + }, + { + "epoch": 3.309593984962406, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9753, + "step": 60022 + }, + { + "epoch": 3.309654135338346, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0626, + "step": 60023 + }, + { + "epoch": 3.3097142857142856, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 60024 + }, + { + "epoch": 3.3097744360902257, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9505, + "step": 60025 + }, + { + "epoch": 3.3098345864661654, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 60026 + }, + { + "epoch": 3.3098947368421054, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 60027 + }, + { + "epoch": 3.309954887218045, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 60028 + }, + { + "epoch": 3.3100150375939847, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1157, + "step": 60029 + }, + { + "epoch": 3.310075187969925, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 60030 + }, + { + "epoch": 3.3101353383458645, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1284, + "step": 60031 + }, + { + "epoch": 3.3101954887218046, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 60032 + }, + { + "epoch": 3.3102556390977442, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9246, + "step": 60033 + }, + { + "epoch": 3.3103157894736843, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 60034 + }, + { + "epoch": 3.310375939849624, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 60035 + }, + { + "epoch": 3.310436090225564, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 60036 + }, + { + "epoch": 3.3104962406015037, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 60037 + }, + { + "epoch": 3.3105563909774434, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 60038 + }, + { + "epoch": 3.3106165413533835, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 60039 + }, + { + "epoch": 3.3106766917293236, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 60040 + }, + { + "epoch": 3.310736842105263, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 60041 + }, + { + "epoch": 3.310796992481203, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 60042 + }, + { + "epoch": 3.310857142857143, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 60043 + }, + { + "epoch": 3.3109172932330826, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.8622, + "step": 60044 + }, + { + "epoch": 3.3109774436090227, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 60045 + }, + { + "epoch": 3.3110375939849623, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 60046 + }, + { + "epoch": 3.3110977443609024, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9376, + "step": 60047 + }, + { + "epoch": 3.311157894736842, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 60048 + }, + { + "epoch": 3.311218045112782, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.8857, + "step": 60049 + }, + { + "epoch": 3.311278195488722, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 60050 + }, + { + "epoch": 3.3113383458646615, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 60051 + }, + { + "epoch": 3.3113984962406016, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 60052 + }, + { + "epoch": 3.311458646616541, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 60053 + }, + { + "epoch": 3.3115187969924813, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 60054 + }, + { + "epoch": 3.311578947368421, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 60055 + }, + { + "epoch": 3.311639097744361, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 60056 + }, + { + "epoch": 3.3116992481203007, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0282, + "step": 60057 + }, + { + "epoch": 3.311759398496241, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 60058 + }, + { + "epoch": 3.3118195488721804, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 60059 + }, + { + "epoch": 3.31187969924812, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 60060 + }, + { + "epoch": 3.31193984962406, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 60061 + }, + { + "epoch": 3.312, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 60062 + }, + { + "epoch": 3.31206015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 60063 + }, + { + "epoch": 3.3121203007518796, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 60064 + }, + { + "epoch": 3.3121804511278197, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0431, + "step": 60065 + }, + { + "epoch": 3.3122406015037593, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 60066 + }, + { + "epoch": 3.3123007518796994, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 60067 + }, + { + "epoch": 3.312360902255639, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 60068 + }, + { + "epoch": 3.3124210526315787, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 60069 + }, + { + "epoch": 3.312481203007519, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 60070 + }, + { + "epoch": 3.3125413533834585, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 60071 + }, + { + "epoch": 3.3126015037593985, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 60072 + }, + { + "epoch": 3.312661654135338, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 60073 + }, + { + "epoch": 3.3127218045112783, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 60074 + }, + { + "epoch": 3.312781954887218, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 60075 + }, + { + "epoch": 3.312842105263158, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 60076 + }, + { + "epoch": 3.3129022556390977, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 60077 + }, + { + "epoch": 3.3129624060150378, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 60078 + }, + { + "epoch": 3.3130225563909774, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 60079 + }, + { + "epoch": 3.3130827067669175, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 60080 + }, + { + "epoch": 3.313142857142857, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 60081 + }, + { + "epoch": 3.313203007518797, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 60082 + }, + { + "epoch": 3.313263157894737, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 60083 + }, + { + "epoch": 3.3133233082706766, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 60084 + }, + { + "epoch": 3.3133834586466167, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 60085 + }, + { + "epoch": 3.3134436090225563, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9332, + "step": 60086 + }, + { + "epoch": 3.3135037593984964, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9533, + "step": 60087 + }, + { + "epoch": 3.313563909774436, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 60088 + }, + { + "epoch": 3.313624060150376, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1658, + "step": 60089 + }, + { + "epoch": 3.313684210526316, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 60090 + }, + { + "epoch": 3.3137443609022554, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0261, + "step": 60091 + }, + { + "epoch": 3.3138045112781955, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 60092 + }, + { + "epoch": 3.313864661654135, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 60093 + }, + { + "epoch": 3.3139248120300753, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 60094 + }, + { + "epoch": 3.313984962406015, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0854, + "step": 60095 + }, + { + "epoch": 3.314045112781955, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 60096 + }, + { + "epoch": 3.3141052631578947, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.906, + "step": 60097 + }, + { + "epoch": 3.3141654135338348, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 60098 + }, + { + "epoch": 3.3142255639097744, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9029, + "step": 60099 + }, + { + "epoch": 3.314285714285714, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 60100 + }, + { + "epoch": 3.314345864661654, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 60101 + }, + { + "epoch": 3.314406015037594, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 60102 + }, + { + "epoch": 3.314466165413534, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 60103 + }, + { + "epoch": 3.3145263157894735, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 60104 + }, + { + "epoch": 3.3145864661654136, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0083, + "step": 60105 + }, + { + "epoch": 3.3146466165413533, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 60106 + }, + { + "epoch": 3.3147067669172934, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 60107 + }, + { + "epoch": 3.314766917293233, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 60108 + }, + { + "epoch": 3.314827067669173, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 60109 + }, + { + "epoch": 3.3148872180451128, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 60110 + }, + { + "epoch": 3.314947368421053, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 60111 + }, + { + "epoch": 3.3150075187969925, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0528, + "step": 60112 + }, + { + "epoch": 3.315067669172932, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 60113 + }, + { + "epoch": 3.3151278195488723, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9334, + "step": 60114 + }, + { + "epoch": 3.315187969924812, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 60115 + }, + { + "epoch": 3.315248120300752, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 60116 + }, + { + "epoch": 3.3153082706766916, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 60117 + }, + { + "epoch": 3.3153684210526317, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9109, + "step": 60118 + }, + { + "epoch": 3.3154285714285714, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 60119 + }, + { + "epoch": 3.3154887218045115, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 60120 + }, + { + "epoch": 3.315548872180451, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 60121 + }, + { + "epoch": 3.315609022556391, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 60122 + }, + { + "epoch": 3.315669172932331, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 60123 + }, + { + "epoch": 3.3157293233082705, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 60124 + }, + { + "epoch": 3.3157894736842106, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 60125 + }, + { + "epoch": 3.3158496240601503, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 60126 + }, + { + "epoch": 3.3159097744360904, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 60127 + }, + { + "epoch": 3.31596992481203, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 60128 + }, + { + "epoch": 3.31603007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0451, + "step": 60129 + }, + { + "epoch": 3.3160902255639098, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 60130 + }, + { + "epoch": 3.3161503759398494, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 60131 + }, + { + "epoch": 3.3162105263157895, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9944, + "step": 60132 + }, + { + "epoch": 3.316270676691729, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 60133 + }, + { + "epoch": 3.3163308270676692, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.961, + "step": 60134 + }, + { + "epoch": 3.316390977443609, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 60135 + }, + { + "epoch": 3.316451127819549, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 60136 + }, + { + "epoch": 3.3165112781954886, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0186, + "step": 60137 + }, + { + "epoch": 3.3165714285714287, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9119, + "step": 60138 + }, + { + "epoch": 3.3166315789473684, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9267, + "step": 60139 + }, + { + "epoch": 3.316691729323308, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 60140 + }, + { + "epoch": 3.316751879699248, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0123, + "step": 60141 + }, + { + "epoch": 3.316812030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0927, + "step": 60142 + }, + { + "epoch": 3.316872180451128, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 60143 + }, + { + "epoch": 3.3169323308270675, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9749, + "step": 60144 + }, + { + "epoch": 3.3169924812030076, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 60145 + }, + { + "epoch": 3.3170526315789473, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 60146 + }, + { + "epoch": 3.3171127819548873, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 60147 + }, + { + "epoch": 3.317172932330827, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9049, + "step": 60148 + }, + { + "epoch": 3.317233082706767, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 60149 + }, + { + "epoch": 3.3172932330827067, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9905, + "step": 60150 + }, + { + "epoch": 3.317353383458647, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 60151 + }, + { + "epoch": 3.3174135338345865, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 60152 + }, + { + "epoch": 3.317473684210526, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 60153 + }, + { + "epoch": 3.3175338345864662, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 60154 + }, + { + "epoch": 3.317593984962406, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 60155 + }, + { + "epoch": 3.317654135338346, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 60156 + }, + { + "epoch": 3.3177142857142856, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.8718, + "step": 60157 + }, + { + "epoch": 3.3177744360902257, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 60158 + }, + { + "epoch": 3.3178345864661654, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 60159 + }, + { + "epoch": 3.3178947368421055, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 60160 + }, + { + "epoch": 3.317954887218045, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9873, + "step": 60161 + }, + { + "epoch": 3.3180150375939848, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 60162 + }, + { + "epoch": 3.318075187969925, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 60163 + }, + { + "epoch": 3.3181353383458645, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 60164 + }, + { + "epoch": 3.3181954887218046, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 60165 + }, + { + "epoch": 3.3182556390977442, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 60166 + }, + { + "epoch": 3.3183157894736843, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 60167 + }, + { + "epoch": 3.318375939849624, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 60168 + }, + { + "epoch": 3.318436090225564, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 60169 + }, + { + "epoch": 3.3184962406015037, + "grad_norm": 1.359375, + "learning_rate": 3e-05, + "loss": 1.1273, + "step": 60170 + }, + { + "epoch": 3.3185563909774434, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 60171 + }, + { + "epoch": 3.3186165413533835, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9025, + "step": 60172 + }, + { + "epoch": 3.3186766917293236, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 60173 + }, + { + "epoch": 3.318736842105263, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0115, + "step": 60174 + }, + { + "epoch": 3.318796992481203, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 60175 + }, + { + "epoch": 3.318857142857143, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 60176 + }, + { + "epoch": 3.3189172932330826, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 60177 + }, + { + "epoch": 3.3189774436090227, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 60178 + }, + { + "epoch": 3.3190375939849623, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9092, + "step": 60179 + }, + { + "epoch": 3.3190977443609024, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 60180 + }, + { + "epoch": 3.319157894736842, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9587, + "step": 60181 + }, + { + "epoch": 3.319218045112782, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.086, + "step": 60182 + }, + { + "epoch": 3.319278195488722, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0463, + "step": 60183 + }, + { + "epoch": 3.3193383458646615, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0957, + "step": 60184 + }, + { + "epoch": 3.3193984962406016, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 60185 + }, + { + "epoch": 3.319458646616541, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 60186 + }, + { + "epoch": 3.3195187969924813, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 60187 + }, + { + "epoch": 3.319578947368421, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 60188 + }, + { + "epoch": 3.319639097744361, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0471, + "step": 60189 + }, + { + "epoch": 3.3196992481203007, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 60190 + }, + { + "epoch": 3.319759398496241, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9298, + "step": 60191 + }, + { + "epoch": 3.3198195488721804, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 60192 + }, + { + "epoch": 3.31987969924812, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9152, + "step": 60193 + }, + { + "epoch": 3.31993984962406, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 60194 + }, + { + "epoch": 3.32, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 60195 + }, + { + "epoch": 3.32006015037594, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 60196 + }, + { + "epoch": 3.3201203007518796, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 60197 + }, + { + "epoch": 3.3201804511278197, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 60198 + }, + { + "epoch": 3.3202406015037593, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9336, + "step": 60199 + }, + { + "epoch": 3.3203007518796994, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 60200 + }, + { + "epoch": 3.320360902255639, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 60201 + }, + { + "epoch": 3.3204210526315787, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 60202 + }, + { + "epoch": 3.320481203007519, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 0.9586, + "step": 60203 + }, + { + "epoch": 3.3205413533834585, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 60204 + }, + { + "epoch": 3.3206015037593986, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 60205 + }, + { + "epoch": 3.320661654135338, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 60206 + }, + { + "epoch": 3.3207218045112783, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 60207 + }, + { + "epoch": 3.320781954887218, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 60208 + }, + { + "epoch": 3.320842105263158, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 60209 + }, + { + "epoch": 3.3209022556390977, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 60210 + }, + { + "epoch": 3.320962406015038, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 60211 + }, + { + "epoch": 3.3210225563909774, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.93, + "step": 60212 + }, + { + "epoch": 3.3210827067669175, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 60213 + }, + { + "epoch": 3.321142857142857, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 60214 + }, + { + "epoch": 3.321203007518797, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 60215 + }, + { + "epoch": 3.321263157894737, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 60216 + }, + { + "epoch": 3.3213233082706766, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9497, + "step": 60217 + }, + { + "epoch": 3.3213834586466167, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 60218 + }, + { + "epoch": 3.3214436090225563, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.2284, + "step": 60219 + }, + { + "epoch": 3.3215037593984964, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 60220 + }, + { + "epoch": 3.321563909774436, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 60221 + }, + { + "epoch": 3.321624060150376, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 60222 + }, + { + "epoch": 3.321684210526316, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 60223 + }, + { + "epoch": 3.3217443609022554, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9389, + "step": 60224 + }, + { + "epoch": 3.3218045112781955, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 60225 + }, + { + "epoch": 3.321864661654135, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 60226 + }, + { + "epoch": 3.3219248120300753, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 60227 + }, + { + "epoch": 3.321984962406015, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 60228 + }, + { + "epoch": 3.322045112781955, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 60229 + }, + { + "epoch": 3.3221052631578947, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0668, + "step": 60230 + }, + { + "epoch": 3.3221654135338348, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0413, + "step": 60231 + }, + { + "epoch": 3.3222255639097744, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 60232 + }, + { + "epoch": 3.322285714285714, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.8986, + "step": 60233 + }, + { + "epoch": 3.322345864661654, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 60234 + }, + { + "epoch": 3.322406015037594, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 60235 + }, + { + "epoch": 3.322466165413534, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 60236 + }, + { + "epoch": 3.3225263157894735, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 60237 + }, + { + "epoch": 3.3225864661654136, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 60238 + }, + { + "epoch": 3.3226466165413533, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 60239 + }, + { + "epoch": 3.3227067669172934, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 60240 + }, + { + "epoch": 3.322766917293233, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.09, + "step": 60241 + }, + { + "epoch": 3.3228270676691727, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1554, + "step": 60242 + }, + { + "epoch": 3.322887218045113, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 60243 + }, + { + "epoch": 3.322947368421053, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 60244 + }, + { + "epoch": 3.3230075187969925, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 60245 + }, + { + "epoch": 3.323067669172932, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9154, + "step": 60246 + }, + { + "epoch": 3.3231278195488723, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9156, + "step": 60247 + }, + { + "epoch": 3.323187969924812, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 60248 + }, + { + "epoch": 3.323248120300752, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 60249 + }, + { + "epoch": 3.3233082706766917, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 60250 + }, + { + "epoch": 3.3233684210526317, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1566, + "step": 60251 + }, + { + "epoch": 3.3234285714285714, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 60252 + }, + { + "epoch": 3.3234887218045115, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 60253 + }, + { + "epoch": 3.323548872180451, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 60254 + }, + { + "epoch": 3.323609022556391, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0718, + "step": 60255 + }, + { + "epoch": 3.323669172932331, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 60256 + }, + { + "epoch": 3.3237293233082705, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 60257 + }, + { + "epoch": 3.3237894736842106, + "grad_norm": 1.2578125, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 60258 + }, + { + "epoch": 3.3238496240601503, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.8895, + "step": 60259 + }, + { + "epoch": 3.3239097744360904, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 60260 + }, + { + "epoch": 3.32396992481203, + "grad_norm": 1.546875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 60261 + }, + { + "epoch": 3.32403007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 60262 + }, + { + "epoch": 3.3240902255639098, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9448, + "step": 60263 + }, + { + "epoch": 3.3241503759398494, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 60264 + }, + { + "epoch": 3.3242105263157895, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 60265 + }, + { + "epoch": 3.324270676691729, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 60266 + }, + { + "epoch": 3.3243308270676692, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9687, + "step": 60267 + }, + { + "epoch": 3.324390977443609, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 60268 + }, + { + "epoch": 3.324451127819549, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 60269 + }, + { + "epoch": 3.3245112781954886, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 60270 + }, + { + "epoch": 3.3245714285714287, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 60271 + }, + { + "epoch": 3.3246315789473684, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 60272 + }, + { + "epoch": 3.324691729323308, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 60273 + }, + { + "epoch": 3.324751879699248, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 60274 + }, + { + "epoch": 3.324812030075188, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.001, + "step": 60275 + }, + { + "epoch": 3.324872180451128, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 60276 + }, + { + "epoch": 3.3249323308270675, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 60277 + }, + { + "epoch": 3.3249924812030076, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0942, + "step": 60278 + }, + { + "epoch": 3.3250526315789473, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0211, + "step": 60279 + }, + { + "epoch": 3.3251127819548874, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 60280 + }, + { + "epoch": 3.325172932330827, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 60281 + }, + { + "epoch": 3.325233082706767, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 60282 + }, + { + "epoch": 3.3252932330827067, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0495, + "step": 60283 + }, + { + "epoch": 3.325353383458647, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 60284 + }, + { + "epoch": 3.3254135338345865, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 60285 + }, + { + "epoch": 3.325473684210526, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9001, + "step": 60286 + }, + { + "epoch": 3.3255338345864662, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 60287 + }, + { + "epoch": 3.325593984962406, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.901, + "step": 60288 + }, + { + "epoch": 3.325654135338346, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9734, + "step": 60289 + }, + { + "epoch": 3.3257142857142856, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 60290 + }, + { + "epoch": 3.3257744360902257, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9978, + "step": 60291 + }, + { + "epoch": 3.3258345864661654, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.1365, + "step": 60292 + }, + { + "epoch": 3.3258947368421055, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9875, + "step": 60293 + }, + { + "epoch": 3.325954887218045, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 60294 + }, + { + "epoch": 3.3260150375939848, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9129, + "step": 60295 + }, + { + "epoch": 3.326075187969925, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 60296 + }, + { + "epoch": 3.3261353383458645, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 60297 + }, + { + "epoch": 3.3261954887218046, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9479, + "step": 60298 + }, + { + "epoch": 3.3262556390977442, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 60299 + }, + { + "epoch": 3.3263157894736843, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 60300 + }, + { + "epoch": 3.326375939849624, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9362, + "step": 60301 + }, + { + "epoch": 3.326436090225564, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 60302 + }, + { + "epoch": 3.3264962406015037, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1484, + "step": 60303 + }, + { + "epoch": 3.3265563909774434, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 60304 + }, + { + "epoch": 3.3266165413533835, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 60305 + }, + { + "epoch": 3.326676691729323, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9059, + "step": 60306 + }, + { + "epoch": 3.326736842105263, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 60307 + }, + { + "epoch": 3.326796992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 60308 + }, + { + "epoch": 3.326857142857143, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 60309 + }, + { + "epoch": 3.3269172932330826, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 60310 + }, + { + "epoch": 3.3269774436090227, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 60311 + }, + { + "epoch": 3.3270375939849623, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 60312 + }, + { + "epoch": 3.3270977443609024, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 60313 + }, + { + "epoch": 3.327157894736842, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 60314 + }, + { + "epoch": 3.327218045112782, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 60315 + }, + { + "epoch": 3.327278195488722, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 60316 + }, + { + "epoch": 3.3273383458646615, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 60317 + }, + { + "epoch": 3.3273984962406016, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 60318 + }, + { + "epoch": 3.3274586466165412, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 60319 + }, + { + "epoch": 3.3275187969924813, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1072, + "step": 60320 + }, + { + "epoch": 3.327578947368421, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 60321 + }, + { + "epoch": 3.327639097744361, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 60322 + }, + { + "epoch": 3.3276992481203007, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 60323 + }, + { + "epoch": 3.327759398496241, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 60324 + }, + { + "epoch": 3.3278195488721805, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 60325 + }, + { + "epoch": 3.32787969924812, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 60326 + }, + { + "epoch": 3.32793984962406, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 60327 + }, + { + "epoch": 3.328, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 60328 + }, + { + "epoch": 3.32806015037594, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 60329 + }, + { + "epoch": 3.3281203007518796, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 60330 + }, + { + "epoch": 3.3281804511278197, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 60331 + }, + { + "epoch": 3.3282406015037593, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9169, + "step": 60332 + }, + { + "epoch": 3.3283007518796994, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 60333 + }, + { + "epoch": 3.328360902255639, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 60334 + }, + { + "epoch": 3.3284210526315787, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 60335 + }, + { + "epoch": 3.328481203007519, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 60336 + }, + { + "epoch": 3.3285413533834585, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 60337 + }, + { + "epoch": 3.3286015037593986, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.927, + "step": 60338 + }, + { + "epoch": 3.328661654135338, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 60339 + }, + { + "epoch": 3.3287218045112783, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 60340 + }, + { + "epoch": 3.328781954887218, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 60341 + }, + { + "epoch": 3.328842105263158, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9427, + "step": 60342 + }, + { + "epoch": 3.3289022556390977, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 60343 + }, + { + "epoch": 3.3289624060150373, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 60344 + }, + { + "epoch": 3.3290225563909774, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 60345 + }, + { + "epoch": 3.3290827067669175, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 60346 + }, + { + "epoch": 3.329142857142857, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 60347 + }, + { + "epoch": 3.329203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0632, + "step": 60348 + }, + { + "epoch": 3.329263157894737, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 60349 + }, + { + "epoch": 3.3293233082706766, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 60350 + }, + { + "epoch": 3.3293834586466167, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.085, + "step": 60351 + }, + { + "epoch": 3.3294436090225563, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 60352 + }, + { + "epoch": 3.3295037593984964, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 60353 + }, + { + "epoch": 3.329563909774436, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9702, + "step": 60354 + }, + { + "epoch": 3.329624060150376, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 60355 + }, + { + "epoch": 3.329684210526316, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0635, + "step": 60356 + }, + { + "epoch": 3.3297443609022555, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0394, + "step": 60357 + }, + { + "epoch": 3.3298045112781955, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 60358 + }, + { + "epoch": 3.329864661654135, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 60359 + }, + { + "epoch": 3.3299248120300753, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 60360 + }, + { + "epoch": 3.329984962406015, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 60361 + }, + { + "epoch": 3.330045112781955, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 60362 + }, + { + "epoch": 3.3301052631578947, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 60363 + }, + { + "epoch": 3.3301654135338348, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 60364 + }, + { + "epoch": 3.3302255639097744, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 60365 + }, + { + "epoch": 3.330285714285714, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9221, + "step": 60366 + }, + { + "epoch": 3.330345864661654, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0482, + "step": 60367 + }, + { + "epoch": 3.330406015037594, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 60368 + }, + { + "epoch": 3.330466165413534, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9259, + "step": 60369 + }, + { + "epoch": 3.3305263157894736, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 60370 + }, + { + "epoch": 3.3305864661654137, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1759, + "step": 60371 + }, + { + "epoch": 3.3306466165413533, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9152, + "step": 60372 + }, + { + "epoch": 3.3307067669172934, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 60373 + }, + { + "epoch": 3.330766917293233, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9279, + "step": 60374 + }, + { + "epoch": 3.3308270676691727, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 60375 + }, + { + "epoch": 3.330887218045113, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 60376 + }, + { + "epoch": 3.330947368421053, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9275, + "step": 60377 + }, + { + "epoch": 3.3310075187969925, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 60378 + }, + { + "epoch": 3.331067669172932, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 60379 + }, + { + "epoch": 3.3311278195488723, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 60380 + }, + { + "epoch": 3.331187969924812, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 60381 + }, + { + "epoch": 3.331248120300752, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9068, + "step": 60382 + }, + { + "epoch": 3.3313082706766917, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 60383 + }, + { + "epoch": 3.3313684210526318, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9518, + "step": 60384 + }, + { + "epoch": 3.3314285714285714, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1019, + "step": 60385 + }, + { + "epoch": 3.3314887218045115, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 60386 + }, + { + "epoch": 3.331548872180451, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0771, + "step": 60387 + }, + { + "epoch": 3.331609022556391, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 60388 + }, + { + "epoch": 3.331669172932331, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 60389 + }, + { + "epoch": 3.3317293233082705, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 60390 + }, + { + "epoch": 3.3317894736842106, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 60391 + }, + { + "epoch": 3.3318496240601503, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.8859, + "step": 60392 + }, + { + "epoch": 3.3319097744360904, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 60393 + }, + { + "epoch": 3.33196992481203, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 60394 + }, + { + "epoch": 3.33203007518797, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9395, + "step": 60395 + }, + { + "epoch": 3.3320902255639098, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9229, + "step": 60396 + }, + { + "epoch": 3.3321503759398494, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9368, + "step": 60397 + }, + { + "epoch": 3.3322105263157895, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 60398 + }, + { + "epoch": 3.332270676691729, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 60399 + }, + { + "epoch": 3.3323308270676693, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 60400 + }, + { + "epoch": 3.332390977443609, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 60401 + }, + { + "epoch": 3.332451127819549, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 60402 + }, + { + "epoch": 3.3325112781954886, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 60403 + }, + { + "epoch": 3.3325714285714287, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 60404 + }, + { + "epoch": 3.3326315789473684, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9163, + "step": 60405 + }, + { + "epoch": 3.332691729323308, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 60406 + }, + { + "epoch": 3.332751879699248, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 60407 + }, + { + "epoch": 3.332812030075188, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 60408 + }, + { + "epoch": 3.332872180451128, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 60409 + }, + { + "epoch": 3.3329323308270675, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 60410 + }, + { + "epoch": 3.3329924812030076, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 60411 + }, + { + "epoch": 3.3330526315789473, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0729, + "step": 60412 + }, + { + "epoch": 3.3331127819548874, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9201, + "step": 60413 + }, + { + "epoch": 3.333172932330827, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 60414 + }, + { + "epoch": 3.333233082706767, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 60415 + }, + { + "epoch": 3.3332932330827068, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 60416 + }, + { + "epoch": 3.333353383458647, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 60417 + }, + { + "epoch": 3.3334135338345865, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 60418 + }, + { + "epoch": 3.333473684210526, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 60419 + }, + { + "epoch": 3.3335338345864662, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 60420 + }, + { + "epoch": 3.333593984962406, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9104, + "step": 60421 + }, + { + "epoch": 3.333654135338346, + "grad_norm": 1.203125, + "learning_rate": 3e-05, + "loss": 1.0926, + "step": 60422 + }, + { + "epoch": 3.3337142857142856, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9535, + "step": 60423 + }, + { + "epoch": 3.3337744360902257, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 60424 + }, + { + "epoch": 3.3338345864661654, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0499, + "step": 60425 + }, + { + "epoch": 3.3338947368421055, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 60426 + }, + { + "epoch": 3.333954887218045, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.94, + "step": 60427 + }, + { + "epoch": 3.3340150375939848, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1353, + "step": 60428 + }, + { + "epoch": 3.334075187969925, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 60429 + }, + { + "epoch": 3.3341353383458645, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 60430 + }, + { + "epoch": 3.3341954887218046, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 60431 + }, + { + "epoch": 3.3342556390977443, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 60432 + }, + { + "epoch": 3.3343157894736843, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 60433 + }, + { + "epoch": 3.334375939849624, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 60434 + }, + { + "epoch": 3.334436090225564, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 60435 + }, + { + "epoch": 3.3344962406015037, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 60436 + }, + { + "epoch": 3.3345563909774434, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 60437 + }, + { + "epoch": 3.3346165413533835, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 60438 + }, + { + "epoch": 3.334676691729323, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 60439 + }, + { + "epoch": 3.334736842105263, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 60440 + }, + { + "epoch": 3.334796992481203, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 60441 + }, + { + "epoch": 3.334857142857143, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 60442 + }, + { + "epoch": 3.3349172932330826, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 60443 + }, + { + "epoch": 3.3349774436090227, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0797, + "step": 60444 + }, + { + "epoch": 3.3350375939849624, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 60445 + }, + { + "epoch": 3.3350977443609024, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1117, + "step": 60446 + }, + { + "epoch": 3.335157894736842, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9286, + "step": 60447 + }, + { + "epoch": 3.335218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 60448 + }, + { + "epoch": 3.335278195488722, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 60449 + }, + { + "epoch": 3.3353383458646615, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 60450 + }, + { + "epoch": 3.3353984962406016, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9147, + "step": 60451 + }, + { + "epoch": 3.3354586466165412, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 60452 + }, + { + "epoch": 3.3355187969924813, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 60453 + }, + { + "epoch": 3.335578947368421, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 60454 + }, + { + "epoch": 3.335639097744361, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 60455 + }, + { + "epoch": 3.3356992481203007, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9494, + "step": 60456 + }, + { + "epoch": 3.335759398496241, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9683, + "step": 60457 + }, + { + "epoch": 3.3358195488721805, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 60458 + }, + { + "epoch": 3.33587969924812, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 60459 + }, + { + "epoch": 3.33593984962406, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 60460 + }, + { + "epoch": 3.336, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 60461 + }, + { + "epoch": 3.33606015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0419, + "step": 60462 + }, + { + "epoch": 3.3361203007518796, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 60463 + }, + { + "epoch": 3.3361804511278197, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0021, + "step": 60464 + }, + { + "epoch": 3.3362406015037593, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 60465 + }, + { + "epoch": 3.3363007518796994, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9253, + "step": 60466 + }, + { + "epoch": 3.336360902255639, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 60467 + }, + { + "epoch": 3.3364210526315787, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 60468 + }, + { + "epoch": 3.336481203007519, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 60469 + }, + { + "epoch": 3.3365413533834585, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 60470 + }, + { + "epoch": 3.3366015037593986, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9255, + "step": 60471 + }, + { + "epoch": 3.336661654135338, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 60472 + }, + { + "epoch": 3.3367218045112783, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9843, + "step": 60473 + }, + { + "epoch": 3.336781954887218, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 60474 + }, + { + "epoch": 3.336842105263158, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.081, + "step": 60475 + }, + { + "epoch": 3.3369022556390977, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 60476 + }, + { + "epoch": 3.3369624060150374, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 60477 + }, + { + "epoch": 3.3370225563909774, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.954, + "step": 60478 + }, + { + "epoch": 3.3370827067669175, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 60479 + }, + { + "epoch": 3.337142857142857, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0095, + "step": 60480 + }, + { + "epoch": 3.337203007518797, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 60481 + }, + { + "epoch": 3.337263157894737, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 60482 + }, + { + "epoch": 3.3373233082706766, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0197, + "step": 60483 + }, + { + "epoch": 3.3373834586466167, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 60484 + }, + { + "epoch": 3.3374436090225563, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9252, + "step": 60485 + }, + { + "epoch": 3.3375037593984964, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 60486 + }, + { + "epoch": 3.337563909774436, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 60487 + }, + { + "epoch": 3.337624060150376, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 60488 + }, + { + "epoch": 3.337684210526316, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9332, + "step": 60489 + }, + { + "epoch": 3.3377443609022555, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9598, + "step": 60490 + }, + { + "epoch": 3.3378045112781956, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 60491 + }, + { + "epoch": 3.337864661654135, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9027, + "step": 60492 + }, + { + "epoch": 3.3379248120300753, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 60493 + }, + { + "epoch": 3.337984962406015, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 60494 + }, + { + "epoch": 3.338045112781955, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 60495 + }, + { + "epoch": 3.3381052631578947, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 60496 + }, + { + "epoch": 3.338165413533835, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0133, + "step": 60497 + }, + { + "epoch": 3.3382255639097744, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 60498 + }, + { + "epoch": 3.338285714285714, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0682, + "step": 60499 + }, + { + "epoch": 3.338345864661654, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 60500 + }, + { + "epoch": 3.338406015037594, + "grad_norm": 0.50390625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 60501 + }, + { + "epoch": 3.338466165413534, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 60502 + }, + { + "epoch": 3.3385263157894736, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 60503 + }, + { + "epoch": 3.3385864661654137, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 60504 + }, + { + "epoch": 3.3386466165413533, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 60505 + }, + { + "epoch": 3.3387067669172934, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 60506 + }, + { + "epoch": 3.338766917293233, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 60507 + }, + { + "epoch": 3.3388270676691727, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.8705, + "step": 60508 + }, + { + "epoch": 3.338887218045113, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1276, + "step": 60509 + }, + { + "epoch": 3.3389473684210524, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 60510 + }, + { + "epoch": 3.3390075187969925, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9735, + "step": 60511 + }, + { + "epoch": 3.339067669172932, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 60512 + }, + { + "epoch": 3.3391278195488723, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9055, + "step": 60513 + }, + { + "epoch": 3.339187969924812, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0689, + "step": 60514 + }, + { + "epoch": 3.339248120300752, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 60515 + }, + { + "epoch": 3.3393082706766917, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 60516 + }, + { + "epoch": 3.3393684210526318, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9306, + "step": 60517 + }, + { + "epoch": 3.3394285714285714, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 60518 + }, + { + "epoch": 3.3394887218045115, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 60519 + }, + { + "epoch": 3.339548872180451, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9958, + "step": 60520 + }, + { + "epoch": 3.339609022556391, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.1595, + "step": 60521 + }, + { + "epoch": 3.339669172932331, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 60522 + }, + { + "epoch": 3.3397293233082705, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 60523 + }, + { + "epoch": 3.3397894736842106, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 60524 + }, + { + "epoch": 3.3398496240601503, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9402, + "step": 60525 + }, + { + "epoch": 3.3399097744360904, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 60526 + }, + { + "epoch": 3.33996992481203, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 60527 + }, + { + "epoch": 3.34003007518797, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9225, + "step": 60528 + }, + { + "epoch": 3.3400902255639098, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 60529 + }, + { + "epoch": 3.3401503759398494, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 60530 + }, + { + "epoch": 3.3402105263157895, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 60531 + }, + { + "epoch": 3.340270676691729, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 60532 + }, + { + "epoch": 3.3403308270676693, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9254, + "step": 60533 + }, + { + "epoch": 3.340390977443609, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9658, + "step": 60534 + }, + { + "epoch": 3.340451127819549, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0822, + "step": 60535 + }, + { + "epoch": 3.3405112781954887, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 60536 + }, + { + "epoch": 3.3405714285714287, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 60537 + }, + { + "epoch": 3.3406315789473684, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 60538 + }, + { + "epoch": 3.340691729323308, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 60539 + }, + { + "epoch": 3.340751879699248, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0269, + "step": 60540 + }, + { + "epoch": 3.340812030075188, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 60541 + }, + { + "epoch": 3.340872180451128, + "grad_norm": 1.4609375, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 60542 + }, + { + "epoch": 3.3409323308270675, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 60543 + }, + { + "epoch": 3.3409924812030076, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 60544 + }, + { + "epoch": 3.3410526315789473, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 60545 + }, + { + "epoch": 3.3411127819548874, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 60546 + }, + { + "epoch": 3.341172932330827, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 60547 + }, + { + "epoch": 3.341233082706767, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 60548 + }, + { + "epoch": 3.3412932330827068, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0696, + "step": 60549 + }, + { + "epoch": 3.341353383458647, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 60550 + }, + { + "epoch": 3.3414135338345865, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 60551 + }, + { + "epoch": 3.341473684210526, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 60552 + }, + { + "epoch": 3.3415338345864662, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0552, + "step": 60553 + }, + { + "epoch": 3.341593984962406, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 60554 + }, + { + "epoch": 3.341654135338346, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 60555 + }, + { + "epoch": 3.3417142857142856, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0786, + "step": 60556 + }, + { + "epoch": 3.3417744360902257, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 60557 + }, + { + "epoch": 3.3418345864661654, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 60558 + }, + { + "epoch": 3.3418947368421055, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 60559 + }, + { + "epoch": 3.341954887218045, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9347, + "step": 60560 + }, + { + "epoch": 3.3420150375939848, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9447, + "step": 60561 + }, + { + "epoch": 3.342075187969925, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0827, + "step": 60562 + }, + { + "epoch": 3.3421353383458645, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 60563 + }, + { + "epoch": 3.3421954887218046, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 60564 + }, + { + "epoch": 3.3422556390977443, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 60565 + }, + { + "epoch": 3.3423157894736844, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 60566 + }, + { + "epoch": 3.342375939849624, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 60567 + }, + { + "epoch": 3.342436090225564, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 60568 + }, + { + "epoch": 3.3424962406015037, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.8856, + "step": 60569 + }, + { + "epoch": 3.3425563909774434, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 60570 + }, + { + "epoch": 3.3426165413533835, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 60571 + }, + { + "epoch": 3.342676691729323, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 60572 + }, + { + "epoch": 3.3427368421052632, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 60573 + }, + { + "epoch": 3.342796992481203, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 60574 + }, + { + "epoch": 3.342857142857143, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9117, + "step": 60575 + }, + { + "epoch": 3.3429172932330826, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 60576 + }, + { + "epoch": 3.3429774436090227, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 60577 + }, + { + "epoch": 3.3430375939849624, + "grad_norm": 1.40625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 60578 + }, + { + "epoch": 3.343097744360902, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 60579 + }, + { + "epoch": 3.343157894736842, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 60580 + }, + { + "epoch": 3.343218045112782, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 60581 + }, + { + "epoch": 3.343278195488722, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9647, + "step": 60582 + }, + { + "epoch": 3.3433383458646615, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 60583 + }, + { + "epoch": 3.3433984962406016, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 60584 + }, + { + "epoch": 3.3434586466165412, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.933, + "step": 60585 + }, + { + "epoch": 3.3435187969924813, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 60586 + }, + { + "epoch": 3.343578947368421, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 60587 + }, + { + "epoch": 3.343639097744361, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 60588 + }, + { + "epoch": 3.3436992481203007, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0896, + "step": 60589 + }, + { + "epoch": 3.343759398496241, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 60590 + }, + { + "epoch": 3.3438195488721805, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 60591 + }, + { + "epoch": 3.34387969924812, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 60592 + }, + { + "epoch": 3.34393984962406, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 60593 + }, + { + "epoch": 3.344, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 60594 + }, + { + "epoch": 3.34406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 60595 + }, + { + "epoch": 3.3441203007518796, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9238, + "step": 60596 + }, + { + "epoch": 3.3441804511278197, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 60597 + }, + { + "epoch": 3.3442406015037593, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 60598 + }, + { + "epoch": 3.3443007518796994, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 60599 + }, + { + "epoch": 3.344360902255639, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0304, + "step": 60600 + }, + { + "epoch": 3.3444210526315787, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9225, + "step": 60601 + }, + { + "epoch": 3.344481203007519, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 60602 + }, + { + "epoch": 3.3445413533834585, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0903, + "step": 60603 + }, + { + "epoch": 3.3446015037593986, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0424, + "step": 60604 + }, + { + "epoch": 3.3446616541353382, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 60605 + }, + { + "epoch": 3.3447218045112783, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9199, + "step": 60606 + }, + { + "epoch": 3.344781954887218, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 60607 + }, + { + "epoch": 3.344842105263158, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 60608 + }, + { + "epoch": 3.3449022556390977, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 60609 + }, + { + "epoch": 3.3449624060150374, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9529, + "step": 60610 + }, + { + "epoch": 3.3450225563909775, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 60611 + }, + { + "epoch": 3.345082706766917, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 60612 + }, + { + "epoch": 3.345142857142857, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 60613 + }, + { + "epoch": 3.345203007518797, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 60614 + }, + { + "epoch": 3.345263157894737, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 60615 + }, + { + "epoch": 3.3453233082706766, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 60616 + }, + { + "epoch": 3.3453834586466167, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0795, + "step": 60617 + }, + { + "epoch": 3.3454436090225563, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 60618 + }, + { + "epoch": 3.3455037593984964, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9431, + "step": 60619 + }, + { + "epoch": 3.345563909774436, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 60620 + }, + { + "epoch": 3.345624060150376, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 60621 + }, + { + "epoch": 3.345684210526316, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 60622 + }, + { + "epoch": 3.3457443609022555, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 60623 + }, + { + "epoch": 3.3458045112781956, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 60624 + }, + { + "epoch": 3.345864661654135, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9085, + "step": 60625 + }, + { + "epoch": 3.3459248120300753, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 60626 + }, + { + "epoch": 3.345984962406015, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 60627 + }, + { + "epoch": 3.346045112781955, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 60628 + }, + { + "epoch": 3.3461052631578947, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 60629 + }, + { + "epoch": 3.346165413533835, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 60630 + }, + { + "epoch": 3.3462255639097744, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 60631 + }, + { + "epoch": 3.346285714285714, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0833, + "step": 60632 + }, + { + "epoch": 3.346345864661654, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 60633 + }, + { + "epoch": 3.346406015037594, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9139, + "step": 60634 + }, + { + "epoch": 3.346466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9304, + "step": 60635 + }, + { + "epoch": 3.3465263157894736, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1008, + "step": 60636 + }, + { + "epoch": 3.3465864661654137, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 60637 + }, + { + "epoch": 3.3466466165413533, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 60638 + }, + { + "epoch": 3.3467067669172934, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 60639 + }, + { + "epoch": 3.346766917293233, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 1.0839, + "step": 60640 + }, + { + "epoch": 3.3468270676691727, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.8743, + "step": 60641 + }, + { + "epoch": 3.346887218045113, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 60642 + }, + { + "epoch": 3.3469473684210524, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 60643 + }, + { + "epoch": 3.3470075187969925, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.901, + "step": 60644 + }, + { + "epoch": 3.347067669172932, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9793, + "step": 60645 + }, + { + "epoch": 3.3471278195488723, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 60646 + }, + { + "epoch": 3.347187969924812, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 60647 + }, + { + "epoch": 3.347248120300752, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1882, + "step": 60648 + }, + { + "epoch": 3.3473082706766917, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 60649 + }, + { + "epoch": 3.3473684210526318, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 60650 + }, + { + "epoch": 3.3474285714285714, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 60651 + }, + { + "epoch": 3.3474887218045115, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 60652 + }, + { + "epoch": 3.347548872180451, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 60653 + }, + { + "epoch": 3.347609022556391, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 60654 + }, + { + "epoch": 3.347669172932331, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 60655 + }, + { + "epoch": 3.3477293233082706, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9256, + "step": 60656 + }, + { + "epoch": 3.3477894736842106, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 60657 + }, + { + "epoch": 3.3478496240601503, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 60658 + }, + { + "epoch": 3.3479097744360904, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 60659 + }, + { + "epoch": 3.34796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 60660 + }, + { + "epoch": 3.34803007518797, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0288, + "step": 60661 + }, + { + "epoch": 3.34809022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 60662 + }, + { + "epoch": 3.3481503759398494, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9782, + "step": 60663 + }, + { + "epoch": 3.3482105263157895, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1319, + "step": 60664 + }, + { + "epoch": 3.348270676691729, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 60665 + }, + { + "epoch": 3.3483308270676693, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.08, + "step": 60666 + }, + { + "epoch": 3.348390977443609, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 60667 + }, + { + "epoch": 3.348451127819549, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 60668 + }, + { + "epoch": 3.3485112781954887, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 60669 + }, + { + "epoch": 3.3485714285714288, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9823, + "step": 60670 + }, + { + "epoch": 3.3486315789473684, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 60671 + }, + { + "epoch": 3.348691729323308, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 60672 + }, + { + "epoch": 3.348751879699248, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 60673 + }, + { + "epoch": 3.348812030075188, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 60674 + }, + { + "epoch": 3.348872180451128, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 60675 + }, + { + "epoch": 3.3489323308270675, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 60676 + }, + { + "epoch": 3.3489924812030076, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 60677 + }, + { + "epoch": 3.3490526315789473, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9327, + "step": 60678 + }, + { + "epoch": 3.3491127819548874, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0208, + "step": 60679 + }, + { + "epoch": 3.349172932330827, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 60680 + }, + { + "epoch": 3.3492330827067667, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 60681 + }, + { + "epoch": 3.3492932330827068, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 60682 + }, + { + "epoch": 3.349353383458647, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.1033, + "step": 60683 + }, + { + "epoch": 3.3494135338345865, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 60684 + }, + { + "epoch": 3.349473684210526, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 60685 + }, + { + "epoch": 3.3495338345864663, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 60686 + }, + { + "epoch": 3.349593984962406, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9374, + "step": 60687 + }, + { + "epoch": 3.349654135338346, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 60688 + }, + { + "epoch": 3.3497142857142856, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 60689 + }, + { + "epoch": 3.3497744360902257, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0148, + "step": 60690 + }, + { + "epoch": 3.3498345864661654, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 60691 + }, + { + "epoch": 3.3498947368421055, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 60692 + }, + { + "epoch": 3.349954887218045, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 60693 + }, + { + "epoch": 3.350015037593985, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 60694 + }, + { + "epoch": 3.350075187969925, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 60695 + }, + { + "epoch": 3.3501353383458645, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 60696 + }, + { + "epoch": 3.3501954887218046, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 60697 + }, + { + "epoch": 3.3502556390977443, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 60698 + }, + { + "epoch": 3.3503157894736844, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 60699 + }, + { + "epoch": 3.350375939849624, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0053, + "step": 60700 + }, + { + "epoch": 3.350436090225564, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0374, + "step": 60701 + }, + { + "epoch": 3.3504962406015038, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 60702 + }, + { + "epoch": 3.3505563909774434, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9936, + "step": 60703 + }, + { + "epoch": 3.3506165413533835, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 60704 + }, + { + "epoch": 3.350676691729323, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 60705 + }, + { + "epoch": 3.3507368421052632, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 60706 + }, + { + "epoch": 3.350796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.8965, + "step": 60707 + }, + { + "epoch": 3.350857142857143, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 60708 + }, + { + "epoch": 3.3509172932330826, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 60709 + }, + { + "epoch": 3.3509774436090227, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 60710 + }, + { + "epoch": 3.3510375939849624, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 60711 + }, + { + "epoch": 3.351097744360902, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 60712 + }, + { + "epoch": 3.351157894736842, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 60713 + }, + { + "epoch": 3.3512180451127818, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9346, + "step": 60714 + }, + { + "epoch": 3.351278195488722, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 60715 + }, + { + "epoch": 3.3513383458646615, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 60716 + }, + { + "epoch": 3.3513984962406016, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 60717 + }, + { + "epoch": 3.3514586466165412, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 60718 + }, + { + "epoch": 3.3515187969924813, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 60719 + }, + { + "epoch": 3.351578947368421, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9158, + "step": 60720 + }, + { + "epoch": 3.351639097744361, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 60721 + }, + { + "epoch": 3.3516992481203007, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 60722 + }, + { + "epoch": 3.351759398496241, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0398, + "step": 60723 + }, + { + "epoch": 3.3518195488721805, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.3055, + "step": 60724 + }, + { + "epoch": 3.35187969924812, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 60725 + }, + { + "epoch": 3.35193984962406, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 60726 + }, + { + "epoch": 3.352, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1862, + "step": 60727 + }, + { + "epoch": 3.35206015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9025, + "step": 60728 + }, + { + "epoch": 3.3521203007518796, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0468, + "step": 60729 + }, + { + "epoch": 3.3521804511278197, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9656, + "step": 60730 + }, + { + "epoch": 3.3522406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 60731 + }, + { + "epoch": 3.3523007518796994, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 60732 + }, + { + "epoch": 3.352360902255639, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 60733 + }, + { + "epoch": 3.3524210526315787, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 60734 + }, + { + "epoch": 3.352481203007519, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 60735 + }, + { + "epoch": 3.3525413533834585, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 60736 + }, + { + "epoch": 3.3526015037593986, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 60737 + }, + { + "epoch": 3.3526616541353382, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9154, + "step": 60738 + }, + { + "epoch": 3.3527218045112783, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 60739 + }, + { + "epoch": 3.352781954887218, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 60740 + }, + { + "epoch": 3.352842105263158, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 60741 + }, + { + "epoch": 3.3529022556390977, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 60742 + }, + { + "epoch": 3.3529624060150374, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 60743 + }, + { + "epoch": 3.3530225563909775, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9517, + "step": 60744 + }, + { + "epoch": 3.353082706766917, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9222, + "step": 60745 + }, + { + "epoch": 3.353142857142857, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 60746 + }, + { + "epoch": 3.353203007518797, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 60747 + }, + { + "epoch": 3.353263157894737, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0609, + "step": 60748 + }, + { + "epoch": 3.3533233082706766, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 60749 + }, + { + "epoch": 3.3533834586466167, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 60750 + }, + { + "epoch": 3.3534436090225563, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 60751 + }, + { + "epoch": 3.3535037593984964, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 60752 + }, + { + "epoch": 3.353563909774436, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0922, + "step": 60753 + }, + { + "epoch": 3.353624060150376, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 60754 + }, + { + "epoch": 3.353684210526316, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0735, + "step": 60755 + }, + { + "epoch": 3.3537443609022555, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0665, + "step": 60756 + }, + { + "epoch": 3.3538045112781956, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9828, + "step": 60757 + }, + { + "epoch": 3.353864661654135, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 60758 + }, + { + "epoch": 3.3539248120300753, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 60759 + }, + { + "epoch": 3.353984962406015, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 60760 + }, + { + "epoch": 3.354045112781955, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 60761 + }, + { + "epoch": 3.3541052631578947, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 60762 + }, + { + "epoch": 3.354165413533835, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 60763 + }, + { + "epoch": 3.3542255639097744, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 60764 + }, + { + "epoch": 3.354285714285714, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 60765 + }, + { + "epoch": 3.354345864661654, + "grad_norm": 1.4921875, + "learning_rate": 3e-05, + "loss": 1.0806, + "step": 60766 + }, + { + "epoch": 3.354406015037594, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 60767 + }, + { + "epoch": 3.354466165413534, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 60768 + }, + { + "epoch": 3.3545263157894736, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 60769 + }, + { + "epoch": 3.3545864661654137, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9263, + "step": 60770 + }, + { + "epoch": 3.3546466165413533, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 60771 + }, + { + "epoch": 3.3547067669172934, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 60772 + }, + { + "epoch": 3.354766917293233, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 60773 + }, + { + "epoch": 3.3548270676691727, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 60774 + }, + { + "epoch": 3.354887218045113, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 60775 + }, + { + "epoch": 3.3549473684210525, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 60776 + }, + { + "epoch": 3.3550075187969925, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 60777 + }, + { + "epoch": 3.355067669172932, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1266, + "step": 60778 + }, + { + "epoch": 3.3551278195488723, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 60779 + }, + { + "epoch": 3.355187969924812, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 60780 + }, + { + "epoch": 3.355248120300752, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 60781 + }, + { + "epoch": 3.3553082706766917, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0233, + "step": 60782 + }, + { + "epoch": 3.3553684210526313, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9236, + "step": 60783 + }, + { + "epoch": 3.3554285714285714, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 60784 + }, + { + "epoch": 3.3554887218045115, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1445, + "step": 60785 + }, + { + "epoch": 3.355548872180451, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 60786 + }, + { + "epoch": 3.355609022556391, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 60787 + }, + { + "epoch": 3.355669172932331, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 60788 + }, + { + "epoch": 3.3557293233082706, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 60789 + }, + { + "epoch": 3.3557894736842107, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 60790 + }, + { + "epoch": 3.3558496240601503, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 60791 + }, + { + "epoch": 3.3559097744360904, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 60792 + }, + { + "epoch": 3.35596992481203, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9345, + "step": 60793 + }, + { + "epoch": 3.35603007518797, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 60794 + }, + { + "epoch": 3.35609022556391, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 60795 + }, + { + "epoch": 3.3561503759398494, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 60796 + }, + { + "epoch": 3.3562105263157895, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1124, + "step": 60797 + }, + { + "epoch": 3.356270676691729, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 60798 + }, + { + "epoch": 3.3563308270676693, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9057, + "step": 60799 + }, + { + "epoch": 3.356390977443609, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 60800 + }, + { + "epoch": 3.356451127819549, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 60801 + }, + { + "epoch": 3.3565112781954887, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0757, + "step": 60802 + }, + { + "epoch": 3.3565714285714288, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 60803 + }, + { + "epoch": 3.3566315789473684, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 60804 + }, + { + "epoch": 3.356691729323308, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 60805 + }, + { + "epoch": 3.356751879699248, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0787, + "step": 60806 + }, + { + "epoch": 3.356812030075188, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 60807 + }, + { + "epoch": 3.356872180451128, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9205, + "step": 60808 + }, + { + "epoch": 3.3569323308270675, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.1006, + "step": 60809 + }, + { + "epoch": 3.3569924812030076, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 60810 + }, + { + "epoch": 3.3570526315789473, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 60811 + }, + { + "epoch": 3.3571127819548874, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 60812 + }, + { + "epoch": 3.357172932330827, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 60813 + }, + { + "epoch": 3.3572330827067667, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 60814 + }, + { + "epoch": 3.3572932330827068, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 60815 + }, + { + "epoch": 3.3573533834586464, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 60816 + }, + { + "epoch": 3.3574135338345865, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0502, + "step": 60817 + }, + { + "epoch": 3.357473684210526, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 60818 + }, + { + "epoch": 3.3575338345864663, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0979, + "step": 60819 + }, + { + "epoch": 3.357593984962406, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 60820 + }, + { + "epoch": 3.357654135338346, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1134, + "step": 60821 + }, + { + "epoch": 3.3577142857142857, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 60822 + }, + { + "epoch": 3.3577744360902257, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 60823 + }, + { + "epoch": 3.3578345864661654, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 60824 + }, + { + "epoch": 3.3578947368421055, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 60825 + }, + { + "epoch": 3.357954887218045, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 60826 + }, + { + "epoch": 3.358015037593985, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1384, + "step": 60827 + }, + { + "epoch": 3.358075187969925, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 60828 + }, + { + "epoch": 3.3581353383458645, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 60829 + }, + { + "epoch": 3.3581954887218046, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 60830 + }, + { + "epoch": 3.3582556390977443, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9284, + "step": 60831 + }, + { + "epoch": 3.3583157894736844, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 60832 + }, + { + "epoch": 3.358375939849624, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 60833 + }, + { + "epoch": 3.358436090225564, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 60834 + }, + { + "epoch": 3.3584962406015038, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 60835 + }, + { + "epoch": 3.3585563909774434, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 60836 + }, + { + "epoch": 3.3586165413533835, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 60837 + }, + { + "epoch": 3.358676691729323, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 60838 + }, + { + "epoch": 3.3587368421052632, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 60839 + }, + { + "epoch": 3.358796992481203, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 60840 + }, + { + "epoch": 3.358857142857143, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 60841 + }, + { + "epoch": 3.3589172932330826, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 60842 + }, + { + "epoch": 3.3589774436090227, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1282, + "step": 60843 + }, + { + "epoch": 3.3590375939849624, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9033, + "step": 60844 + }, + { + "epoch": 3.359097744360902, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 60845 + }, + { + "epoch": 3.359157894736842, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9393, + "step": 60846 + }, + { + "epoch": 3.3592180451127818, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 60847 + }, + { + "epoch": 3.359278195488722, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 60848 + }, + { + "epoch": 3.3593383458646615, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.1576, + "step": 60849 + }, + { + "epoch": 3.3593984962406016, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 60850 + }, + { + "epoch": 3.3594586466165413, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 60851 + }, + { + "epoch": 3.3595187969924813, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 60852 + }, + { + "epoch": 3.359578947368421, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 60853 + }, + { + "epoch": 3.359639097744361, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 60854 + }, + { + "epoch": 3.3596992481203007, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 60855 + }, + { + "epoch": 3.359759398496241, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 60856 + }, + { + "epoch": 3.3598195488721805, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0138, + "step": 60857 + }, + { + "epoch": 3.35987969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 60858 + }, + { + "epoch": 3.3599398496240602, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 60859 + }, + { + "epoch": 3.36, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9519, + "step": 60860 + }, + { + "epoch": 3.36006015037594, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0638, + "step": 60861 + }, + { + "epoch": 3.3601203007518796, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 60862 + }, + { + "epoch": 3.3601804511278197, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.1793, + "step": 60863 + }, + { + "epoch": 3.3602406015037594, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 60864 + }, + { + "epoch": 3.3603007518796995, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 60865 + }, + { + "epoch": 3.360360902255639, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 60866 + }, + { + "epoch": 3.3604210526315788, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0236, + "step": 60867 + }, + { + "epoch": 3.360481203007519, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 60868 + }, + { + "epoch": 3.3605413533834585, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 60869 + }, + { + "epoch": 3.3606015037593986, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 60870 + }, + { + "epoch": 3.3606616541353382, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 60871 + }, + { + "epoch": 3.3607218045112783, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 60872 + }, + { + "epoch": 3.360781954887218, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.024, + "step": 60873 + }, + { + "epoch": 3.360842105263158, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 60874 + }, + { + "epoch": 3.3609022556390977, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.1083, + "step": 60875 + }, + { + "epoch": 3.3609624060150374, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 60876 + }, + { + "epoch": 3.3610225563909775, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1798, + "step": 60877 + }, + { + "epoch": 3.361082706766917, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 60878 + }, + { + "epoch": 3.361142857142857, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 60879 + }, + { + "epoch": 3.361203007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 60880 + }, + { + "epoch": 3.361263157894737, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0764, + "step": 60881 + }, + { + "epoch": 3.3613233082706766, + "grad_norm": 1.3515625, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 60882 + }, + { + "epoch": 3.3613834586466167, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 60883 + }, + { + "epoch": 3.3614436090225563, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 60884 + }, + { + "epoch": 3.361503759398496, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 60885 + }, + { + "epoch": 3.361563909774436, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9309, + "step": 60886 + }, + { + "epoch": 3.361624060150376, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 60887 + }, + { + "epoch": 3.361684210526316, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 60888 + }, + { + "epoch": 3.3617443609022555, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 60889 + }, + { + "epoch": 3.3618045112781956, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 60890 + }, + { + "epoch": 3.3618646616541352, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 60891 + }, + { + "epoch": 3.3619248120300753, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0044, + "step": 60892 + }, + { + "epoch": 3.361984962406015, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9622, + "step": 60893 + }, + { + "epoch": 3.362045112781955, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9079, + "step": 60894 + }, + { + "epoch": 3.3621052631578947, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 60895 + }, + { + "epoch": 3.362165413533835, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9593, + "step": 60896 + }, + { + "epoch": 3.3622255639097745, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 60897 + }, + { + "epoch": 3.362285714285714, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 60898 + }, + { + "epoch": 3.362345864661654, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 60899 + }, + { + "epoch": 3.362406015037594, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.1755, + "step": 60900 + }, + { + "epoch": 3.362466165413534, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0667, + "step": 60901 + }, + { + "epoch": 3.3625263157894736, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9755, + "step": 60902 + }, + { + "epoch": 3.3625864661654137, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 60903 + }, + { + "epoch": 3.3626466165413533, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 60904 + }, + { + "epoch": 3.3627067669172934, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 60905 + }, + { + "epoch": 3.362766917293233, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 60906 + }, + { + "epoch": 3.3628270676691727, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 60907 + }, + { + "epoch": 3.362887218045113, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 60908 + }, + { + "epoch": 3.3629473684210525, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 60909 + }, + { + "epoch": 3.3630075187969926, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 60910 + }, + { + "epoch": 3.363067669172932, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 60911 + }, + { + "epoch": 3.3631278195488723, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9796, + "step": 60912 + }, + { + "epoch": 3.363187969924812, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 60913 + }, + { + "epoch": 3.363248120300752, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0055, + "step": 60914 + }, + { + "epoch": 3.3633082706766917, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 60915 + }, + { + "epoch": 3.3633684210526313, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 60916 + }, + { + "epoch": 3.3634285714285714, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 60917 + }, + { + "epoch": 3.3634887218045115, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0891, + "step": 60918 + }, + { + "epoch": 3.363548872180451, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 60919 + }, + { + "epoch": 3.363609022556391, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 60920 + }, + { + "epoch": 3.363669172932331, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 60921 + }, + { + "epoch": 3.3637293233082706, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 60922 + }, + { + "epoch": 3.3637894736842107, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 60923 + }, + { + "epoch": 3.3638496240601503, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 60924 + }, + { + "epoch": 3.3639097744360904, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 60925 + }, + { + "epoch": 3.36396992481203, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.1474, + "step": 60926 + }, + { + "epoch": 3.36403007518797, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 60927 + }, + { + "epoch": 3.36409022556391, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0861, + "step": 60928 + }, + { + "epoch": 3.3641503759398494, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 60929 + }, + { + "epoch": 3.3642105263157895, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 60930 + }, + { + "epoch": 3.364270676691729, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 60931 + }, + { + "epoch": 3.3643308270676693, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 60932 + }, + { + "epoch": 3.364390977443609, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 60933 + }, + { + "epoch": 3.364451127819549, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 60934 + }, + { + "epoch": 3.3645112781954887, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 60935 + }, + { + "epoch": 3.3645714285714288, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.1028, + "step": 60936 + }, + { + "epoch": 3.3646315789473684, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 60937 + }, + { + "epoch": 3.364691729323308, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.1005, + "step": 60938 + }, + { + "epoch": 3.364751879699248, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 60939 + }, + { + "epoch": 3.364812030075188, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0749, + "step": 60940 + }, + { + "epoch": 3.364872180451128, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 60941 + }, + { + "epoch": 3.3649323308270676, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 60942 + }, + { + "epoch": 3.3649924812030076, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 60943 + }, + { + "epoch": 3.3650526315789473, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 60944 + }, + { + "epoch": 3.3651127819548874, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9192, + "step": 60945 + }, + { + "epoch": 3.365172932330827, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 60946 + }, + { + "epoch": 3.3652330827067667, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.8529, + "step": 60947 + }, + { + "epoch": 3.365293233082707, + "grad_norm": 1.90625, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 60948 + }, + { + "epoch": 3.3653533834586464, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 60949 + }, + { + "epoch": 3.3654135338345865, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9525, + "step": 60950 + }, + { + "epoch": 3.365473684210526, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 60951 + }, + { + "epoch": 3.3655338345864663, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9964, + "step": 60952 + }, + { + "epoch": 3.365593984962406, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 60953 + }, + { + "epoch": 3.365654135338346, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0466, + "step": 60954 + }, + { + "epoch": 3.3657142857142857, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9816, + "step": 60955 + }, + { + "epoch": 3.3657744360902258, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 60956 + }, + { + "epoch": 3.3658345864661654, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0484, + "step": 60957 + }, + { + "epoch": 3.3658947368421055, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 60958 + }, + { + "epoch": 3.365954887218045, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 60959 + }, + { + "epoch": 3.366015037593985, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 60960 + }, + { + "epoch": 3.366075187969925, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0744, + "step": 60961 + }, + { + "epoch": 3.3661353383458645, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 60962 + }, + { + "epoch": 3.3661954887218046, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9874, + "step": 60963 + }, + { + "epoch": 3.3662556390977443, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 60964 + }, + { + "epoch": 3.3663157894736844, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9511, + "step": 60965 + }, + { + "epoch": 3.366375939849624, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9519, + "step": 60966 + }, + { + "epoch": 3.366436090225564, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9462, + "step": 60967 + }, + { + "epoch": 3.3664962406015038, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 60968 + }, + { + "epoch": 3.3665563909774434, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.8909, + "step": 60969 + }, + { + "epoch": 3.3666165413533835, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 60970 + }, + { + "epoch": 3.366676691729323, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.93, + "step": 60971 + }, + { + "epoch": 3.3667368421052632, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9164, + "step": 60972 + }, + { + "epoch": 3.366796992481203, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 60973 + }, + { + "epoch": 3.366857142857143, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.925, + "step": 60974 + }, + { + "epoch": 3.3669172932330826, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 60975 + }, + { + "epoch": 3.3669774436090227, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 60976 + }, + { + "epoch": 3.3670375939849624, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 60977 + }, + { + "epoch": 3.367097744360902, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9302, + "step": 60978 + }, + { + "epoch": 3.367157894736842, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 60979 + }, + { + "epoch": 3.3672180451127818, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 60980 + }, + { + "epoch": 3.367278195488722, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 60981 + }, + { + "epoch": 3.3673383458646615, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 60982 + }, + { + "epoch": 3.3673984962406016, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1009, + "step": 60983 + }, + { + "epoch": 3.3674586466165413, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0564, + "step": 60984 + }, + { + "epoch": 3.3675187969924814, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 60985 + }, + { + "epoch": 3.367578947368421, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 60986 + }, + { + "epoch": 3.3676390977443607, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 60987 + }, + { + "epoch": 3.3676992481203007, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 60988 + }, + { + "epoch": 3.367759398496241, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9349, + "step": 60989 + }, + { + "epoch": 3.3678195488721805, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 60990 + }, + { + "epoch": 3.36787969924812, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 60991 + }, + { + "epoch": 3.3679398496240602, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9394, + "step": 60992 + }, + { + "epoch": 3.368, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1224, + "step": 60993 + }, + { + "epoch": 3.36806015037594, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9767, + "step": 60994 + }, + { + "epoch": 3.3681203007518796, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0518, + "step": 60995 + }, + { + "epoch": 3.3681804511278197, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 60996 + }, + { + "epoch": 3.3682406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9178, + "step": 60997 + }, + { + "epoch": 3.3683007518796995, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 60998 + }, + { + "epoch": 3.368360902255639, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 60999 + }, + { + "epoch": 3.3684210526315788, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 61000 + }, + { + "epoch": 3.368481203007519, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9172, + "step": 61001 + }, + { + "epoch": 3.3685413533834585, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9297, + "step": 61002 + }, + { + "epoch": 3.3686015037593986, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 61003 + }, + { + "epoch": 3.3686616541353382, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0998, + "step": 61004 + }, + { + "epoch": 3.3687218045112783, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 61005 + }, + { + "epoch": 3.368781954887218, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 61006 + }, + { + "epoch": 3.368842105263158, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 61007 + }, + { + "epoch": 3.3689022556390977, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 61008 + }, + { + "epoch": 3.3689624060150374, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 61009 + }, + { + "epoch": 3.3690225563909775, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 61010 + }, + { + "epoch": 3.369082706766917, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0992, + "step": 61011 + }, + { + "epoch": 3.369142857142857, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.1401, + "step": 61012 + }, + { + "epoch": 3.369203007518797, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0393, + "step": 61013 + }, + { + "epoch": 3.369263157894737, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 61014 + }, + { + "epoch": 3.3693233082706766, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 61015 + }, + { + "epoch": 3.3693834586466167, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 61016 + }, + { + "epoch": 3.3694436090225564, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 61017 + }, + { + "epoch": 3.369503759398496, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 61018 + }, + { + "epoch": 3.369563909774436, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 61019 + }, + { + "epoch": 3.369624060150376, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 61020 + }, + { + "epoch": 3.369684210526316, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0165, + "step": 61021 + }, + { + "epoch": 3.3697443609022555, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 61022 + }, + { + "epoch": 3.3698045112781956, + "grad_norm": 1.75, + "learning_rate": 3e-05, + "loss": 0.9068, + "step": 61023 + }, + { + "epoch": 3.3698646616541352, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 61024 + }, + { + "epoch": 3.3699248120300753, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.046, + "step": 61025 + }, + { + "epoch": 3.369984962406015, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 61026 + }, + { + "epoch": 3.370045112781955, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 61027 + }, + { + "epoch": 3.3701052631578947, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.93, + "step": 61028 + }, + { + "epoch": 3.370165413533835, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 61029 + }, + { + "epoch": 3.3702255639097745, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1208, + "step": 61030 + }, + { + "epoch": 3.370285714285714, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 61031 + }, + { + "epoch": 3.370345864661654, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 61032 + }, + { + "epoch": 3.370406015037594, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 61033 + }, + { + "epoch": 3.370466165413534, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 61034 + }, + { + "epoch": 3.3705263157894736, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 61035 + }, + { + "epoch": 3.3705864661654137, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 61036 + }, + { + "epoch": 3.3706466165413533, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9093, + "step": 61037 + }, + { + "epoch": 3.3707067669172934, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 61038 + }, + { + "epoch": 3.370766917293233, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.874, + "step": 61039 + }, + { + "epoch": 3.3708270676691727, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 61040 + }, + { + "epoch": 3.370887218045113, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 61041 + }, + { + "epoch": 3.3709473684210525, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 61042 + }, + { + "epoch": 3.3710075187969926, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9617, + "step": 61043 + }, + { + "epoch": 3.371067669172932, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 61044 + }, + { + "epoch": 3.3711278195488723, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 61045 + }, + { + "epoch": 3.371187969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 61046 + }, + { + "epoch": 3.371248120300752, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 61047 + }, + { + "epoch": 3.3713082706766917, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 61048 + }, + { + "epoch": 3.3713684210526313, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9371, + "step": 61049 + }, + { + "epoch": 3.3714285714285714, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 61050 + }, + { + "epoch": 3.371488721804511, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 61051 + }, + { + "epoch": 3.371548872180451, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 61052 + }, + { + "epoch": 3.371609022556391, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 61053 + }, + { + "epoch": 3.371669172932331, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0777, + "step": 61054 + }, + { + "epoch": 3.3717293233082706, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 61055 + }, + { + "epoch": 3.3717894736842107, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9061, + "step": 61056 + }, + { + "epoch": 3.3718496240601503, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 61057 + }, + { + "epoch": 3.3719097744360904, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9925, + "step": 61058 + }, + { + "epoch": 3.37196992481203, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1244, + "step": 61059 + }, + { + "epoch": 3.37203007518797, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 61060 + }, + { + "epoch": 3.37209022556391, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9238, + "step": 61061 + }, + { + "epoch": 3.3721503759398495, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 61062 + }, + { + "epoch": 3.3722105263157895, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 61063 + }, + { + "epoch": 3.372270676691729, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 61064 + }, + { + "epoch": 3.3723308270676693, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 61065 + }, + { + "epoch": 3.372390977443609, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 61066 + }, + { + "epoch": 3.372451127819549, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 61067 + }, + { + "epoch": 3.3725112781954887, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 61068 + }, + { + "epoch": 3.3725714285714288, + "grad_norm": 1.1640625, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 61069 + }, + { + "epoch": 3.3726315789473684, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 61070 + }, + { + "epoch": 3.372691729323308, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0127, + "step": 61071 + }, + { + "epoch": 3.372751879699248, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 61072 + }, + { + "epoch": 3.372812030075188, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1267, + "step": 61073 + }, + { + "epoch": 3.372872180451128, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 61074 + }, + { + "epoch": 3.3729323308270676, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 61075 + }, + { + "epoch": 3.3729924812030077, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 61076 + }, + { + "epoch": 3.3730526315789473, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 61077 + }, + { + "epoch": 3.3731127819548874, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 61078 + }, + { + "epoch": 3.373172932330827, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 61079 + }, + { + "epoch": 3.3732330827067667, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 61080 + }, + { + "epoch": 3.373293233082707, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 61081 + }, + { + "epoch": 3.3733533834586464, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 61082 + }, + { + "epoch": 3.3734135338345865, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0411, + "step": 61083 + }, + { + "epoch": 3.373473684210526, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 61084 + }, + { + "epoch": 3.3735338345864663, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 61085 + }, + { + "epoch": 3.373593984962406, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0405, + "step": 61086 + }, + { + "epoch": 3.373654135338346, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0661, + "step": 61087 + }, + { + "epoch": 3.3737142857142857, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 61088 + }, + { + "epoch": 3.3737744360902253, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 61089 + }, + { + "epoch": 3.3738345864661654, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 61090 + }, + { + "epoch": 3.3738947368421055, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 61091 + }, + { + "epoch": 3.373954887218045, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.1539, + "step": 61092 + }, + { + "epoch": 3.374015037593985, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9411, + "step": 61093 + }, + { + "epoch": 3.374075187969925, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 61094 + }, + { + "epoch": 3.3741353383458645, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9384, + "step": 61095 + }, + { + "epoch": 3.3741954887218046, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 61096 + }, + { + "epoch": 3.3742556390977443, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 61097 + }, + { + "epoch": 3.3743157894736844, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 61098 + }, + { + "epoch": 3.374375939849624, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9333, + "step": 61099 + }, + { + "epoch": 3.374436090225564, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 61100 + }, + { + "epoch": 3.3744962406015038, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 61101 + }, + { + "epoch": 3.3745563909774434, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 61102 + }, + { + "epoch": 3.3746165413533835, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0275, + "step": 61103 + }, + { + "epoch": 3.374676691729323, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 61104 + }, + { + "epoch": 3.3747368421052633, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 61105 + }, + { + "epoch": 3.374796992481203, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 61106 + }, + { + "epoch": 3.374857142857143, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9805, + "step": 61107 + }, + { + "epoch": 3.3749172932330826, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9264, + "step": 61108 + }, + { + "epoch": 3.3749774436090227, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9219, + "step": 61109 + }, + { + "epoch": 3.3750375939849624, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9328, + "step": 61110 + }, + { + "epoch": 3.375097744360902, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9243, + "step": 61111 + }, + { + "epoch": 3.375157894736842, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.956, + "step": 61112 + }, + { + "epoch": 3.375218045112782, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 61113 + }, + { + "epoch": 3.375278195488722, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 61114 + }, + { + "epoch": 3.3753383458646615, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 61115 + }, + { + "epoch": 3.3753984962406016, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.956, + "step": 61116 + }, + { + "epoch": 3.3754586466165413, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 61117 + }, + { + "epoch": 3.3755187969924814, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0519, + "step": 61118 + }, + { + "epoch": 3.375578947368421, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0774, + "step": 61119 + }, + { + "epoch": 3.3756390977443607, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 61120 + }, + { + "epoch": 3.3756992481203008, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 61121 + }, + { + "epoch": 3.375759398496241, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1063, + "step": 61122 + }, + { + "epoch": 3.3758195488721805, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 61123 + }, + { + "epoch": 3.37587969924812, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 61124 + }, + { + "epoch": 3.3759398496240602, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.933, + "step": 61125 + }, + { + "epoch": 3.376, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0862, + "step": 61126 + }, + { + "epoch": 3.37606015037594, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 61127 + }, + { + "epoch": 3.3761203007518796, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 61128 + }, + { + "epoch": 3.3761804511278197, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 61129 + }, + { + "epoch": 3.3762406015037594, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9922, + "step": 61130 + }, + { + "epoch": 3.3763007518796995, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 61131 + }, + { + "epoch": 3.376360902255639, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0644, + "step": 61132 + }, + { + "epoch": 3.3764210526315788, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 61133 + }, + { + "epoch": 3.376481203007519, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1753, + "step": 61134 + }, + { + "epoch": 3.3765413533834585, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 61135 + }, + { + "epoch": 3.3766015037593986, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0773, + "step": 61136 + }, + { + "epoch": 3.3766616541353383, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9485, + "step": 61137 + }, + { + "epoch": 3.3767218045112783, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 61138 + }, + { + "epoch": 3.376781954887218, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 61139 + }, + { + "epoch": 3.376842105263158, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 61140 + }, + { + "epoch": 3.3769022556390977, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 61141 + }, + { + "epoch": 3.3769624060150374, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 61142 + }, + { + "epoch": 3.3770225563909775, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0009, + "step": 61143 + }, + { + "epoch": 3.377082706766917, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.057, + "step": 61144 + }, + { + "epoch": 3.3771428571428572, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 61145 + }, + { + "epoch": 3.377203007518797, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 61146 + }, + { + "epoch": 3.377263157894737, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.947, + "step": 61147 + }, + { + "epoch": 3.3773233082706766, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1764, + "step": 61148 + }, + { + "epoch": 3.3773834586466167, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 61149 + }, + { + "epoch": 3.3774436090225564, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0423, + "step": 61150 + }, + { + "epoch": 3.377503759398496, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 61151 + }, + { + "epoch": 3.377563909774436, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 61152 + }, + { + "epoch": 3.3776240601503758, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.941, + "step": 61153 + }, + { + "epoch": 3.377684210526316, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 61154 + }, + { + "epoch": 3.3777443609022555, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9833, + "step": 61155 + }, + { + "epoch": 3.3778045112781956, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 61156 + }, + { + "epoch": 3.3778646616541352, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 61157 + }, + { + "epoch": 3.3779248120300753, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9885, + "step": 61158 + }, + { + "epoch": 3.377984962406015, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 61159 + }, + { + "epoch": 3.378045112781955, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 61160 + }, + { + "epoch": 3.3781052631578947, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.954, + "step": 61161 + }, + { + "epoch": 3.378165413533835, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0748, + "step": 61162 + }, + { + "epoch": 3.3782255639097745, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 61163 + }, + { + "epoch": 3.378285714285714, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 61164 + }, + { + "epoch": 3.378345864661654, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 61165 + }, + { + "epoch": 3.378406015037594, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 61166 + }, + { + "epoch": 3.378466165413534, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 61167 + }, + { + "epoch": 3.3785263157894736, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 61168 + }, + { + "epoch": 3.3785864661654137, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 61169 + }, + { + "epoch": 3.3786466165413533, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 61170 + }, + { + "epoch": 3.3787067669172934, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 61171 + }, + { + "epoch": 3.378766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 61172 + }, + { + "epoch": 3.3788270676691727, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 61173 + }, + { + "epoch": 3.378887218045113, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 61174 + }, + { + "epoch": 3.3789473684210525, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 61175 + }, + { + "epoch": 3.3790075187969926, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0306, + "step": 61176 + }, + { + "epoch": 3.379067669172932, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 61177 + }, + { + "epoch": 3.3791278195488723, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 61178 + }, + { + "epoch": 3.379187969924812, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.8992, + "step": 61179 + }, + { + "epoch": 3.379248120300752, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0517, + "step": 61180 + }, + { + "epoch": 3.3793082706766917, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 61181 + }, + { + "epoch": 3.3793684210526314, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0298, + "step": 61182 + }, + { + "epoch": 3.3794285714285714, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 61183 + }, + { + "epoch": 3.379488721804511, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 61184 + }, + { + "epoch": 3.379548872180451, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.072, + "step": 61185 + }, + { + "epoch": 3.379609022556391, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.8867, + "step": 61186 + }, + { + "epoch": 3.379669172932331, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 61187 + }, + { + "epoch": 3.3797293233082706, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 61188 + }, + { + "epoch": 3.3797894736842107, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 61189 + }, + { + "epoch": 3.3798496240601503, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 61190 + }, + { + "epoch": 3.3799097744360904, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 61191 + }, + { + "epoch": 3.37996992481203, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 61192 + }, + { + "epoch": 3.38003007518797, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 61193 + }, + { + "epoch": 3.38009022556391, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9095, + "step": 61194 + }, + { + "epoch": 3.3801503759398495, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 61195 + }, + { + "epoch": 3.3802105263157896, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 61196 + }, + { + "epoch": 3.380270676691729, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.8786, + "step": 61197 + }, + { + "epoch": 3.3803308270676693, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 61198 + }, + { + "epoch": 3.380390977443609, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.1514, + "step": 61199 + }, + { + "epoch": 3.380451127819549, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 61200 + }, + { + "epoch": 3.3805112781954887, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9346, + "step": 61201 + }, + { + "epoch": 3.380571428571429, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 61202 + }, + { + "epoch": 3.3806315789473684, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 61203 + }, + { + "epoch": 3.380691729323308, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 61204 + }, + { + "epoch": 3.380751879699248, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9157, + "step": 61205 + }, + { + "epoch": 3.380812030075188, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9435, + "step": 61206 + }, + { + "epoch": 3.380872180451128, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0459, + "step": 61207 + }, + { + "epoch": 3.3809323308270676, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 61208 + }, + { + "epoch": 3.3809924812030077, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0976, + "step": 61209 + }, + { + "epoch": 3.3810526315789473, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9492, + "step": 61210 + }, + { + "epoch": 3.3811127819548874, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9987, + "step": 61211 + }, + { + "epoch": 3.381172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 61212 + }, + { + "epoch": 3.3812330827067667, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9386, + "step": 61213 + }, + { + "epoch": 3.381293233082707, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9608, + "step": 61214 + }, + { + "epoch": 3.3813533834586464, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 61215 + }, + { + "epoch": 3.3814135338345865, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 61216 + }, + { + "epoch": 3.381473684210526, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 61217 + }, + { + "epoch": 3.3815338345864663, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 61218 + }, + { + "epoch": 3.381593984962406, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 61219 + }, + { + "epoch": 3.381654135338346, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 61220 + }, + { + "epoch": 3.3817142857142857, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 61221 + }, + { + "epoch": 3.3817744360902253, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9076, + "step": 61222 + }, + { + "epoch": 3.3818345864661654, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 61223 + }, + { + "epoch": 3.3818947368421055, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9665, + "step": 61224 + }, + { + "epoch": 3.381954887218045, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 61225 + }, + { + "epoch": 3.382015037593985, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 61226 + }, + { + "epoch": 3.382075187969925, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.954, + "step": 61227 + }, + { + "epoch": 3.3821353383458646, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 61228 + }, + { + "epoch": 3.3821954887218046, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 61229 + }, + { + "epoch": 3.3822556390977443, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0514, + "step": 61230 + }, + { + "epoch": 3.3823157894736844, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 61231 + }, + { + "epoch": 3.382375939849624, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0008, + "step": 61232 + }, + { + "epoch": 3.382436090225564, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0281, + "step": 61233 + }, + { + "epoch": 3.382496240601504, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 61234 + }, + { + "epoch": 3.3825563909774434, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 61235 + }, + { + "epoch": 3.3826165413533835, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0847, + "step": 61236 + }, + { + "epoch": 3.382676691729323, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1135, + "step": 61237 + }, + { + "epoch": 3.3827368421052633, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9919, + "step": 61238 + }, + { + "epoch": 3.382796992481203, + "grad_norm": 1.296875, + "learning_rate": 3e-05, + "loss": 1.1468, + "step": 61239 + }, + { + "epoch": 3.382857142857143, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 61240 + }, + { + "epoch": 3.3829172932330827, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 61241 + }, + { + "epoch": 3.3829774436090227, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 61242 + }, + { + "epoch": 3.3830375939849624, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9707, + "step": 61243 + }, + { + "epoch": 3.383097744360902, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 61244 + }, + { + "epoch": 3.383157894736842, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 61245 + }, + { + "epoch": 3.383218045112782, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 61246 + }, + { + "epoch": 3.383278195488722, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 61247 + }, + { + "epoch": 3.3833383458646615, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0193, + "step": 61248 + }, + { + "epoch": 3.3833984962406016, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 61249 + }, + { + "epoch": 3.3834586466165413, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9555, + "step": 61250 + }, + { + "epoch": 3.3835187969924814, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9951, + "step": 61251 + }, + { + "epoch": 3.383578947368421, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9338, + "step": 61252 + }, + { + "epoch": 3.3836390977443607, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9312, + "step": 61253 + }, + { + "epoch": 3.3836992481203008, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 61254 + }, + { + "epoch": 3.3837593984962404, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 61255 + }, + { + "epoch": 3.3838195488721805, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0525, + "step": 61256 + }, + { + "epoch": 3.38387969924812, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0607, + "step": 61257 + }, + { + "epoch": 3.3839398496240602, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 61258 + }, + { + "epoch": 3.384, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9351, + "step": 61259 + }, + { + "epoch": 3.38406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.995, + "step": 61260 + }, + { + "epoch": 3.3841203007518796, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 61261 + }, + { + "epoch": 3.3841804511278197, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 61262 + }, + { + "epoch": 3.3842406015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 61263 + }, + { + "epoch": 3.3843007518796995, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 61264 + }, + { + "epoch": 3.384360902255639, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 61265 + }, + { + "epoch": 3.3844210526315788, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 61266 + }, + { + "epoch": 3.384481203007519, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 61267 + }, + { + "epoch": 3.3845413533834585, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 61268 + }, + { + "epoch": 3.3846015037593986, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 61269 + }, + { + "epoch": 3.3846616541353383, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 61270 + }, + { + "epoch": 3.3847218045112784, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 61271 + }, + { + "epoch": 3.384781954887218, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 61272 + }, + { + "epoch": 3.384842105263158, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 61273 + }, + { + "epoch": 3.3849022556390977, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 61274 + }, + { + "epoch": 3.3849624060150374, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 61275 + }, + { + "epoch": 3.3850225563909775, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 61276 + }, + { + "epoch": 3.385082706766917, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 61277 + }, + { + "epoch": 3.3851428571428572, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 61278 + }, + { + "epoch": 3.385203007518797, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.927, + "step": 61279 + }, + { + "epoch": 3.385263157894737, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9984, + "step": 61280 + }, + { + "epoch": 3.3853233082706766, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 61281 + }, + { + "epoch": 3.3853834586466167, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 61282 + }, + { + "epoch": 3.3854436090225564, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 61283 + }, + { + "epoch": 3.385503759398496, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 61284 + }, + { + "epoch": 3.385563909774436, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0883, + "step": 61285 + }, + { + "epoch": 3.3856240601503758, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 61286 + }, + { + "epoch": 3.385684210526316, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 61287 + }, + { + "epoch": 3.3857443609022555, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.952, + "step": 61288 + }, + { + "epoch": 3.3858045112781956, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 61289 + }, + { + "epoch": 3.3858646616541352, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9312, + "step": 61290 + }, + { + "epoch": 3.3859248120300753, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9425, + "step": 61291 + }, + { + "epoch": 3.385984962406015, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 61292 + }, + { + "epoch": 3.386045112781955, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.1065, + "step": 61293 + }, + { + "epoch": 3.3861052631578947, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 61294 + }, + { + "epoch": 3.386165413533835, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 61295 + }, + { + "epoch": 3.3862255639097745, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0071, + "step": 61296 + }, + { + "epoch": 3.386285714285714, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0583, + "step": 61297 + }, + { + "epoch": 3.386345864661654, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 61298 + }, + { + "epoch": 3.386406015037594, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0104, + "step": 61299 + }, + { + "epoch": 3.386466165413534, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.92, + "step": 61300 + }, + { + "epoch": 3.3865263157894736, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 61301 + }, + { + "epoch": 3.3865864661654137, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1293, + "step": 61302 + }, + { + "epoch": 3.3866466165413533, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9554, + "step": 61303 + }, + { + "epoch": 3.3867067669172934, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9531, + "step": 61304 + }, + { + "epoch": 3.386766917293233, + "grad_norm": 1.6328125, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 61305 + }, + { + "epoch": 3.3868270676691727, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 61306 + }, + { + "epoch": 3.386887218045113, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 61307 + }, + { + "epoch": 3.3869473684210525, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 61308 + }, + { + "epoch": 3.3870075187969926, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 61309 + }, + { + "epoch": 3.3870676691729322, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 61310 + }, + { + "epoch": 3.3871278195488723, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 61311 + }, + { + "epoch": 3.387187969924812, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 61312 + }, + { + "epoch": 3.387248120300752, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 61313 + }, + { + "epoch": 3.3873082706766917, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 61314 + }, + { + "epoch": 3.3873684210526314, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.3094, + "step": 61315 + }, + { + "epoch": 3.3874285714285715, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9037, + "step": 61316 + }, + { + "epoch": 3.387488721804511, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0371, + "step": 61317 + }, + { + "epoch": 3.387548872180451, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.05, + "step": 61318 + }, + { + "epoch": 3.387609022556391, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 61319 + }, + { + "epoch": 3.387669172932331, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9242, + "step": 61320 + }, + { + "epoch": 3.3877293233082706, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 61321 + }, + { + "epoch": 3.3877894736842107, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 61322 + }, + { + "epoch": 3.3878496240601503, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9751, + "step": 61323 + }, + { + "epoch": 3.38790977443609, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9552, + "step": 61324 + }, + { + "epoch": 3.38796992481203, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1359, + "step": 61325 + }, + { + "epoch": 3.38803007518797, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1067, + "step": 61326 + }, + { + "epoch": 3.38809022556391, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1058, + "step": 61327 + }, + { + "epoch": 3.3881503759398495, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 61328 + }, + { + "epoch": 3.3882105263157896, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 61329 + }, + { + "epoch": 3.388270676691729, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 61330 + }, + { + "epoch": 3.3883308270676693, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 61331 + }, + { + "epoch": 3.388390977443609, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 61332 + }, + { + "epoch": 3.388451127819549, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9487, + "step": 61333 + }, + { + "epoch": 3.3885112781954887, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 61334 + }, + { + "epoch": 3.388571428571429, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 61335 + }, + { + "epoch": 3.3886315789473684, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1002, + "step": 61336 + }, + { + "epoch": 3.388691729323308, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 61337 + }, + { + "epoch": 3.388751879699248, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 61338 + }, + { + "epoch": 3.388812030075188, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.913, + "step": 61339 + }, + { + "epoch": 3.388872180451128, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 61340 + }, + { + "epoch": 3.3889323308270676, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 61341 + }, + { + "epoch": 3.3889924812030077, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 61342 + }, + { + "epoch": 3.3890526315789473, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 61343 + }, + { + "epoch": 3.3891127819548874, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 61344 + }, + { + "epoch": 3.389172932330827, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9852, + "step": 61345 + }, + { + "epoch": 3.3892330827067667, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 61346 + }, + { + "epoch": 3.389293233082707, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 61347 + }, + { + "epoch": 3.3893533834586465, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 61348 + }, + { + "epoch": 3.3894135338345865, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 61349 + }, + { + "epoch": 3.389473684210526, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 61350 + }, + { + "epoch": 3.3895338345864663, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9519, + "step": 61351 + }, + { + "epoch": 3.389593984962406, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9909, + "step": 61352 + }, + { + "epoch": 3.389654135338346, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 61353 + }, + { + "epoch": 3.3897142857142857, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9894, + "step": 61354 + }, + { + "epoch": 3.3897744360902253, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1191, + "step": 61355 + }, + { + "epoch": 3.3898345864661654, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0253, + "step": 61356 + }, + { + "epoch": 3.389894736842105, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 61357 + }, + { + "epoch": 3.389954887218045, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9202, + "step": 61358 + }, + { + "epoch": 3.390015037593985, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 61359 + }, + { + "epoch": 3.390075187969925, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 61360 + }, + { + "epoch": 3.3901353383458646, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9383, + "step": 61361 + }, + { + "epoch": 3.3901954887218047, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 61362 + }, + { + "epoch": 3.3902556390977443, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 61363 + }, + { + "epoch": 3.3903157894736844, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 61364 + }, + { + "epoch": 3.390375939849624, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 61365 + }, + { + "epoch": 3.390436090225564, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 61366 + }, + { + "epoch": 3.390496240601504, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 61367 + }, + { + "epoch": 3.3905563909774434, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9429, + "step": 61368 + }, + { + "epoch": 3.3906165413533835, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9403, + "step": 61369 + }, + { + "epoch": 3.390676691729323, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9692, + "step": 61370 + }, + { + "epoch": 3.3907368421052633, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 61371 + }, + { + "epoch": 3.390796992481203, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 61372 + }, + { + "epoch": 3.390857142857143, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.1733, + "step": 61373 + }, + { + "epoch": 3.3909172932330827, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 61374 + }, + { + "epoch": 3.3909774436090228, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 61375 + }, + { + "epoch": 3.3910375939849624, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 61376 + }, + { + "epoch": 3.391097744360902, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 61377 + }, + { + "epoch": 3.391157894736842, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 61378 + }, + { + "epoch": 3.391218045112782, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 61379 + }, + { + "epoch": 3.391278195488722, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 61380 + }, + { + "epoch": 3.3913383458646615, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 61381 + }, + { + "epoch": 3.3913984962406016, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.977, + "step": 61382 + }, + { + "epoch": 3.3914586466165413, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 61383 + }, + { + "epoch": 3.3915187969924814, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 61384 + }, + { + "epoch": 3.391578947368421, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 61385 + }, + { + "epoch": 3.3916390977443607, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0949, + "step": 61386 + }, + { + "epoch": 3.3916992481203008, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.8885, + "step": 61387 + }, + { + "epoch": 3.3917593984962404, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 61388 + }, + { + "epoch": 3.3918195488721805, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 61389 + }, + { + "epoch": 3.39187969924812, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 61390 + }, + { + "epoch": 3.3919398496240603, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0608, + "step": 61391 + }, + { + "epoch": 3.392, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 61392 + }, + { + "epoch": 3.39206015037594, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.934, + "step": 61393 + }, + { + "epoch": 3.3921203007518796, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 61394 + }, + { + "epoch": 3.3921804511278197, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9791, + "step": 61395 + }, + { + "epoch": 3.3922406015037594, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.939, + "step": 61396 + }, + { + "epoch": 3.3923007518796995, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 61397 + }, + { + "epoch": 3.392360902255639, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 61398 + }, + { + "epoch": 3.392421052631579, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 61399 + }, + { + "epoch": 3.392481203007519, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 61400 + }, + { + "epoch": 3.3925413533834585, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9167, + "step": 61401 + }, + { + "epoch": 3.3926015037593986, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9779, + "step": 61402 + }, + { + "epoch": 3.3926616541353383, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0415, + "step": 61403 + }, + { + "epoch": 3.3927218045112784, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 61404 + }, + { + "epoch": 3.392781954887218, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0593, + "step": 61405 + }, + { + "epoch": 3.392842105263158, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 61406 + }, + { + "epoch": 3.3929022556390978, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 61407 + }, + { + "epoch": 3.3929624060150374, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 61408 + }, + { + "epoch": 3.3930225563909775, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0532, + "step": 61409 + }, + { + "epoch": 3.393082706766917, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 61410 + }, + { + "epoch": 3.3931428571428572, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 61411 + }, + { + "epoch": 3.393203007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 61412 + }, + { + "epoch": 3.393263157894737, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 61413 + }, + { + "epoch": 3.3933233082706766, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 61414 + }, + { + "epoch": 3.3933834586466167, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 61415 + }, + { + "epoch": 3.3934436090225564, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 61416 + }, + { + "epoch": 3.393503759398496, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 61417 + }, + { + "epoch": 3.393563909774436, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 61418 + }, + { + "epoch": 3.3936240601503758, + "grad_norm": 2.96875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 61419 + }, + { + "epoch": 3.393684210526316, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 61420 + }, + { + "epoch": 3.3937443609022555, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9694, + "step": 61421 + }, + { + "epoch": 3.3938045112781956, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 61422 + }, + { + "epoch": 3.3938646616541353, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0571, + "step": 61423 + }, + { + "epoch": 3.3939248120300753, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 61424 + }, + { + "epoch": 3.393984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.113, + "step": 61425 + }, + { + "epoch": 3.3940451127819546, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0596, + "step": 61426 + }, + { + "epoch": 3.3941052631578947, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 61427 + }, + { + "epoch": 3.394165413533835, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 61428 + }, + { + "epoch": 3.3942255639097745, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9631, + "step": 61429 + }, + { + "epoch": 3.394285714285714, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9254, + "step": 61430 + }, + { + "epoch": 3.394345864661654, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9395, + "step": 61431 + }, + { + "epoch": 3.394406015037594, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 61432 + }, + { + "epoch": 3.394466165413534, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 61433 + }, + { + "epoch": 3.3945263157894736, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 61434 + }, + { + "epoch": 3.3945864661654137, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 61435 + }, + { + "epoch": 3.3946466165413534, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.1142, + "step": 61436 + }, + { + "epoch": 3.3947067669172935, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 61437 + }, + { + "epoch": 3.394766917293233, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 61438 + }, + { + "epoch": 3.3948270676691727, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 61439 + }, + { + "epoch": 3.394887218045113, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 61440 + }, + { + "epoch": 3.3949473684210525, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 61441 + }, + { + "epoch": 3.3950075187969926, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 61442 + }, + { + "epoch": 3.3950676691729322, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9349, + "step": 61443 + }, + { + "epoch": 3.3951278195488723, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 61444 + }, + { + "epoch": 3.395187969924812, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 61445 + }, + { + "epoch": 3.395248120300752, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 61446 + }, + { + "epoch": 3.3953082706766917, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9464, + "step": 61447 + }, + { + "epoch": 3.3953684210526314, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 61448 + }, + { + "epoch": 3.3954285714285715, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 61449 + }, + { + "epoch": 3.395488721804511, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 61450 + }, + { + "epoch": 3.395548872180451, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 61451 + }, + { + "epoch": 3.395609022556391, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0759, + "step": 61452 + }, + { + "epoch": 3.395669172932331, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.923, + "step": 61453 + }, + { + "epoch": 3.3957293233082706, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 61454 + }, + { + "epoch": 3.3957894736842107, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9932, + "step": 61455 + }, + { + "epoch": 3.3958496240601503, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9974, + "step": 61456 + }, + { + "epoch": 3.39590977443609, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 61457 + }, + { + "epoch": 3.39596992481203, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 61458 + }, + { + "epoch": 3.3960300751879697, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 61459 + }, + { + "epoch": 3.39609022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 61460 + }, + { + "epoch": 3.3961503759398495, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 61461 + }, + { + "epoch": 3.3962105263157896, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 61462 + }, + { + "epoch": 3.396270676691729, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 61463 + }, + { + "epoch": 3.3963308270676693, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.8876, + "step": 61464 + }, + { + "epoch": 3.396390977443609, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 61465 + }, + { + "epoch": 3.396451127819549, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0366, + "step": 61466 + }, + { + "epoch": 3.3965112781954887, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0435, + "step": 61467 + }, + { + "epoch": 3.396571428571429, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 61468 + }, + { + "epoch": 3.3966315789473684, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9282, + "step": 61469 + }, + { + "epoch": 3.396691729323308, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9347, + "step": 61470 + }, + { + "epoch": 3.396751879699248, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 61471 + }, + { + "epoch": 3.396812030075188, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 61472 + }, + { + "epoch": 3.396872180451128, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0651, + "step": 61473 + }, + { + "epoch": 3.3969323308270676, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 61474 + }, + { + "epoch": 3.3969924812030077, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 61475 + }, + { + "epoch": 3.3970526315789473, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9721, + "step": 61476 + }, + { + "epoch": 3.3971127819548874, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 61477 + }, + { + "epoch": 3.397172932330827, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 61478 + }, + { + "epoch": 3.3972330827067667, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 61479 + }, + { + "epoch": 3.397293233082707, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 61480 + }, + { + "epoch": 3.3973533834586465, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.2893, + "step": 61481 + }, + { + "epoch": 3.3974135338345866, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 61482 + }, + { + "epoch": 3.397473684210526, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 61483 + }, + { + "epoch": 3.3975338345864663, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 61484 + }, + { + "epoch": 3.397593984962406, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9234, + "step": 61485 + }, + { + "epoch": 3.397654135338346, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9123, + "step": 61486 + }, + { + "epoch": 3.3977142857142857, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 61487 + }, + { + "epoch": 3.3977744360902253, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 61488 + }, + { + "epoch": 3.3978345864661654, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 61489 + }, + { + "epoch": 3.397894736842105, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 61490 + }, + { + "epoch": 3.397954887218045, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.1263, + "step": 61491 + }, + { + "epoch": 3.398015037593985, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 61492 + }, + { + "epoch": 3.398075187969925, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 61493 + }, + { + "epoch": 3.3981353383458646, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 61494 + }, + { + "epoch": 3.3981954887218047, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 61495 + }, + { + "epoch": 3.3982556390977443, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9536, + "step": 61496 + }, + { + "epoch": 3.3983157894736844, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 61497 + }, + { + "epoch": 3.398375939849624, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 61498 + }, + { + "epoch": 3.398436090225564, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 61499 + }, + { + "epoch": 3.398496240601504, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.054, + "step": 61500 + }, + { + "epoch": 3.3985563909774434, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 61501 + }, + { + "epoch": 3.3986165413533835, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 61502 + }, + { + "epoch": 3.398676691729323, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 61503 + }, + { + "epoch": 3.3987368421052633, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.902, + "step": 61504 + }, + { + "epoch": 3.398796992481203, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9454, + "step": 61505 + }, + { + "epoch": 3.398857142857143, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 61506 + }, + { + "epoch": 3.3989172932330827, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 61507 + }, + { + "epoch": 3.3989774436090228, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 61508 + }, + { + "epoch": 3.3990375939849624, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 61509 + }, + { + "epoch": 3.399097744360902, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 61510 + }, + { + "epoch": 3.399157894736842, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 61511 + }, + { + "epoch": 3.399218045112782, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 61512 + }, + { + "epoch": 3.399278195488722, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 61513 + }, + { + "epoch": 3.3993383458646615, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 61514 + }, + { + "epoch": 3.3993984962406016, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 61515 + }, + { + "epoch": 3.3994586466165413, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 61516 + }, + { + "epoch": 3.3995187969924814, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 61517 + }, + { + "epoch": 3.399578947368421, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 61518 + }, + { + "epoch": 3.3996390977443607, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9498, + "step": 61519 + }, + { + "epoch": 3.3996992481203008, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 61520 + }, + { + "epoch": 3.3997593984962404, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 61521 + }, + { + "epoch": 3.3998195488721805, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 61522 + }, + { + "epoch": 3.39987969924812, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 61523 + }, + { + "epoch": 3.3999398496240603, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 61524 + }, + { + "epoch": 3.4, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 61525 + }, + { + "epoch": 3.40006015037594, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0065, + "step": 61526 + }, + { + "epoch": 3.4001203007518797, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 61527 + }, + { + "epoch": 3.4001804511278193, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 61528 + }, + { + "epoch": 3.4002406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9384, + "step": 61529 + }, + { + "epoch": 3.4003007518796995, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 61530 + }, + { + "epoch": 3.400360902255639, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 61531 + }, + { + "epoch": 3.400421052631579, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 61532 + }, + { + "epoch": 3.400481203007519, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 61533 + }, + { + "epoch": 3.4005413533834585, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0537, + "step": 61534 + }, + { + "epoch": 3.4006015037593986, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 61535 + }, + { + "epoch": 3.4006616541353383, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 61536 + }, + { + "epoch": 3.4007218045112784, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0274, + "step": 61537 + }, + { + "epoch": 3.400781954887218, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 61538 + }, + { + "epoch": 3.400842105263158, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 61539 + }, + { + "epoch": 3.4009022556390978, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9503, + "step": 61540 + }, + { + "epoch": 3.4009624060150374, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 61541 + }, + { + "epoch": 3.4010225563909775, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2134, + "step": 61542 + }, + { + "epoch": 3.401082706766917, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 61543 + }, + { + "epoch": 3.4011428571428572, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1345, + "step": 61544 + }, + { + "epoch": 3.401203007518797, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9451, + "step": 61545 + }, + { + "epoch": 3.401263157894737, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 61546 + }, + { + "epoch": 3.4013233082706766, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 61547 + }, + { + "epoch": 3.4013834586466167, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 61548 + }, + { + "epoch": 3.4014436090225564, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 61549 + }, + { + "epoch": 3.401503759398496, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9374, + "step": 61550 + }, + { + "epoch": 3.401563909774436, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 61551 + }, + { + "epoch": 3.4016240601503758, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 61552 + }, + { + "epoch": 3.401684210526316, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9506, + "step": 61553 + }, + { + "epoch": 3.4017443609022555, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 61554 + }, + { + "epoch": 3.4018045112781956, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 61555 + }, + { + "epoch": 3.4018646616541353, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9239, + "step": 61556 + }, + { + "epoch": 3.4019248120300754, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 61557 + }, + { + "epoch": 3.401984962406015, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0335, + "step": 61558 + }, + { + "epoch": 3.4020451127819546, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 61559 + }, + { + "epoch": 3.4021052631578947, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 61560 + }, + { + "epoch": 3.4021654135338344, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0966, + "step": 61561 + }, + { + "epoch": 3.4022255639097745, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0426, + "step": 61562 + }, + { + "epoch": 3.402285714285714, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 61563 + }, + { + "epoch": 3.4023458646616542, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 61564 + }, + { + "epoch": 3.402406015037594, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 61565 + }, + { + "epoch": 3.402466165413534, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 61566 + }, + { + "epoch": 3.4025263157894736, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0507, + "step": 61567 + }, + { + "epoch": 3.4025864661654137, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 61568 + }, + { + "epoch": 3.4026466165413534, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9408, + "step": 61569 + }, + { + "epoch": 3.4027067669172935, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.004, + "step": 61570 + }, + { + "epoch": 3.402766917293233, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 61571 + }, + { + "epoch": 3.4028270676691728, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 61572 + }, + { + "epoch": 3.402887218045113, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 61573 + }, + { + "epoch": 3.4029473684210525, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 61574 + }, + { + "epoch": 3.4030075187969926, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 61575 + }, + { + "epoch": 3.4030676691729322, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9254, + "step": 61576 + }, + { + "epoch": 3.4031278195488723, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 61577 + }, + { + "epoch": 3.403187969924812, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.968, + "step": 61578 + }, + { + "epoch": 3.403248120300752, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 61579 + }, + { + "epoch": 3.4033082706766917, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9259, + "step": 61580 + }, + { + "epoch": 3.4033684210526314, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 61581 + }, + { + "epoch": 3.4034285714285715, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9505, + "step": 61582 + }, + { + "epoch": 3.403488721804511, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9289, + "step": 61583 + }, + { + "epoch": 3.403548872180451, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 61584 + }, + { + "epoch": 3.403609022556391, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 61585 + }, + { + "epoch": 3.403669172932331, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 61586 + }, + { + "epoch": 3.4037293233082706, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0375, + "step": 61587 + }, + { + "epoch": 3.4037894736842107, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 61588 + }, + { + "epoch": 3.4038496240601503, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 61589 + }, + { + "epoch": 3.40390977443609, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 61590 + }, + { + "epoch": 3.40396992481203, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 61591 + }, + { + "epoch": 3.4040300751879697, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 61592 + }, + { + "epoch": 3.40409022556391, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 61593 + }, + { + "epoch": 3.4041503759398495, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0672, + "step": 61594 + }, + { + "epoch": 3.4042105263157896, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0677, + "step": 61595 + }, + { + "epoch": 3.4042706766917292, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 61596 + }, + { + "epoch": 3.4043308270676693, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 61597 + }, + { + "epoch": 3.404390977443609, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0117, + "step": 61598 + }, + { + "epoch": 3.404451127819549, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 61599 + }, + { + "epoch": 3.4045112781954887, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 61600 + }, + { + "epoch": 3.404571428571429, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9501, + "step": 61601 + }, + { + "epoch": 3.4046315789473685, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1258, + "step": 61602 + }, + { + "epoch": 3.404691729323308, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 61603 + }, + { + "epoch": 3.404751879699248, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9835, + "step": 61604 + }, + { + "epoch": 3.404812030075188, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 61605 + }, + { + "epoch": 3.404872180451128, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 61606 + }, + { + "epoch": 3.4049323308270676, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 61607 + }, + { + "epoch": 3.4049924812030077, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 61608 + }, + { + "epoch": 3.4050526315789473, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 61609 + }, + { + "epoch": 3.4051127819548874, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 61610 + }, + { + "epoch": 3.405172932330827, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0736, + "step": 61611 + }, + { + "epoch": 3.4052330827067667, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.098, + "step": 61612 + }, + { + "epoch": 3.405293233082707, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9888, + "step": 61613 + }, + { + "epoch": 3.4053533834586465, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 61614 + }, + { + "epoch": 3.4054135338345866, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0043, + "step": 61615 + }, + { + "epoch": 3.405473684210526, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9469, + "step": 61616 + }, + { + "epoch": 3.4055338345864663, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 61617 + }, + { + "epoch": 3.405593984962406, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0232, + "step": 61618 + }, + { + "epoch": 3.405654135338346, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9036, + "step": 61619 + }, + { + "epoch": 3.4057142857142857, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0572, + "step": 61620 + }, + { + "epoch": 3.4057744360902253, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 61621 + }, + { + "epoch": 3.4058345864661654, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0504, + "step": 61622 + }, + { + "epoch": 3.405894736842105, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 61623 + }, + { + "epoch": 3.405954887218045, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9251, + "step": 61624 + }, + { + "epoch": 3.406015037593985, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0036, + "step": 61625 + }, + { + "epoch": 3.406075187969925, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 61626 + }, + { + "epoch": 3.4061353383458646, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 0.8195, + "step": 61627 + }, + { + "epoch": 3.4061954887218047, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0662, + "step": 61628 + }, + { + "epoch": 3.4062556390977443, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9187, + "step": 61629 + }, + { + "epoch": 3.406315789473684, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0985, + "step": 61630 + }, + { + "epoch": 3.406375939849624, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 61631 + }, + { + "epoch": 3.406436090225564, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 61632 + }, + { + "epoch": 3.406496240601504, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9522, + "step": 61633 + }, + { + "epoch": 3.4065563909774434, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 61634 + }, + { + "epoch": 3.4066165413533835, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.2136, + "step": 61635 + }, + { + "epoch": 3.406676691729323, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 61636 + }, + { + "epoch": 3.4067368421052633, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 61637 + }, + { + "epoch": 3.406796992481203, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9276, + "step": 61638 + }, + { + "epoch": 3.406857142857143, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 61639 + }, + { + "epoch": 3.4069172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 61640 + }, + { + "epoch": 3.4069774436090228, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 61641 + }, + { + "epoch": 3.4070375939849624, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0327, + "step": 61642 + }, + { + "epoch": 3.407097744360902, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 61643 + }, + { + "epoch": 3.407157894736842, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 61644 + }, + { + "epoch": 3.407218045112782, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 61645 + }, + { + "epoch": 3.407278195488722, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9272, + "step": 61646 + }, + { + "epoch": 3.4073383458646616, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 61647 + }, + { + "epoch": 3.4073984962406016, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 61648 + }, + { + "epoch": 3.4074586466165413, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.941, + "step": 61649 + }, + { + "epoch": 3.4075187969924814, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 61650 + }, + { + "epoch": 3.407578947368421, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 61651 + }, + { + "epoch": 3.4076390977443607, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 61652 + }, + { + "epoch": 3.407699248120301, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0649, + "step": 61653 + }, + { + "epoch": 3.4077593984962404, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 61654 + }, + { + "epoch": 3.4078195488721805, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 61655 + }, + { + "epoch": 3.40787969924812, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 61656 + }, + { + "epoch": 3.4079398496240603, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 61657 + }, + { + "epoch": 3.408, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9531, + "step": 61658 + }, + { + "epoch": 3.40806015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 61659 + }, + { + "epoch": 3.4081203007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 61660 + }, + { + "epoch": 3.4081804511278193, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1105, + "step": 61661 + }, + { + "epoch": 3.4082406015037594, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0192, + "step": 61662 + }, + { + "epoch": 3.4083007518796995, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 61663 + }, + { + "epoch": 3.408360902255639, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.036, + "step": 61664 + }, + { + "epoch": 3.408421052631579, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0333, + "step": 61665 + }, + { + "epoch": 3.408481203007519, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 61666 + }, + { + "epoch": 3.4085413533834585, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 61667 + }, + { + "epoch": 3.4086015037593986, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9389, + "step": 61668 + }, + { + "epoch": 3.4086616541353383, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 61669 + }, + { + "epoch": 3.4087218045112784, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 61670 + }, + { + "epoch": 3.408781954887218, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 61671 + }, + { + "epoch": 3.408842105263158, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 61672 + }, + { + "epoch": 3.4089022556390978, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9616, + "step": 61673 + }, + { + "epoch": 3.4089624060150374, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1085, + "step": 61674 + }, + { + "epoch": 3.4090225563909775, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0105, + "step": 61675 + }, + { + "epoch": 3.409082706766917, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 61676 + }, + { + "epoch": 3.4091428571428573, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 61677 + }, + { + "epoch": 3.409203007518797, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 61678 + }, + { + "epoch": 3.409263157894737, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 61679 + }, + { + "epoch": 3.4093233082706766, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 61680 + }, + { + "epoch": 3.4093834586466167, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 61681 + }, + { + "epoch": 3.4094436090225564, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 61682 + }, + { + "epoch": 3.409503759398496, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0778, + "step": 61683 + }, + { + "epoch": 3.409563909774436, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9865, + "step": 61684 + }, + { + "epoch": 3.409624060150376, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9268, + "step": 61685 + }, + { + "epoch": 3.409684210526316, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 61686 + }, + { + "epoch": 3.4097443609022555, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.8852, + "step": 61687 + }, + { + "epoch": 3.4098045112781956, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0131, + "step": 61688 + }, + { + "epoch": 3.4098646616541353, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 61689 + }, + { + "epoch": 3.4099248120300754, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 61690 + }, + { + "epoch": 3.409984962406015, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9293, + "step": 61691 + }, + { + "epoch": 3.4100451127819547, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9071, + "step": 61692 + }, + { + "epoch": 3.4101052631578948, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 61693 + }, + { + "epoch": 3.4101654135338344, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 61694 + }, + { + "epoch": 3.4102255639097745, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 61695 + }, + { + "epoch": 3.410285714285714, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 61696 + }, + { + "epoch": 3.4103458646616542, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 61697 + }, + { + "epoch": 3.410406015037594, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 61698 + }, + { + "epoch": 3.410466165413534, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 61699 + }, + { + "epoch": 3.4105263157894736, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 61700 + }, + { + "epoch": 3.4105864661654137, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 61701 + }, + { + "epoch": 3.4106466165413534, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0267, + "step": 61702 + }, + { + "epoch": 3.4107067669172935, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1316, + "step": 61703 + }, + { + "epoch": 3.410766917293233, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 61704 + }, + { + "epoch": 3.4108270676691728, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 61705 + }, + { + "epoch": 3.410887218045113, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 61706 + }, + { + "epoch": 3.4109473684210525, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.996, + "step": 61707 + }, + { + "epoch": 3.4110075187969926, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 61708 + }, + { + "epoch": 3.4110676691729322, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.1361, + "step": 61709 + }, + { + "epoch": 3.4111278195488723, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9451, + "step": 61710 + }, + { + "epoch": 3.411187969924812, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.8988, + "step": 61711 + }, + { + "epoch": 3.411248120300752, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 61712 + }, + { + "epoch": 3.4113082706766917, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 61713 + }, + { + "epoch": 3.4113684210526314, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 61714 + }, + { + "epoch": 3.4114285714285715, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 61715 + }, + { + "epoch": 3.411488721804511, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 61716 + }, + { + "epoch": 3.411548872180451, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 61717 + }, + { + "epoch": 3.411609022556391, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 61718 + }, + { + "epoch": 3.411669172932331, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0628, + "step": 61719 + }, + { + "epoch": 3.4117293233082706, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0046, + "step": 61720 + }, + { + "epoch": 3.4117894736842107, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 61721 + }, + { + "epoch": 3.4118496240601504, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 61722 + }, + { + "epoch": 3.41190977443609, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0841, + "step": 61723 + }, + { + "epoch": 3.41196992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 61724 + }, + { + "epoch": 3.4120300751879697, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 61725 + }, + { + "epoch": 3.41209022556391, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9212, + "step": 61726 + }, + { + "epoch": 3.4121503759398495, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0313, + "step": 61727 + }, + { + "epoch": 3.4122105263157896, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 61728 + }, + { + "epoch": 3.4122706766917292, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 61729 + }, + { + "epoch": 3.4123308270676693, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 61730 + }, + { + "epoch": 3.412390977443609, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 61731 + }, + { + "epoch": 3.4124511278195486, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0765, + "step": 61732 + }, + { + "epoch": 3.4125112781954887, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 61733 + }, + { + "epoch": 3.412571428571429, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 61734 + }, + { + "epoch": 3.4126315789473685, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 61735 + }, + { + "epoch": 3.412691729323308, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 61736 + }, + { + "epoch": 3.412751879699248, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 61737 + }, + { + "epoch": 3.412812030075188, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 61738 + }, + { + "epoch": 3.412872180451128, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 61739 + }, + { + "epoch": 3.4129323308270676, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 61740 + }, + { + "epoch": 3.4129924812030077, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9679, + "step": 61741 + }, + { + "epoch": 3.4130526315789473, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9837, + "step": 61742 + }, + { + "epoch": 3.4131127819548874, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 61743 + }, + { + "epoch": 3.413172932330827, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 61744 + }, + { + "epoch": 3.4132330827067667, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0944, + "step": 61745 + }, + { + "epoch": 3.413293233082707, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 61746 + }, + { + "epoch": 3.4133533834586465, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.936, + "step": 61747 + }, + { + "epoch": 3.4134135338345866, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9784, + "step": 61748 + }, + { + "epoch": 3.413473684210526, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9988, + "step": 61749 + }, + { + "epoch": 3.4135338345864663, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 61750 + }, + { + "epoch": 3.413593984962406, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0929, + "step": 61751 + }, + { + "epoch": 3.413654135338346, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9544, + "step": 61752 + }, + { + "epoch": 3.4137142857142857, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 61753 + }, + { + "epoch": 3.4137744360902254, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0157, + "step": 61754 + }, + { + "epoch": 3.4138345864661654, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0271, + "step": 61755 + }, + { + "epoch": 3.413894736842105, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 61756 + }, + { + "epoch": 3.413954887218045, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 61757 + }, + { + "epoch": 3.414015037593985, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9402, + "step": 61758 + }, + { + "epoch": 3.414075187969925, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0911, + "step": 61759 + }, + { + "epoch": 3.4141353383458646, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 61760 + }, + { + "epoch": 3.4141954887218047, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 61761 + }, + { + "epoch": 3.4142556390977443, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 61762 + }, + { + "epoch": 3.414315789473684, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 61763 + }, + { + "epoch": 3.414375939849624, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9529, + "step": 61764 + }, + { + "epoch": 3.414436090225564, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 61765 + }, + { + "epoch": 3.414496240601504, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1416, + "step": 61766 + }, + { + "epoch": 3.4145563909774435, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9242, + "step": 61767 + }, + { + "epoch": 3.4146165413533835, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 61768 + }, + { + "epoch": 3.414676691729323, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 61769 + }, + { + "epoch": 3.4147368421052633, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.8746, + "step": 61770 + }, + { + "epoch": 3.414796992481203, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 61771 + }, + { + "epoch": 3.414857142857143, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.1669, + "step": 61772 + }, + { + "epoch": 3.4149172932330827, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 61773 + }, + { + "epoch": 3.414977443609023, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0314, + "step": 61774 + }, + { + "epoch": 3.4150375939849624, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.007, + "step": 61775 + }, + { + "epoch": 3.415097744360902, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1186, + "step": 61776 + }, + { + "epoch": 3.415157894736842, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 61777 + }, + { + "epoch": 3.415218045112782, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 61778 + }, + { + "epoch": 3.415278195488722, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 61779 + }, + { + "epoch": 3.4153383458646616, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9383, + "step": 61780 + }, + { + "epoch": 3.4153984962406017, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0383, + "step": 61781 + }, + { + "epoch": 3.4154586466165413, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9801, + "step": 61782 + }, + { + "epoch": 3.4155187969924814, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0188, + "step": 61783 + }, + { + "epoch": 3.415578947368421, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 61784 + }, + { + "epoch": 3.4156390977443607, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1866, + "step": 61785 + }, + { + "epoch": 3.415699248120301, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 61786 + }, + { + "epoch": 3.4157593984962404, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 61787 + }, + { + "epoch": 3.4158195488721805, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9349, + "step": 61788 + }, + { + "epoch": 3.41587969924812, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 61789 + }, + { + "epoch": 3.4159398496240603, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0925, + "step": 61790 + }, + { + "epoch": 3.416, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.8964, + "step": 61791 + }, + { + "epoch": 3.41606015037594, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0295, + "step": 61792 + }, + { + "epoch": 3.4161203007518797, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 61793 + }, + { + "epoch": 3.4161804511278193, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.8972, + "step": 61794 + }, + { + "epoch": 3.4162406015037594, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0816, + "step": 61795 + }, + { + "epoch": 3.416300751879699, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 61796 + }, + { + "epoch": 3.416360902255639, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 61797 + }, + { + "epoch": 3.416421052631579, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0889, + "step": 61798 + }, + { + "epoch": 3.416481203007519, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 61799 + }, + { + "epoch": 3.4165413533834585, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9104, + "step": 61800 + }, + { + "epoch": 3.4166015037593986, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9723, + "step": 61801 + }, + { + "epoch": 3.4166616541353383, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.053, + "step": 61802 + }, + { + "epoch": 3.4167218045112784, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 61803 + }, + { + "epoch": 3.416781954887218, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1021, + "step": 61804 + }, + { + "epoch": 3.416842105263158, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9061, + "step": 61805 + }, + { + "epoch": 3.4169022556390978, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 61806 + }, + { + "epoch": 3.4169624060150374, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9591, + "step": 61807 + }, + { + "epoch": 3.4170225563909775, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0339, + "step": 61808 + }, + { + "epoch": 3.417082706766917, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 61809 + }, + { + "epoch": 3.4171428571428573, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0567, + "step": 61810 + }, + { + "epoch": 3.417203007518797, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 61811 + }, + { + "epoch": 3.417263157894737, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 61812 + }, + { + "epoch": 3.4173233082706767, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 61813 + }, + { + "epoch": 3.4173834586466167, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 61814 + }, + { + "epoch": 3.4174436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 61815 + }, + { + "epoch": 3.417503759398496, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 61816 + }, + { + "epoch": 3.417563909774436, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 61817 + }, + { + "epoch": 3.417624060150376, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0221, + "step": 61818 + }, + { + "epoch": 3.417684210526316, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 61819 + }, + { + "epoch": 3.4177443609022555, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9101, + "step": 61820 + }, + { + "epoch": 3.4178045112781956, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0781, + "step": 61821 + }, + { + "epoch": 3.4178646616541353, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 61822 + }, + { + "epoch": 3.4179248120300754, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 61823 + }, + { + "epoch": 3.417984962406015, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 61824 + }, + { + "epoch": 3.4180451127819547, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 61825 + }, + { + "epoch": 3.4181052631578948, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 61826 + }, + { + "epoch": 3.4181654135338344, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 61827 + }, + { + "epoch": 3.4182255639097745, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9408, + "step": 61828 + }, + { + "epoch": 3.418285714285714, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.183, + "step": 61829 + }, + { + "epoch": 3.4183458646616542, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 61830 + }, + { + "epoch": 3.418406015037594, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 61831 + }, + { + "epoch": 3.418466165413534, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0711, + "step": 61832 + }, + { + "epoch": 3.4185263157894736, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9066, + "step": 61833 + }, + { + "epoch": 3.4185864661654133, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 61834 + }, + { + "epoch": 3.4186466165413534, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1068, + "step": 61835 + }, + { + "epoch": 3.4187067669172935, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 61836 + }, + { + "epoch": 3.418766917293233, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.956, + "step": 61837 + }, + { + "epoch": 3.4188270676691728, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 61838 + }, + { + "epoch": 3.418887218045113, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 61839 + }, + { + "epoch": 3.4189473684210525, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 61840 + }, + { + "epoch": 3.4190075187969926, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0586, + "step": 61841 + }, + { + "epoch": 3.4190676691729323, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0039, + "step": 61842 + }, + { + "epoch": 3.4191278195488723, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9428, + "step": 61843 + }, + { + "epoch": 3.419187969924812, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9158, + "step": 61844 + }, + { + "epoch": 3.419248120300752, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0321, + "step": 61845 + }, + { + "epoch": 3.4193082706766917, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9772, + "step": 61846 + }, + { + "epoch": 3.4193684210526314, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 61847 + }, + { + "epoch": 3.4194285714285715, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 61848 + }, + { + "epoch": 3.419488721804511, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 61849 + }, + { + "epoch": 3.4195488721804512, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 61850 + }, + { + "epoch": 3.419609022556391, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9792, + "step": 61851 + }, + { + "epoch": 3.419669172932331, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 61852 + }, + { + "epoch": 3.4197293233082706, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 61853 + }, + { + "epoch": 3.4197894736842107, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 61854 + }, + { + "epoch": 3.4198496240601504, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 61855 + }, + { + "epoch": 3.41990977443609, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9802, + "step": 61856 + }, + { + "epoch": 3.41996992481203, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1079, + "step": 61857 + }, + { + "epoch": 3.4200300751879698, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9573, + "step": 61858 + }, + { + "epoch": 3.42009022556391, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 61859 + }, + { + "epoch": 3.4201503759398495, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 61860 + }, + { + "epoch": 3.4202105263157896, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 61861 + }, + { + "epoch": 3.4202706766917292, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0476, + "step": 61862 + }, + { + "epoch": 3.4203308270676693, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 61863 + }, + { + "epoch": 3.420390977443609, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.1649, + "step": 61864 + }, + { + "epoch": 3.4204511278195486, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9186, + "step": 61865 + }, + { + "epoch": 3.4205112781954887, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 61866 + }, + { + "epoch": 3.420571428571429, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.077, + "step": 61867 + }, + { + "epoch": 3.4206315789473685, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 0.9157, + "step": 61868 + }, + { + "epoch": 3.420691729323308, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 61869 + }, + { + "epoch": 3.420751879699248, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 61870 + }, + { + "epoch": 3.420812030075188, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 61871 + }, + { + "epoch": 3.420872180451128, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9778, + "step": 61872 + }, + { + "epoch": 3.4209323308270676, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9719, + "step": 61873 + }, + { + "epoch": 3.4209924812030077, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 61874 + }, + { + "epoch": 3.4210526315789473, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0204, + "step": 61875 + }, + { + "epoch": 3.4211127819548874, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 61876 + }, + { + "epoch": 3.421172932330827, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 61877 + }, + { + "epoch": 3.4212330827067667, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9684, + "step": 61878 + }, + { + "epoch": 3.421293233082707, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0541, + "step": 61879 + }, + { + "epoch": 3.4213533834586465, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 61880 + }, + { + "epoch": 3.4214135338345866, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.002, + "step": 61881 + }, + { + "epoch": 3.4214736842105262, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.059, + "step": 61882 + }, + { + "epoch": 3.4215338345864663, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 61883 + }, + { + "epoch": 3.421593984962406, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 61884 + }, + { + "epoch": 3.421654135338346, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0206, + "step": 61885 + }, + { + "epoch": 3.4217142857142857, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0512, + "step": 61886 + }, + { + "epoch": 3.4217744360902254, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 61887 + }, + { + "epoch": 3.4218345864661655, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 61888 + }, + { + "epoch": 3.421894736842105, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0601, + "step": 61889 + }, + { + "epoch": 3.421954887218045, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0501, + "step": 61890 + }, + { + "epoch": 3.422015037593985, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 61891 + }, + { + "epoch": 3.422075187969925, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1374, + "step": 61892 + }, + { + "epoch": 3.4221353383458646, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 61893 + }, + { + "epoch": 3.4221954887218047, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0329, + "step": 61894 + }, + { + "epoch": 3.4222556390977443, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1692, + "step": 61895 + }, + { + "epoch": 3.422315789473684, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.021, + "step": 61896 + }, + { + "epoch": 3.422375939849624, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9799, + "step": 61897 + }, + { + "epoch": 3.4224360902255637, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 61898 + }, + { + "epoch": 3.422496240601504, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9384, + "step": 61899 + }, + { + "epoch": 3.4225563909774435, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0785, + "step": 61900 + }, + { + "epoch": 3.4226165413533836, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.066, + "step": 61901 + }, + { + "epoch": 3.422676691729323, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 61902 + }, + { + "epoch": 3.4227368421052633, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 61903 + }, + { + "epoch": 3.422796992481203, + "grad_norm": 1.0859375, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 61904 + }, + { + "epoch": 3.422857142857143, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1471, + "step": 61905 + }, + { + "epoch": 3.4229172932330827, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9426, + "step": 61906 + }, + { + "epoch": 3.422977443609023, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1617, + "step": 61907 + }, + { + "epoch": 3.4230375939849624, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 61908 + }, + { + "epoch": 3.423097744360902, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9986, + "step": 61909 + }, + { + "epoch": 3.423157894736842, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1187, + "step": 61910 + }, + { + "epoch": 3.423218045112782, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 61911 + }, + { + "epoch": 3.423278195488722, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 61912 + }, + { + "epoch": 3.4233383458646616, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 61913 + }, + { + "epoch": 3.4233984962406017, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0826, + "step": 61914 + }, + { + "epoch": 3.4234586466165413, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 61915 + }, + { + "epoch": 3.4235187969924814, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0363, + "step": 61916 + }, + { + "epoch": 3.423578947368421, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 61917 + }, + { + "epoch": 3.4236390977443607, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 61918 + }, + { + "epoch": 3.423699248120301, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9876, + "step": 61919 + }, + { + "epoch": 3.4237593984962404, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.912, + "step": 61920 + }, + { + "epoch": 3.4238195488721805, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1843, + "step": 61921 + }, + { + "epoch": 3.42387969924812, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0828, + "step": 61922 + }, + { + "epoch": 3.4239398496240603, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 61923 + }, + { + "epoch": 3.424, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 61924 + }, + { + "epoch": 3.42406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9596, + "step": 61925 + }, + { + "epoch": 3.4241203007518797, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 61926 + }, + { + "epoch": 3.4241804511278193, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0511, + "step": 61927 + }, + { + "epoch": 3.4242406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 61928 + }, + { + "epoch": 3.424300751879699, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 61929 + }, + { + "epoch": 3.424360902255639, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 61930 + }, + { + "epoch": 3.424421052631579, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 61931 + }, + { + "epoch": 3.424481203007519, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0076, + "step": 61932 + }, + { + "epoch": 3.4245413533834586, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9416, + "step": 61933 + }, + { + "epoch": 3.4246015037593986, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 61934 + }, + { + "epoch": 3.4246616541353383, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 61935 + }, + { + "epoch": 3.4247218045112784, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 61936 + }, + { + "epoch": 3.424781954887218, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.1023, + "step": 61937 + }, + { + "epoch": 3.424842105263158, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9916, + "step": 61938 + }, + { + "epoch": 3.424902255639098, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 61939 + }, + { + "epoch": 3.4249624060150374, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9385, + "step": 61940 + }, + { + "epoch": 3.4250225563909775, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 61941 + }, + { + "epoch": 3.425082706766917, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9556, + "step": 61942 + }, + { + "epoch": 3.4251428571428573, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 61943 + }, + { + "epoch": 3.425203007518797, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 61944 + }, + { + "epoch": 3.425263157894737, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.9124, + "step": 61945 + }, + { + "epoch": 3.4253233082706767, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 61946 + }, + { + "epoch": 3.4253834586466168, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 61947 + }, + { + "epoch": 3.4254436090225564, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 61948 + }, + { + "epoch": 3.425503759398496, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 61949 + }, + { + "epoch": 3.425563909774436, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 61950 + }, + { + "epoch": 3.425624060150376, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 61951 + }, + { + "epoch": 3.425684210526316, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.948, + "step": 61952 + }, + { + "epoch": 3.4257443609022555, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0612, + "step": 61953 + }, + { + "epoch": 3.4258045112781956, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.0717, + "step": 61954 + }, + { + "epoch": 3.4258646616541353, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 61955 + }, + { + "epoch": 3.4259248120300754, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0643, + "step": 61956 + }, + { + "epoch": 3.425984962406015, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 61957 + }, + { + "epoch": 3.4260451127819547, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 61958 + }, + { + "epoch": 3.4261052631578948, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0639, + "step": 61959 + }, + { + "epoch": 3.4261654135338344, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0494, + "step": 61960 + }, + { + "epoch": 3.4262255639097745, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 61961 + }, + { + "epoch": 3.426285714285714, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0796, + "step": 61962 + }, + { + "epoch": 3.4263458646616543, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 61963 + }, + { + "epoch": 3.426406015037594, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 61964 + }, + { + "epoch": 3.426466165413534, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9303, + "step": 61965 + }, + { + "epoch": 3.4265263157894736, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.911, + "step": 61966 + }, + { + "epoch": 3.4265864661654133, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0636, + "step": 61967 + }, + { + "epoch": 3.4266466165413534, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 61968 + }, + { + "epoch": 3.4267067669172935, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9374, + "step": 61969 + }, + { + "epoch": 3.426766917293233, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9655, + "step": 61970 + }, + { + "epoch": 3.4268270676691728, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 61971 + }, + { + "epoch": 3.426887218045113, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0956, + "step": 61972 + }, + { + "epoch": 3.4269473684210525, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 61973 + }, + { + "epoch": 3.4270075187969926, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.062, + "step": 61974 + }, + { + "epoch": 3.4270676691729323, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 61975 + }, + { + "epoch": 3.4271278195488724, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.1542, + "step": 61976 + }, + { + "epoch": 3.427187969924812, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1322, + "step": 61977 + }, + { + "epoch": 3.427248120300752, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0746, + "step": 61978 + }, + { + "epoch": 3.4273082706766917, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9179, + "step": 61979 + }, + { + "epoch": 3.4273684210526314, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 61980 + }, + { + "epoch": 3.4274285714285715, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 61981 + }, + { + "epoch": 3.427488721804511, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 61982 + }, + { + "epoch": 3.4275488721804512, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0447, + "step": 61983 + }, + { + "epoch": 3.427609022556391, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 61984 + }, + { + "epoch": 3.427669172932331, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 61985 + }, + { + "epoch": 3.4277293233082706, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 61986 + }, + { + "epoch": 3.4277894736842107, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 61987 + }, + { + "epoch": 3.4278496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 61988 + }, + { + "epoch": 3.42790977443609, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0096, + "step": 61989 + }, + { + "epoch": 3.42796992481203, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 61990 + }, + { + "epoch": 3.4280300751879698, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 61991 + }, + { + "epoch": 3.42809022556391, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 61992 + }, + { + "epoch": 3.4281503759398495, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9399, + "step": 61993 + }, + { + "epoch": 3.4282105263157896, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 61994 + }, + { + "epoch": 3.4282706766917292, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 61995 + }, + { + "epoch": 3.4283308270676693, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 61996 + }, + { + "epoch": 3.428390977443609, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9737, + "step": 61997 + }, + { + "epoch": 3.4284511278195486, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 61998 + }, + { + "epoch": 3.4285112781954887, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 61999 + }, + { + "epoch": 3.4285714285714284, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 62000 + }, + { + "epoch": 3.4286315789473685, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0972, + "step": 62001 + }, + { + "epoch": 3.428691729323308, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 62002 + }, + { + "epoch": 3.428751879699248, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 62003 + }, + { + "epoch": 3.428812030075188, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9262, + "step": 62004 + }, + { + "epoch": 3.428872180451128, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 62005 + }, + { + "epoch": 3.4289323308270676, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 62006 + }, + { + "epoch": 3.4289924812030077, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 62007 + }, + { + "epoch": 3.4290526315789474, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 62008 + }, + { + "epoch": 3.4291127819548874, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 62009 + }, + { + "epoch": 3.429172932330827, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 62010 + }, + { + "epoch": 3.4292330827067667, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 62011 + }, + { + "epoch": 3.429293233082707, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 62012 + }, + { + "epoch": 3.4293533834586465, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9714, + "step": 62013 + }, + { + "epoch": 3.4294135338345866, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 62014 + }, + { + "epoch": 3.4294736842105262, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9693, + "step": 62015 + }, + { + "epoch": 3.4295338345864663, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0006, + "step": 62016 + }, + { + "epoch": 3.429593984962406, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 62017 + }, + { + "epoch": 3.429654135338346, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1037, + "step": 62018 + }, + { + "epoch": 3.4297142857142857, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 62019 + }, + { + "epoch": 3.4297744360902254, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.125, + "step": 62020 + }, + { + "epoch": 3.4298345864661655, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0198, + "step": 62021 + }, + { + "epoch": 3.429894736842105, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 62022 + }, + { + "epoch": 3.429954887218045, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0722, + "step": 62023 + }, + { + "epoch": 3.430015037593985, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 62024 + }, + { + "epoch": 3.430075187969925, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 62025 + }, + { + "epoch": 3.4301353383458646, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 62026 + }, + { + "epoch": 3.4301954887218047, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 62027 + }, + { + "epoch": 3.4302556390977443, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 62028 + }, + { + "epoch": 3.430315789473684, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1926, + "step": 62029 + }, + { + "epoch": 3.430375939849624, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 62030 + }, + { + "epoch": 3.4304360902255637, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 0.9395, + "step": 62031 + }, + { + "epoch": 3.430496240601504, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 62032 + }, + { + "epoch": 3.4305563909774435, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.1188, + "step": 62033 + }, + { + "epoch": 3.4306165413533836, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1601, + "step": 62034 + }, + { + "epoch": 3.430676691729323, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 62035 + }, + { + "epoch": 3.4307368421052633, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 62036 + }, + { + "epoch": 3.430796992481203, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9183, + "step": 62037 + }, + { + "epoch": 3.430857142857143, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 62038 + }, + { + "epoch": 3.4309172932330827, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 62039 + }, + { + "epoch": 3.430977443609023, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0573, + "step": 62040 + }, + { + "epoch": 3.4310375939849624, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 62041 + }, + { + "epoch": 3.431097744360902, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9647, + "step": 62042 + }, + { + "epoch": 3.431157894736842, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 62043 + }, + { + "epoch": 3.431218045112782, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 62044 + }, + { + "epoch": 3.431278195488722, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 62045 + }, + { + "epoch": 3.4313383458646616, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.902, + "step": 62046 + }, + { + "epoch": 3.4313984962406017, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9954, + "step": 62047 + }, + { + "epoch": 3.4314586466165413, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 62048 + }, + { + "epoch": 3.4315187969924814, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0215, + "step": 62049 + }, + { + "epoch": 3.431578947368421, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.9601, + "step": 62050 + }, + { + "epoch": 3.4316390977443607, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1004, + "step": 62051 + }, + { + "epoch": 3.431699248120301, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 62052 + }, + { + "epoch": 3.4317593984962405, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0194, + "step": 62053 + }, + { + "epoch": 3.4318195488721805, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 62054 + }, + { + "epoch": 3.43187969924812, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9705, + "step": 62055 + }, + { + "epoch": 3.4319398496240603, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9424, + "step": 62056 + }, + { + "epoch": 3.432, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 62057 + }, + { + "epoch": 3.43206015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0054, + "step": 62058 + }, + { + "epoch": 3.4321203007518797, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0442, + "step": 62059 + }, + { + "epoch": 3.4321804511278193, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0681, + "step": 62060 + }, + { + "epoch": 3.4322406015037594, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0202, + "step": 62061 + }, + { + "epoch": 3.432300751879699, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9929, + "step": 62062 + }, + { + "epoch": 3.432360902255639, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 62063 + }, + { + "epoch": 3.432421052631579, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 62064 + }, + { + "epoch": 3.432481203007519, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 62065 + }, + { + "epoch": 3.4325413533834586, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 62066 + }, + { + "epoch": 3.4326015037593987, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 62067 + }, + { + "epoch": 3.4326616541353383, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 62068 + }, + { + "epoch": 3.432721804511278, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.1469, + "step": 62069 + }, + { + "epoch": 3.432781954887218, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0467, + "step": 62070 + }, + { + "epoch": 3.432842105263158, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 62071 + }, + { + "epoch": 3.432902255639098, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9279, + "step": 62072 + }, + { + "epoch": 3.4329624060150374, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 62073 + }, + { + "epoch": 3.4330225563909775, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 62074 + }, + { + "epoch": 3.433082706766917, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1076, + "step": 62075 + }, + { + "epoch": 3.4331428571428573, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0846, + "step": 62076 + }, + { + "epoch": 3.433203007518797, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1057, + "step": 62077 + }, + { + "epoch": 3.433263157894737, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9149, + "step": 62078 + }, + { + "epoch": 3.4333233082706767, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.1391, + "step": 62079 + }, + { + "epoch": 3.4333834586466168, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0026, + "step": 62080 + }, + { + "epoch": 3.4334436090225564, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0908, + "step": 62081 + }, + { + "epoch": 3.433503759398496, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 62082 + }, + { + "epoch": 3.433563909774436, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0378, + "step": 62083 + }, + { + "epoch": 3.433624060150376, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9285, + "step": 62084 + }, + { + "epoch": 3.433684210526316, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 62085 + }, + { + "epoch": 3.4337443609022555, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 62086 + }, + { + "epoch": 3.4338045112781956, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0032, + "step": 62087 + }, + { + "epoch": 3.4338646616541353, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 62088 + }, + { + "epoch": 3.4339248120300754, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.1885, + "step": 62089 + }, + { + "epoch": 3.433984962406015, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.1598, + "step": 62090 + }, + { + "epoch": 3.4340451127819547, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9829, + "step": 62091 + }, + { + "epoch": 3.4341052631578948, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 62092 + }, + { + "epoch": 3.4341654135338344, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0538, + "step": 62093 + }, + { + "epoch": 3.4342255639097745, + "grad_norm": 1.2109375, + "learning_rate": 3e-05, + "loss": 1.0977, + "step": 62094 + }, + { + "epoch": 3.434285714285714, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0035, + "step": 62095 + }, + { + "epoch": 3.4343458646616543, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.986, + "step": 62096 + }, + { + "epoch": 3.434406015037594, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9863, + "step": 62097 + }, + { + "epoch": 3.434466165413534, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 62098 + }, + { + "epoch": 3.4345263157894736, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9465, + "step": 62099 + }, + { + "epoch": 3.4345864661654133, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 62100 + }, + { + "epoch": 3.4346466165413534, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 62101 + }, + { + "epoch": 3.434706766917293, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 62102 + }, + { + "epoch": 3.434766917293233, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 62103 + }, + { + "epoch": 3.434827067669173, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 62104 + }, + { + "epoch": 3.434887218045113, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 62105 + }, + { + "epoch": 3.4349473684210525, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9589, + "step": 62106 + }, + { + "epoch": 3.4350075187969926, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 62107 + }, + { + "epoch": 3.4350676691729323, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.1203, + "step": 62108 + }, + { + "epoch": 3.4351278195488724, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9409, + "step": 62109 + }, + { + "epoch": 3.435187969924812, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 62110 + }, + { + "epoch": 3.435248120300752, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 62111 + }, + { + "epoch": 3.4353082706766918, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9243, + "step": 62112 + }, + { + "epoch": 3.4353684210526314, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 0.92, + "step": 62113 + }, + { + "epoch": 3.4354285714285715, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 62114 + }, + { + "epoch": 3.435488721804511, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0257, + "step": 62115 + }, + { + "epoch": 3.4355488721804512, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0334, + "step": 62116 + }, + { + "epoch": 3.435609022556391, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 62117 + }, + { + "epoch": 3.435669172932331, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 62118 + }, + { + "epoch": 3.4357293233082706, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 62119 + }, + { + "epoch": 3.4357894736842107, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 62120 + }, + { + "epoch": 3.4358496240601504, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 62121 + }, + { + "epoch": 3.43590977443609, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9695, + "step": 62122 + }, + { + "epoch": 3.43596992481203, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0172, + "step": 62123 + }, + { + "epoch": 3.4360300751879698, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 62124 + }, + { + "epoch": 3.43609022556391, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.1383, + "step": 62125 + }, + { + "epoch": 3.4361503759398495, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0568, + "step": 62126 + }, + { + "epoch": 3.4362105263157896, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0089, + "step": 62127 + }, + { + "epoch": 3.4362706766917293, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 62128 + }, + { + "epoch": 3.4363308270676693, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 62129 + }, + { + "epoch": 3.436390977443609, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9545, + "step": 62130 + }, + { + "epoch": 3.4364511278195486, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0755, + "step": 62131 + }, + { + "epoch": 3.4365112781954887, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9896, + "step": 62132 + }, + { + "epoch": 3.4365714285714284, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 62133 + }, + { + "epoch": 3.4366315789473685, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 62134 + }, + { + "epoch": 3.436691729323308, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9953, + "step": 62135 + }, + { + "epoch": 3.4367518796992482, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 62136 + }, + { + "epoch": 3.436812030075188, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 62137 + }, + { + "epoch": 3.436872180451128, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 1.0047, + "step": 62138 + }, + { + "epoch": 3.4369323308270676, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 0.9187, + "step": 62139 + }, + { + "epoch": 3.4369924812030077, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 62140 + }, + { + "epoch": 3.4370526315789474, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0544, + "step": 62141 + }, + { + "epoch": 3.4371127819548875, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 62142 + }, + { + "epoch": 3.437172932330827, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 62143 + }, + { + "epoch": 3.4372330827067668, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0305, + "step": 62144 + }, + { + "epoch": 3.437293233082707, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9019, + "step": 62145 + }, + { + "epoch": 3.4373533834586465, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 62146 + }, + { + "epoch": 3.4374135338345866, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 62147 + }, + { + "epoch": 3.4374736842105262, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9415, + "step": 62148 + }, + { + "epoch": 3.4375338345864663, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 62149 + }, + { + "epoch": 3.437593984962406, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 62150 + }, + { + "epoch": 3.437654135338346, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 62151 + }, + { + "epoch": 3.4377142857142857, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0699, + "step": 62152 + }, + { + "epoch": 3.4377744360902254, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 62153 + }, + { + "epoch": 3.4378345864661655, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 62154 + }, + { + "epoch": 3.437894736842105, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 62155 + }, + { + "epoch": 3.437954887218045, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 62156 + }, + { + "epoch": 3.438015037593985, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9258, + "step": 62157 + }, + { + "epoch": 3.438075187969925, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0625, + "step": 62158 + }, + { + "epoch": 3.4381353383458646, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.069, + "step": 62159 + }, + { + "epoch": 3.4381954887218047, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 62160 + }, + { + "epoch": 3.4382556390977443, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 62161 + }, + { + "epoch": 3.438315789473684, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9728, + "step": 62162 + }, + { + "epoch": 3.438375939849624, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 62163 + }, + { + "epoch": 3.4384360902255637, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 62164 + }, + { + "epoch": 3.438496240601504, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 62165 + }, + { + "epoch": 3.4385563909774435, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 62166 + }, + { + "epoch": 3.4386165413533836, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9644, + "step": 62167 + }, + { + "epoch": 3.438676691729323, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 62168 + }, + { + "epoch": 3.4387368421052633, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 62169 + }, + { + "epoch": 3.438796992481203, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9457, + "step": 62170 + }, + { + "epoch": 3.4388571428571426, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9639, + "step": 62171 + }, + { + "epoch": 3.4389172932330827, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 62172 + }, + { + "epoch": 3.438977443609023, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9995, + "step": 62173 + }, + { + "epoch": 3.4390375939849624, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 62174 + }, + { + "epoch": 3.439097744360902, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 62175 + }, + { + "epoch": 3.439157894736842, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0279, + "step": 62176 + }, + { + "epoch": 3.439218045112782, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 62177 + }, + { + "epoch": 3.439278195488722, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 62178 + }, + { + "epoch": 3.4393383458646616, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 62179 + }, + { + "epoch": 3.4393984962406017, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 62180 + }, + { + "epoch": 3.4394586466165413, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 62181 + }, + { + "epoch": 3.4395187969924814, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9582, + "step": 62182 + }, + { + "epoch": 3.439578947368421, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0161, + "step": 62183 + }, + { + "epoch": 3.4396390977443607, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 62184 + }, + { + "epoch": 3.439699248120301, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 62185 + }, + { + "epoch": 3.4397593984962405, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 62186 + }, + { + "epoch": 3.4398195488721806, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 62187 + }, + { + "epoch": 3.43987969924812, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 62188 + }, + { + "epoch": 3.4399398496240603, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 1.0734, + "step": 62189 + }, + { + "epoch": 3.44, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0353, + "step": 62190 + }, + { + "epoch": 3.44006015037594, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 62191 + }, + { + "epoch": 3.4401203007518797, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 62192 + }, + { + "epoch": 3.4401804511278193, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1407, + "step": 62193 + }, + { + "epoch": 3.4402406015037594, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9387, + "step": 62194 + }, + { + "epoch": 3.440300751879699, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 62195 + }, + { + "epoch": 3.440360902255639, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 62196 + }, + { + "epoch": 3.440421052631579, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 62197 + }, + { + "epoch": 3.440481203007519, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 62198 + }, + { + "epoch": 3.4405413533834586, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 62199 + }, + { + "epoch": 3.4406015037593987, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9133, + "step": 62200 + }, + { + "epoch": 3.4406616541353383, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 62201 + }, + { + "epoch": 3.440721804511278, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 62202 + }, + { + "epoch": 3.440781954887218, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 62203 + }, + { + "epoch": 3.4408421052631577, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 62204 + }, + { + "epoch": 3.440902255639098, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.075, + "step": 62205 + }, + { + "epoch": 3.4409624060150374, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0264, + "step": 62206 + }, + { + "epoch": 3.4410225563909775, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9269, + "step": 62207 + }, + { + "epoch": 3.441082706766917, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9659, + "step": 62208 + }, + { + "epoch": 3.4411428571428573, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 62209 + }, + { + "epoch": 3.441203007518797, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 62210 + }, + { + "epoch": 3.441263157894737, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.2933, + "step": 62211 + }, + { + "epoch": 3.4413233082706767, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9146, + "step": 62212 + }, + { + "epoch": 3.4413834586466168, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 62213 + }, + { + "epoch": 3.4414436090225564, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 62214 + }, + { + "epoch": 3.441503759398496, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 62215 + }, + { + "epoch": 3.441563909774436, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 62216 + }, + { + "epoch": 3.441624060150376, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 62217 + }, + { + "epoch": 3.441684210526316, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 62218 + }, + { + "epoch": 3.4417443609022556, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0073, + "step": 62219 + }, + { + "epoch": 3.4418045112781956, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 62220 + }, + { + "epoch": 3.4418646616541353, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0454, + "step": 62221 + }, + { + "epoch": 3.4419248120300754, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0119, + "step": 62222 + }, + { + "epoch": 3.441984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0455, + "step": 62223 + }, + { + "epoch": 3.4420451127819547, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.938, + "step": 62224 + }, + { + "epoch": 3.442105263157895, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.857, + "step": 62225 + }, + { + "epoch": 3.4421654135338344, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 62226 + }, + { + "epoch": 3.4422255639097745, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 62227 + }, + { + "epoch": 3.442285714285714, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 62228 + }, + { + "epoch": 3.4423458646616543, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 62229 + }, + { + "epoch": 3.442406015037594, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0933, + "step": 62230 + }, + { + "epoch": 3.442466165413534, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 62231 + }, + { + "epoch": 3.4425263157894737, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1582, + "step": 62232 + }, + { + "epoch": 3.4425864661654133, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9303, + "step": 62233 + }, + { + "epoch": 3.4426466165413534, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 62234 + }, + { + "epoch": 3.442706766917293, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.8633, + "step": 62235 + }, + { + "epoch": 3.442766917293233, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0818, + "step": 62236 + }, + { + "epoch": 3.442827067669173, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9887, + "step": 62237 + }, + { + "epoch": 3.442887218045113, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 62238 + }, + { + "epoch": 3.4429473684210525, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1385, + "step": 62239 + }, + { + "epoch": 3.4430075187969926, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 62240 + }, + { + "epoch": 3.4430676691729323, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 62241 + }, + { + "epoch": 3.4431278195488724, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9847, + "step": 62242 + }, + { + "epoch": 3.443187969924812, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.1024, + "step": 62243 + }, + { + "epoch": 3.443248120300752, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 62244 + }, + { + "epoch": 3.4433082706766918, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.948, + "step": 62245 + }, + { + "epoch": 3.4433684210526314, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 62246 + }, + { + "epoch": 3.4434285714285715, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 62247 + }, + { + "epoch": 3.443488721804511, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 62248 + }, + { + "epoch": 3.4435488721804512, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 62249 + }, + { + "epoch": 3.443609022556391, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0548, + "step": 62250 + }, + { + "epoch": 3.443669172932331, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 1.1306, + "step": 62251 + }, + { + "epoch": 3.4437293233082706, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 62252 + }, + { + "epoch": 3.4437894736842107, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.933, + "step": 62253 + }, + { + "epoch": 3.4438496240601504, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9855, + "step": 62254 + }, + { + "epoch": 3.44390977443609, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0509, + "step": 62255 + }, + { + "epoch": 3.44396992481203, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 62256 + }, + { + "epoch": 3.4440300751879698, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 62257 + }, + { + "epoch": 3.44409022556391, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.905, + "step": 62258 + }, + { + "epoch": 3.4441503759398495, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 62259 + }, + { + "epoch": 3.4442105263157896, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 62260 + }, + { + "epoch": 3.4442706766917293, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9982, + "step": 62261 + }, + { + "epoch": 3.4443308270676694, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 62262 + }, + { + "epoch": 3.444390977443609, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0229, + "step": 62263 + }, + { + "epoch": 3.4444511278195487, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 62264 + }, + { + "epoch": 3.4445112781954887, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9513, + "step": 62265 + }, + { + "epoch": 3.4445714285714284, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0562, + "step": 62266 + }, + { + "epoch": 3.4446315789473685, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 62267 + }, + { + "epoch": 3.444691729323308, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 62268 + }, + { + "epoch": 3.4447518796992482, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 62269 + }, + { + "epoch": 3.444812030075188, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 62270 + }, + { + "epoch": 3.444872180451128, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0753, + "step": 62271 + }, + { + "epoch": 3.4449323308270676, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 62272 + }, + { + "epoch": 3.4449924812030073, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 62273 + }, + { + "epoch": 3.4450526315789474, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 62274 + }, + { + "epoch": 3.4451127819548875, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 62275 + }, + { + "epoch": 3.445172932330827, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9725, + "step": 62276 + }, + { + "epoch": 3.4452330827067668, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 62277 + }, + { + "epoch": 3.445293233082707, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.047, + "step": 62278 + }, + { + "epoch": 3.4453533834586465, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 62279 + }, + { + "epoch": 3.4454135338345866, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 62280 + }, + { + "epoch": 3.4454736842105262, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9254, + "step": 62281 + }, + { + "epoch": 3.4455338345864663, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 62282 + }, + { + "epoch": 3.445593984962406, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 62283 + }, + { + "epoch": 3.445654135338346, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 62284 + }, + { + "epoch": 3.4457142857142857, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 62285 + }, + { + "epoch": 3.4457744360902254, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9998, + "step": 62286 + }, + { + "epoch": 3.4458345864661655, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 62287 + }, + { + "epoch": 3.445894736842105, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 62288 + }, + { + "epoch": 3.445954887218045, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 62289 + }, + { + "epoch": 3.446015037593985, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 62290 + }, + { + "epoch": 3.446075187969925, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 62291 + }, + { + "epoch": 3.4461353383458646, + "grad_norm": 0.5, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 62292 + }, + { + "epoch": 3.4461954887218047, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0064, + "step": 62293 + }, + { + "epoch": 3.4462556390977443, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9902, + "step": 62294 + }, + { + "epoch": 3.446315789473684, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 62295 + }, + { + "epoch": 3.446375939849624, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0555, + "step": 62296 + }, + { + "epoch": 3.4464360902255637, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 62297 + }, + { + "epoch": 3.446496240601504, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 62298 + }, + { + "epoch": 3.4465563909774435, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9365, + "step": 62299 + }, + { + "epoch": 3.4466165413533836, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9975, + "step": 62300 + }, + { + "epoch": 3.4466766917293232, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 62301 + }, + { + "epoch": 3.4467368421052633, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 62302 + }, + { + "epoch": 3.446796992481203, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 62303 + }, + { + "epoch": 3.4468571428571426, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 62304 + }, + { + "epoch": 3.4469172932330827, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 62305 + }, + { + "epoch": 3.4469774436090224, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 62306 + }, + { + "epoch": 3.4470375939849625, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.8643, + "step": 62307 + }, + { + "epoch": 3.447097744360902, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9131, + "step": 62308 + }, + { + "epoch": 3.447157894736842, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 62309 + }, + { + "epoch": 3.447218045112782, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 62310 + }, + { + "epoch": 3.447278195488722, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 62311 + }, + { + "epoch": 3.4473383458646616, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 62312 + }, + { + "epoch": 3.4473984962406017, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 62313 + }, + { + "epoch": 3.4474586466165413, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9732, + "step": 62314 + }, + { + "epoch": 3.4475187969924814, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 62315 + }, + { + "epoch": 3.447578947368421, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.1025, + "step": 62316 + }, + { + "epoch": 3.4476390977443607, + "grad_norm": 1.3046875, + "learning_rate": 3e-05, + "loss": 1.0709, + "step": 62317 + }, + { + "epoch": 3.447699248120301, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 62318 + }, + { + "epoch": 3.4477593984962405, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 62319 + }, + { + "epoch": 3.4478195488721806, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0088, + "step": 62320 + }, + { + "epoch": 3.44787969924812, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 62321 + }, + { + "epoch": 3.4479398496240603, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0359, + "step": 62322 + }, + { + "epoch": 3.448, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0873, + "step": 62323 + }, + { + "epoch": 3.44806015037594, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9268, + "step": 62324 + }, + { + "epoch": 3.4481203007518797, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 62325 + }, + { + "epoch": 3.4481804511278193, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 62326 + }, + { + "epoch": 3.4482406015037594, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 62327 + }, + { + "epoch": 3.448300751879699, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 62328 + }, + { + "epoch": 3.448360902255639, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0881, + "step": 62329 + }, + { + "epoch": 3.448421052631579, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0653, + "step": 62330 + }, + { + "epoch": 3.448481203007519, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9313, + "step": 62331 + }, + { + "epoch": 3.4485413533834586, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9736, + "step": 62332 + }, + { + "epoch": 3.4486015037593987, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0262, + "step": 62333 + }, + { + "epoch": 3.4486616541353383, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 62334 + }, + { + "epoch": 3.448721804511278, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.945, + "step": 62335 + }, + { + "epoch": 3.448781954887218, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 62336 + }, + { + "epoch": 3.4488421052631577, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 62337 + }, + { + "epoch": 3.448902255639098, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0464, + "step": 62338 + }, + { + "epoch": 3.4489624060150375, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.8725, + "step": 62339 + }, + { + "epoch": 3.4490225563909775, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9785, + "step": 62340 + }, + { + "epoch": 3.449082706766917, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9339, + "step": 62341 + }, + { + "epoch": 3.4491428571428573, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 62342 + }, + { + "epoch": 3.449203007518797, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0098, + "step": 62343 + }, + { + "epoch": 3.449263157894737, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0129, + "step": 62344 + }, + { + "epoch": 3.4493233082706767, + "grad_norm": 1.1171875, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 62345 + }, + { + "epoch": 3.4493834586466168, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 62346 + }, + { + "epoch": 3.4494436090225564, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0582, + "step": 62347 + }, + { + "epoch": 3.449503759398496, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 62348 + }, + { + "epoch": 3.449563909774436, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 62349 + }, + { + "epoch": 3.449624060150376, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9488, + "step": 62350 + }, + { + "epoch": 3.449684210526316, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9743, + "step": 62351 + }, + { + "epoch": 3.4497443609022556, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.8837, + "step": 62352 + }, + { + "epoch": 3.4498045112781957, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 62353 + }, + { + "epoch": 3.4498646616541353, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 0.9952, + "step": 62354 + }, + { + "epoch": 3.4499248120300754, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0091, + "step": 62355 + }, + { + "epoch": 3.449984962406015, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9972, + "step": 62356 + }, + { + "epoch": 3.4500451127819547, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0492, + "step": 62357 + }, + { + "epoch": 3.450105263157895, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 62358 + }, + { + "epoch": 3.4501654135338344, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 62359 + }, + { + "epoch": 3.4502255639097745, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 62360 + }, + { + "epoch": 3.450285714285714, + "grad_norm": 1.1875, + "learning_rate": 3e-05, + "loss": 0.9217, + "step": 62361 + }, + { + "epoch": 3.4503458646616543, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 62362 + }, + { + "epoch": 3.450406015037594, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 62363 + }, + { + "epoch": 3.450466165413534, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 62364 + }, + { + "epoch": 3.4505263157894737, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 62365 + }, + { + "epoch": 3.4505864661654133, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.1205, + "step": 62366 + }, + { + "epoch": 3.4506466165413534, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9678, + "step": 62367 + }, + { + "epoch": 3.450706766917293, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 62368 + }, + { + "epoch": 3.450766917293233, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.874, + "step": 62369 + }, + { + "epoch": 3.450827067669173, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 62370 + }, + { + "epoch": 3.450887218045113, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 62371 + }, + { + "epoch": 3.4509473684210525, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 62372 + }, + { + "epoch": 3.4510075187969926, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9993, + "step": 62373 + }, + { + "epoch": 3.4510676691729323, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 62374 + }, + { + "epoch": 3.451127819548872, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.2063, + "step": 62375 + }, + { + "epoch": 3.451187969924812, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.029, + "step": 62376 + }, + { + "epoch": 3.451248120300752, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 62377 + }, + { + "epoch": 3.4513082706766918, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 62378 + }, + { + "epoch": 3.4513684210526314, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0103, + "step": 62379 + }, + { + "epoch": 3.4514285714285715, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 62380 + }, + { + "epoch": 3.451488721804511, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 62381 + }, + { + "epoch": 3.4515488721804513, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.1001, + "step": 62382 + }, + { + "epoch": 3.451609022556391, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 62383 + }, + { + "epoch": 3.451669172932331, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9502, + "step": 62384 + }, + { + "epoch": 3.4517293233082706, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 62385 + }, + { + "epoch": 3.4517894736842107, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 62386 + }, + { + "epoch": 3.4518496240601504, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 62387 + }, + { + "epoch": 3.45190977443609, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 62388 + }, + { + "epoch": 3.45196992481203, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9157, + "step": 62389 + }, + { + "epoch": 3.45203007518797, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 62390 + }, + { + "epoch": 3.45209022556391, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 62391 + }, + { + "epoch": 3.4521503759398495, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0707, + "step": 62392 + }, + { + "epoch": 3.4522105263157896, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9577, + "step": 62393 + }, + { + "epoch": 3.4522706766917293, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 0.8976, + "step": 62394 + }, + { + "epoch": 3.4523308270676694, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 62395 + }, + { + "epoch": 3.452390977443609, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0691, + "step": 62396 + }, + { + "epoch": 3.4524511278195487, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 62397 + }, + { + "epoch": 3.4525112781954888, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9641, + "step": 62398 + }, + { + "epoch": 3.4525714285714284, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0163, + "step": 62399 + }, + { + "epoch": 3.4526315789473685, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 62400 + }, + { + "epoch": 3.452691729323308, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 62401 + }, + { + "epoch": 3.4527518796992482, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.981, + "step": 62402 + }, + { + "epoch": 3.452812030075188, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 62403 + }, + { + "epoch": 3.452872180451128, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0357, + "step": 62404 + }, + { + "epoch": 3.4529323308270676, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 62405 + }, + { + "epoch": 3.4529924812030073, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0214, + "step": 62406 + }, + { + "epoch": 3.4530526315789474, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 62407 + }, + { + "epoch": 3.4531127819548875, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.93, + "step": 62408 + }, + { + "epoch": 3.453172932330827, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0647, + "step": 62409 + }, + { + "epoch": 3.4532330827067668, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0595, + "step": 62410 + }, + { + "epoch": 3.453293233082707, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 62411 + }, + { + "epoch": 3.4533533834586465, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9588, + "step": 62412 + }, + { + "epoch": 3.4534135338345866, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9715, + "step": 62413 + }, + { + "epoch": 3.4534736842105263, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 62414 + }, + { + "epoch": 3.4535338345864663, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0094, + "step": 62415 + }, + { + "epoch": 3.453593984962406, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 62416 + }, + { + "epoch": 3.453654135338346, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 62417 + }, + { + "epoch": 3.4537142857142857, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0199, + "step": 62418 + }, + { + "epoch": 3.4537744360902254, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9188, + "step": 62419 + }, + { + "epoch": 3.4538345864661655, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 62420 + }, + { + "epoch": 3.453894736842105, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9562, + "step": 62421 + }, + { + "epoch": 3.4539548872180452, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0853, + "step": 62422 + }, + { + "epoch": 3.454015037593985, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.8873, + "step": 62423 + }, + { + "epoch": 3.454075187969925, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0815, + "step": 62424 + }, + { + "epoch": 3.4541353383458646, + "grad_norm": 1.265625, + "learning_rate": 3e-05, + "loss": 1.1297, + "step": 62425 + }, + { + "epoch": 3.4541954887218047, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9839, + "step": 62426 + }, + { + "epoch": 3.4542556390977444, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9858, + "step": 62427 + }, + { + "epoch": 3.454315789473684, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 62428 + }, + { + "epoch": 3.454375939849624, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9771, + "step": 62429 + }, + { + "epoch": 3.4544360902255637, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 62430 + }, + { + "epoch": 3.454496240601504, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 62431 + }, + { + "epoch": 3.4545563909774435, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 62432 + }, + { + "epoch": 3.4546165413533836, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9349, + "step": 62433 + }, + { + "epoch": 3.4546766917293232, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 62434 + }, + { + "epoch": 3.4547368421052633, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9568, + "step": 62435 + }, + { + "epoch": 3.454796992481203, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 62436 + }, + { + "epoch": 3.4548571428571426, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 62437 + }, + { + "epoch": 3.4549172932330827, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 62438 + }, + { + "epoch": 3.4549774436090224, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9271, + "step": 62439 + }, + { + "epoch": 3.4550375939849625, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9701, + "step": 62440 + }, + { + "epoch": 3.455097744360902, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9472, + "step": 62441 + }, + { + "epoch": 3.455157894736842, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 62442 + }, + { + "epoch": 3.455218045112782, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 62443 + }, + { + "epoch": 3.455278195488722, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9499, + "step": 62444 + }, + { + "epoch": 3.4553383458646616, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 62445 + }, + { + "epoch": 3.4553984962406017, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 62446 + }, + { + "epoch": 3.4554586466165413, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 62447 + }, + { + "epoch": 3.4555187969924814, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 62448 + }, + { + "epoch": 3.455578947368421, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9637, + "step": 62449 + }, + { + "epoch": 3.4556390977443607, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.1007, + "step": 62450 + }, + { + "epoch": 3.455699248120301, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0213, + "step": 62451 + }, + { + "epoch": 3.4557593984962405, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 62452 + }, + { + "epoch": 3.4558195488721806, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9981, + "step": 62453 + }, + { + "epoch": 3.45587969924812, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 62454 + }, + { + "epoch": 3.4559398496240603, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0077, + "step": 62455 + }, + { + "epoch": 3.456, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 62456 + }, + { + "epoch": 3.45606015037594, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 62457 + }, + { + "epoch": 3.4561203007518797, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 62458 + }, + { + "epoch": 3.4561804511278194, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0599, + "step": 62459 + }, + { + "epoch": 3.4562406015037594, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0974, + "step": 62460 + }, + { + "epoch": 3.456300751879699, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9521, + "step": 62461 + }, + { + "epoch": 3.456360902255639, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0887, + "step": 62462 + }, + { + "epoch": 3.456421052631579, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.1625, + "step": 62463 + }, + { + "epoch": 3.456481203007519, + "grad_norm": 0.9765625, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 62464 + }, + { + "epoch": 3.4565413533834586, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9169, + "step": 62465 + }, + { + "epoch": 3.4566015037593987, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 0.99, + "step": 62466 + }, + { + "epoch": 3.4566616541353383, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 62467 + }, + { + "epoch": 3.456721804511278, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 62468 + }, + { + "epoch": 3.456781954887218, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 0.9868, + "step": 62469 + }, + { + "epoch": 3.4568421052631577, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9585, + "step": 62470 + }, + { + "epoch": 3.456902255639098, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9285, + "step": 62471 + }, + { + "epoch": 3.4569624060150375, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.056, + "step": 62472 + }, + { + "epoch": 3.4570225563909776, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 62473 + }, + { + "epoch": 3.457082706766917, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 62474 + }, + { + "epoch": 3.4571428571428573, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.9267, + "step": 62475 + }, + { + "epoch": 3.457203007518797, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0917, + "step": 62476 + }, + { + "epoch": 3.4572631578947366, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9574, + "step": 62477 + }, + { + "epoch": 3.4573233082706767, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 62478 + }, + { + "epoch": 3.457383458646617, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9149, + "step": 62479 + }, + { + "epoch": 3.4574436090225564, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 62480 + }, + { + "epoch": 3.457503759398496, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.035, + "step": 62481 + }, + { + "epoch": 3.457563909774436, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9242, + "step": 62482 + }, + { + "epoch": 3.457624060150376, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0569, + "step": 62483 + }, + { + "epoch": 3.457684210526316, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 0.9361, + "step": 62484 + }, + { + "epoch": 3.4577443609022556, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0173, + "step": 62485 + }, + { + "epoch": 3.4578045112781957, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9226, + "step": 62486 + }, + { + "epoch": 3.4578646616541353, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 62487 + }, + { + "epoch": 3.4579248120300754, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 62488 + }, + { + "epoch": 3.457984962406015, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9319, + "step": 62489 + }, + { + "epoch": 3.4580451127819547, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 62490 + }, + { + "epoch": 3.458105263157895, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 62491 + }, + { + "epoch": 3.4581654135338344, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 62492 + }, + { + "epoch": 3.4582255639097745, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 62493 + }, + { + "epoch": 3.458285714285714, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0445, + "step": 62494 + }, + { + "epoch": 3.4583458646616543, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.963, + "step": 62495 + }, + { + "epoch": 3.458406015037594, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1062, + "step": 62496 + }, + { + "epoch": 3.458466165413534, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 62497 + }, + { + "epoch": 3.4585263157894737, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 62498 + }, + { + "epoch": 3.4585864661654133, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 62499 + }, + { + "epoch": 3.4586466165413534, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 62500 + }, + { + "epoch": 3.458706766917293, + "grad_norm": 1.171875, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 62501 + }, + { + "epoch": 3.458766917293233, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0143, + "step": 62502 + }, + { + "epoch": 3.458827067669173, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9507, + "step": 62503 + }, + { + "epoch": 3.458887218045113, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0142, + "step": 62504 + }, + { + "epoch": 3.4589473684210525, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 62505 + }, + { + "epoch": 3.4590075187969926, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.1109, + "step": 62506 + }, + { + "epoch": 3.4590676691729323, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9132, + "step": 62507 + }, + { + "epoch": 3.459127819548872, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0184, + "step": 62508 + }, + { + "epoch": 3.459187969924812, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 62509 + }, + { + "epoch": 3.459248120300752, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 62510 + }, + { + "epoch": 3.4593082706766918, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 62511 + }, + { + "epoch": 3.4593684210526314, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 62512 + }, + { + "epoch": 3.4594285714285715, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0905, + "step": 62513 + }, + { + "epoch": 3.459488721804511, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 62514 + }, + { + "epoch": 3.4595488721804513, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 62515 + }, + { + "epoch": 3.459609022556391, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0037, + "step": 62516 + }, + { + "epoch": 3.459669172932331, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9746, + "step": 62517 + }, + { + "epoch": 3.4597293233082707, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9961, + "step": 62518 + }, + { + "epoch": 3.4597894736842107, + "grad_norm": 0.48828125, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 62519 + }, + { + "epoch": 3.4598496240601504, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 62520 + }, + { + "epoch": 3.45990977443609, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 62521 + }, + { + "epoch": 3.45996992481203, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9904, + "step": 62522 + }, + { + "epoch": 3.46003007518797, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 62523 + }, + { + "epoch": 3.46009022556391, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9635, + "step": 62524 + }, + { + "epoch": 3.4601503759398495, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9851, + "step": 62525 + }, + { + "epoch": 3.4602105263157896, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 62526 + }, + { + "epoch": 3.4602706766917293, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 62527 + }, + { + "epoch": 3.4603308270676694, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0721, + "step": 62528 + }, + { + "epoch": 3.460390977443609, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 62529 + }, + { + "epoch": 3.4604511278195487, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 62530 + }, + { + "epoch": 3.4605112781954888, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0624, + "step": 62531 + }, + { + "epoch": 3.4605714285714284, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 62532 + }, + { + "epoch": 3.4606315789473685, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 62533 + }, + { + "epoch": 3.460691729323308, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 62534 + }, + { + "epoch": 3.4607518796992482, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 62535 + }, + { + "epoch": 3.460812030075188, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9358, + "step": 62536 + }, + { + "epoch": 3.460872180451128, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 62537 + }, + { + "epoch": 3.4609323308270676, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9689, + "step": 62538 + }, + { + "epoch": 3.4609924812030073, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9822, + "step": 62539 + }, + { + "epoch": 3.4610526315789474, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.0918, + "step": 62540 + }, + { + "epoch": 3.461112781954887, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 62541 + }, + { + "epoch": 3.461172932330827, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 62542 + }, + { + "epoch": 3.4612330827067668, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 0.9475, + "step": 62543 + }, + { + "epoch": 3.461293233082707, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0563, + "step": 62544 + }, + { + "epoch": 3.4613533834586465, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9043, + "step": 62545 + }, + { + "epoch": 3.4614135338345866, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9157, + "step": 62546 + }, + { + "epoch": 3.4614736842105263, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9697, + "step": 62547 + }, + { + "epoch": 3.4615338345864664, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 62548 + }, + { + "epoch": 3.461593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 62549 + }, + { + "epoch": 3.461654135338346, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1074, + "step": 62550 + }, + { + "epoch": 3.4617142857142857, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.994, + "step": 62551 + }, + { + "epoch": 3.4617744360902254, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0179, + "step": 62552 + }, + { + "epoch": 3.4618345864661655, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 62553 + }, + { + "epoch": 3.461894736842105, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 62554 + }, + { + "epoch": 3.4619548872180452, + "grad_norm": 1.1796875, + "learning_rate": 3e-05, + "loss": 1.0546, + "step": 62555 + }, + { + "epoch": 3.462015037593985, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.1179, + "step": 62556 + }, + { + "epoch": 3.462075187969925, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 62557 + }, + { + "epoch": 3.4621353383458646, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9257, + "step": 62558 + }, + { + "epoch": 3.4621954887218047, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 62559 + }, + { + "epoch": 3.4622556390977444, + "grad_norm": 1.4921875, + "learning_rate": 3e-05, + "loss": 1.0935, + "step": 62560 + }, + { + "epoch": 3.462315789473684, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 62561 + }, + { + "epoch": 3.462375939849624, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 62562 + }, + { + "epoch": 3.4624360902255638, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9794, + "step": 62563 + }, + { + "epoch": 3.462496240601504, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9943, + "step": 62564 + }, + { + "epoch": 3.4625563909774435, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9272, + "step": 62565 + }, + { + "epoch": 3.4626165413533836, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 62566 + }, + { + "epoch": 3.4626766917293232, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 62567 + }, + { + "epoch": 3.4627368421052633, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 62568 + }, + { + "epoch": 3.462796992481203, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 62569 + }, + { + "epoch": 3.4628571428571426, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9268, + "step": 62570 + }, + { + "epoch": 3.4629172932330827, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0263, + "step": 62571 + }, + { + "epoch": 3.4629774436090224, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 62572 + }, + { + "epoch": 3.4630375939849625, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9681, + "step": 62573 + }, + { + "epoch": 3.463097744360902, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9262, + "step": 62574 + }, + { + "epoch": 3.463157894736842, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 62575 + }, + { + "epoch": 3.463218045112782, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0483, + "step": 62576 + }, + { + "epoch": 3.463278195488722, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 62577 + }, + { + "epoch": 3.4633383458646616, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9238, + "step": 62578 + }, + { + "epoch": 3.4633984962406013, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0195, + "step": 62579 + }, + { + "epoch": 3.4634586466165413, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9627, + "step": 62580 + }, + { + "epoch": 3.4635187969924814, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1216, + "step": 62581 + }, + { + "epoch": 3.463578947368421, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 62582 + }, + { + "epoch": 3.4636390977443607, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 62583 + }, + { + "epoch": 3.463699248120301, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 62584 + }, + { + "epoch": 3.4637593984962405, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 62585 + }, + { + "epoch": 3.4638195488721806, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.923, + "step": 62586 + }, + { + "epoch": 3.4638796992481202, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9234, + "step": 62587 + }, + { + "epoch": 3.4639398496240603, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 62588 + }, + { + "epoch": 3.464, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.9361, + "step": 62589 + }, + { + "epoch": 3.46406015037594, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 62590 + }, + { + "epoch": 3.4641203007518797, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 62591 + }, + { + "epoch": 3.4641804511278194, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.1093, + "step": 62592 + }, + { + "epoch": 3.4642406015037595, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9871, + "step": 62593 + }, + { + "epoch": 3.464300751879699, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 62594 + }, + { + "epoch": 3.464360902255639, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 62595 + }, + { + "epoch": 3.464421052631579, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9667, + "step": 62596 + }, + { + "epoch": 3.464481203007519, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0498, + "step": 62597 + }, + { + "epoch": 3.4645413533834586, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0465, + "step": 62598 + }, + { + "epoch": 3.4646015037593987, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 62599 + }, + { + "epoch": 3.4646616541353383, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0731, + "step": 62600 + }, + { + "epoch": 3.464721804511278, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.014, + "step": 62601 + }, + { + "epoch": 3.464781954887218, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.1299, + "step": 62602 + }, + { + "epoch": 3.4648421052631577, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9675, + "step": 62603 + }, + { + "epoch": 3.464902255639098, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 62604 + }, + { + "epoch": 3.4649624060150375, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 62605 + }, + { + "epoch": 3.4650225563909776, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0209, + "step": 62606 + }, + { + "epoch": 3.465082706766917, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.1153, + "step": 62607 + }, + { + "epoch": 3.4651428571428573, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0869, + "step": 62608 + }, + { + "epoch": 3.465203007518797, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.1296, + "step": 62609 + }, + { + "epoch": 3.4652631578947366, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0801, + "step": 62610 + }, + { + "epoch": 3.4653233082706767, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9133, + "step": 62611 + }, + { + "epoch": 3.465383458646617, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9628, + "step": 62612 + }, + { + "epoch": 3.4654436090225564, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 62613 + }, + { + "epoch": 3.465503759398496, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9315, + "step": 62614 + }, + { + "epoch": 3.465563909774436, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0902, + "step": 62615 + }, + { + "epoch": 3.465624060150376, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9548, + "step": 62616 + }, + { + "epoch": 3.465684210526316, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9023, + "step": 62617 + }, + { + "epoch": 3.4657443609022556, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 62618 + }, + { + "epoch": 3.4658045112781957, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9129, + "step": 62619 + }, + { + "epoch": 3.4658646616541353, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9068, + "step": 62620 + }, + { + "epoch": 3.4659248120300754, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0674, + "step": 62621 + }, + { + "epoch": 3.465984962406015, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9417, + "step": 62622 + }, + { + "epoch": 3.4660451127819547, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 62623 + }, + { + "epoch": 3.466105263157895, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 0.9645, + "step": 62624 + }, + { + "epoch": 3.4661654135338344, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 62625 + }, + { + "epoch": 3.4662255639097745, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 62626 + }, + { + "epoch": 3.466285714285714, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 1.108, + "step": 62627 + }, + { + "epoch": 3.4663458646616543, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 62628 + }, + { + "epoch": 3.466406015037594, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 62629 + }, + { + "epoch": 3.466466165413534, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 62630 + }, + { + "epoch": 3.4665263157894737, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9082, + "step": 62631 + }, + { + "epoch": 3.4665864661654133, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 62632 + }, + { + "epoch": 3.4666466165413534, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 62633 + }, + { + "epoch": 3.466706766917293, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0938, + "step": 62634 + }, + { + "epoch": 3.466766917293233, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.1405, + "step": 62635 + }, + { + "epoch": 3.466827067669173, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 62636 + }, + { + "epoch": 3.466887218045113, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 62637 + }, + { + "epoch": 3.4669473684210526, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0185, + "step": 62638 + }, + { + "epoch": 3.4670075187969926, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9333, + "step": 62639 + }, + { + "epoch": 3.4670676691729323, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 62640 + }, + { + "epoch": 3.467127819548872, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 62641 + }, + { + "epoch": 3.467187969924812, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 62642 + }, + { + "epoch": 3.4672481203007517, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0255, + "step": 62643 + }, + { + "epoch": 3.467308270676692, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 62644 + }, + { + "epoch": 3.4673684210526314, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 62645 + }, + { + "epoch": 3.4674285714285715, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0059, + "step": 62646 + }, + { + "epoch": 3.467488721804511, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 62647 + }, + { + "epoch": 3.4675488721804513, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 62648 + }, + { + "epoch": 3.467609022556391, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0931, + "step": 62649 + }, + { + "epoch": 3.467669172932331, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0063, + "step": 62650 + }, + { + "epoch": 3.4677293233082707, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 62651 + }, + { + "epoch": 3.4677894736842108, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9233, + "step": 62652 + }, + { + "epoch": 3.4678496240601504, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9624, + "step": 62653 + }, + { + "epoch": 3.46790977443609, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 62654 + }, + { + "epoch": 3.46796992481203, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0093, + "step": 62655 + }, + { + "epoch": 3.46803007518797, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0248, + "step": 62656 + }, + { + "epoch": 3.46809022556391, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0892, + "step": 62657 + }, + { + "epoch": 3.4681503759398495, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.1165, + "step": 62658 + }, + { + "epoch": 3.4682105263157896, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.951, + "step": 62659 + }, + { + "epoch": 3.4682706766917293, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0995, + "step": 62660 + }, + { + "epoch": 3.4683308270676694, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 62661 + }, + { + "epoch": 3.468390977443609, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 62662 + }, + { + "epoch": 3.4684511278195487, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 62663 + }, + { + "epoch": 3.4685112781954888, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 62664 + }, + { + "epoch": 3.4685714285714284, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 62665 + }, + { + "epoch": 3.4686315789473685, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 62666 + }, + { + "epoch": 3.468691729323308, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9441, + "step": 62667 + }, + { + "epoch": 3.4687518796992483, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9688, + "step": 62668 + }, + { + "epoch": 3.468812030075188, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 62669 + }, + { + "epoch": 3.468872180451128, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9738, + "step": 62670 + }, + { + "epoch": 3.4689323308270676, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 62671 + }, + { + "epoch": 3.4689924812030073, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9621, + "step": 62672 + }, + { + "epoch": 3.4690526315789474, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0297, + "step": 62673 + }, + { + "epoch": 3.469112781954887, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.2207, + "step": 62674 + }, + { + "epoch": 3.469172932330827, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 62675 + }, + { + "epoch": 3.469233082706767, + "grad_norm": 0.490234375, + "learning_rate": 3e-05, + "loss": 0.9191, + "step": 62676 + }, + { + "epoch": 3.469293233082707, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.8716, + "step": 62677 + }, + { + "epoch": 3.4693533834586465, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9189, + "step": 62678 + }, + { + "epoch": 3.4694135338345866, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 62679 + }, + { + "epoch": 3.4694736842105263, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0404, + "step": 62680 + }, + { + "epoch": 3.4695338345864664, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 62681 + }, + { + "epoch": 3.469593984962406, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 62682 + }, + { + "epoch": 3.469654135338346, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9696, + "step": 62683 + }, + { + "epoch": 3.4697142857142858, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 62684 + }, + { + "epoch": 3.4697744360902254, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9808, + "step": 62685 + }, + { + "epoch": 3.4698345864661655, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 62686 + }, + { + "epoch": 3.469894736842105, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9363, + "step": 62687 + }, + { + "epoch": 3.4699548872180452, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9431, + "step": 62688 + }, + { + "epoch": 3.470015037593985, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 62689 + }, + { + "epoch": 3.470075187969925, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 62690 + }, + { + "epoch": 3.4701353383458646, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 62691 + }, + { + "epoch": 3.4701954887218047, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9386, + "step": 62692 + }, + { + "epoch": 3.4702556390977444, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0705, + "step": 62693 + }, + { + "epoch": 3.470315789473684, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9906, + "step": 62694 + }, + { + "epoch": 3.470375939849624, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 62695 + }, + { + "epoch": 3.4704360902255638, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.074, + "step": 62696 + }, + { + "epoch": 3.470496240601504, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 62697 + }, + { + "epoch": 3.4705563909774435, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 62698 + }, + { + "epoch": 3.4706165413533836, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 62699 + }, + { + "epoch": 3.4706766917293232, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0475, + "step": 62700 + }, + { + "epoch": 3.4707368421052633, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9267, + "step": 62701 + }, + { + "epoch": 3.470796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0824, + "step": 62702 + }, + { + "epoch": 3.4708571428571426, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 62703 + }, + { + "epoch": 3.4709172932330827, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.019, + "step": 62704 + }, + { + "epoch": 3.4709774436090224, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1551, + "step": 62705 + }, + { + "epoch": 3.4710375939849625, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1859, + "step": 62706 + }, + { + "epoch": 3.471097744360902, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 62707 + }, + { + "epoch": 3.471157894736842, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0708, + "step": 62708 + }, + { + "epoch": 3.471218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 62709 + }, + { + "epoch": 3.471278195488722, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 62710 + }, + { + "epoch": 3.4713383458646616, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 62711 + }, + { + "epoch": 3.4713984962406013, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 62712 + }, + { + "epoch": 3.4714586466165414, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0497, + "step": 62713 + }, + { + "epoch": 3.4715187969924814, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 62714 + }, + { + "epoch": 3.471578947368421, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9229, + "step": 62715 + }, + { + "epoch": 3.4716390977443607, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 62716 + }, + { + "epoch": 3.471699248120301, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 62717 + }, + { + "epoch": 3.4717593984962405, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0217, + "step": 62718 + }, + { + "epoch": 3.4718195488721806, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 62719 + }, + { + "epoch": 3.4718796992481202, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0406, + "step": 62720 + }, + { + "epoch": 3.4719398496240603, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0377, + "step": 62721 + }, + { + "epoch": 3.472, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 62722 + }, + { + "epoch": 3.47206015037594, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 62723 + }, + { + "epoch": 3.4721203007518797, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9538, + "step": 62724 + }, + { + "epoch": 3.4721804511278194, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0747, + "step": 62725 + }, + { + "epoch": 3.4722406015037595, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0919, + "step": 62726 + }, + { + "epoch": 3.472300751879699, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0062, + "step": 62727 + }, + { + "epoch": 3.472360902255639, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9703, + "step": 62728 + }, + { + "epoch": 3.472421052631579, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0256, + "step": 62729 + }, + { + "epoch": 3.472481203007519, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0964, + "step": 62730 + }, + { + "epoch": 3.4725413533834586, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 62731 + }, + { + "epoch": 3.4726015037593987, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.962, + "step": 62732 + }, + { + "epoch": 3.4726616541353383, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9576, + "step": 62733 + }, + { + "epoch": 3.472721804511278, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 62734 + }, + { + "epoch": 3.472781954887218, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0545, + "step": 62735 + }, + { + "epoch": 3.4728421052631577, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.1255, + "step": 62736 + }, + { + "epoch": 3.472902255639098, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 62737 + }, + { + "epoch": 3.4729624060150375, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 62738 + }, + { + "epoch": 3.4730225563909776, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.949, + "step": 62739 + }, + { + "epoch": 3.473082706766917, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9419, + "step": 62740 + }, + { + "epoch": 3.4731428571428573, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0412, + "step": 62741 + }, + { + "epoch": 3.473203007518797, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 1.0134, + "step": 62742 + }, + { + "epoch": 3.4732631578947366, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9509, + "step": 62743 + }, + { + "epoch": 3.4733233082706767, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9163, + "step": 62744 + }, + { + "epoch": 3.4733834586466164, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 62745 + }, + { + "epoch": 3.4734436090225564, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0126, + "step": 62746 + }, + { + "epoch": 3.473503759398496, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 62747 + }, + { + "epoch": 3.473563909774436, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9218, + "step": 62748 + }, + { + "epoch": 3.473624060150376, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0136, + "step": 62749 + }, + { + "epoch": 3.473684210526316, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 62750 + }, + { + "epoch": 3.4737443609022556, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0001, + "step": 62751 + }, + { + "epoch": 3.4738045112781957, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9947, + "step": 62752 + }, + { + "epoch": 3.4738646616541353, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 62753 + }, + { + "epoch": 3.4739248120300754, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 62754 + }, + { + "epoch": 3.473984962406015, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0292, + "step": 62755 + }, + { + "epoch": 3.4740451127819547, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0481, + "step": 62756 + }, + { + "epoch": 3.474105263157895, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.955, + "step": 62757 + }, + { + "epoch": 3.4741654135338345, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 62758 + }, + { + "epoch": 3.4742255639097746, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 62759 + }, + { + "epoch": 3.474285714285714, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9323, + "step": 62760 + }, + { + "epoch": 3.4743458646616543, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9957, + "step": 62761 + }, + { + "epoch": 3.474406015037594, + "grad_norm": 1.09375, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 62762 + }, + { + "epoch": 3.474466165413534, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 62763 + }, + { + "epoch": 3.4745263157894737, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.873, + "step": 62764 + }, + { + "epoch": 3.4745864661654133, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 62765 + }, + { + "epoch": 3.4746466165413534, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 62766 + }, + { + "epoch": 3.474706766917293, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0449, + "step": 62767 + }, + { + "epoch": 3.474766917293233, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 62768 + }, + { + "epoch": 3.474827067669173, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 62769 + }, + { + "epoch": 3.474887218045113, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9526, + "step": 62770 + }, + { + "epoch": 3.4749473684210526, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9182, + "step": 62771 + }, + { + "epoch": 3.4750075187969927, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 62772 + }, + { + "epoch": 3.4750676691729323, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.8784, + "step": 62773 + }, + { + "epoch": 3.475127819548872, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0196, + "step": 62774 + }, + { + "epoch": 3.475187969924812, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9763, + "step": 62775 + }, + { + "epoch": 3.4752481203007517, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1239, + "step": 62776 + }, + { + "epoch": 3.475308270676692, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 62777 + }, + { + "epoch": 3.4753684210526314, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 0.9726, + "step": 62778 + }, + { + "epoch": 3.4754285714285715, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0249, + "step": 62779 + }, + { + "epoch": 3.475488721804511, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 62780 + }, + { + "epoch": 3.4755488721804513, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 62781 + }, + { + "epoch": 3.475609022556391, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.023, + "step": 62782 + }, + { + "epoch": 3.475669172932331, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 62783 + }, + { + "epoch": 3.4757293233082707, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0182, + "step": 62784 + }, + { + "epoch": 3.4757894736842108, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0328, + "step": 62785 + }, + { + "epoch": 3.4758496240601504, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 62786 + }, + { + "epoch": 3.47590977443609, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0308, + "step": 62787 + }, + { + "epoch": 3.47596992481203, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 62788 + }, + { + "epoch": 3.47603007518797, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 62789 + }, + { + "epoch": 3.47609022556391, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 62790 + }, + { + "epoch": 3.4761503759398495, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.041, + "step": 62791 + }, + { + "epoch": 3.4762105263157896, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.028, + "step": 62792 + }, + { + "epoch": 3.4762706766917293, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 62793 + }, + { + "epoch": 3.4763308270676694, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0303, + "step": 62794 + }, + { + "epoch": 3.476390977443609, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0554, + "step": 62795 + }, + { + "epoch": 3.4764511278195487, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9915, + "step": 62796 + }, + { + "epoch": 3.4765112781954888, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 62797 + }, + { + "epoch": 3.4765714285714284, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 62798 + }, + { + "epoch": 3.4766315789473685, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9854, + "step": 62799 + }, + { + "epoch": 3.476691729323308, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 1.0153, + "step": 62800 + }, + { + "epoch": 3.4767518796992483, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9739, + "step": 62801 + }, + { + "epoch": 3.476812030075188, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 62802 + }, + { + "epoch": 3.476872180451128, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 0.9583, + "step": 62803 + }, + { + "epoch": 3.4769323308270677, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 62804 + }, + { + "epoch": 3.4769924812030073, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 62805 + }, + { + "epoch": 3.4770526315789474, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0701, + "step": 62806 + }, + { + "epoch": 3.477112781954887, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 62807 + }, + { + "epoch": 3.477172932330827, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 62808 + }, + { + "epoch": 3.477233082706767, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 62809 + }, + { + "epoch": 3.477293233082707, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 62810 + }, + { + "epoch": 3.4773533834586465, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.1513, + "step": 62811 + }, + { + "epoch": 3.4774135338345866, + "grad_norm": 0.5078125, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 62812 + }, + { + "epoch": 3.4774736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9594, + "step": 62813 + }, + { + "epoch": 3.477533834586466, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0615, + "step": 62814 + }, + { + "epoch": 3.477593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 62815 + }, + { + "epoch": 3.477654135338346, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9825, + "step": 62816 + }, + { + "epoch": 3.4777142857142858, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0539, + "step": 62817 + }, + { + "epoch": 3.4777744360902254, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.8982, + "step": 62818 + }, + { + "epoch": 3.4778345864661655, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0588, + "step": 62819 + }, + { + "epoch": 3.477894736842105, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 62820 + }, + { + "epoch": 3.4779548872180452, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 62821 + }, + { + "epoch": 3.478015037593985, + "grad_norm": 0.5625, + "learning_rate": 3e-05, + "loss": 1.0114, + "step": 62822 + }, + { + "epoch": 3.478075187969925, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 62823 + }, + { + "epoch": 3.4781353383458646, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 62824 + }, + { + "epoch": 3.4781954887218047, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9937, + "step": 62825 + }, + { + "epoch": 3.4782556390977444, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9333, + "step": 62826 + }, + { + "epoch": 3.478315789473684, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 62827 + }, + { + "epoch": 3.478375939849624, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 62828 + }, + { + "epoch": 3.4784360902255638, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 62829 + }, + { + "epoch": 3.478496240601504, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9963, + "step": 62830 + }, + { + "epoch": 3.4785563909774435, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9773, + "step": 62831 + }, + { + "epoch": 3.4786165413533836, + "grad_norm": 0.5234375, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 62832 + }, + { + "epoch": 3.4786766917293233, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 62833 + }, + { + "epoch": 3.4787368421052633, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9478, + "step": 62834 + }, + { + "epoch": 3.478796992481203, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.978, + "step": 62835 + }, + { + "epoch": 3.4788571428571426, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9184, + "step": 62836 + }, + { + "epoch": 3.4789172932330827, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0319, + "step": 62837 + }, + { + "epoch": 3.4789774436090224, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.1574, + "step": 62838 + }, + { + "epoch": 3.4790375939849625, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9862, + "step": 62839 + }, + { + "epoch": 3.479097744360902, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 62840 + }, + { + "epoch": 3.4791578947368422, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 62841 + }, + { + "epoch": 3.479218045112782, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 62842 + }, + { + "epoch": 3.479278195488722, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9877, + "step": 62843 + }, + { + "epoch": 3.4793383458646616, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 62844 + }, + { + "epoch": 3.4793984962406013, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9118, + "step": 62845 + }, + { + "epoch": 3.4794586466165414, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9997, + "step": 62846 + }, + { + "epoch": 3.479518796992481, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0069, + "step": 62847 + }, + { + "epoch": 3.479578947368421, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0551, + "step": 62848 + }, + { + "epoch": 3.4796390977443608, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 62849 + }, + { + "epoch": 3.479699248120301, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0597, + "step": 62850 + }, + { + "epoch": 3.4797593984962405, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0273, + "step": 62851 + }, + { + "epoch": 3.4798195488721806, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 62852 + }, + { + "epoch": 3.4798796992481202, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9291, + "step": 62853 + }, + { + "epoch": 3.4799398496240603, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 62854 + }, + { + "epoch": 3.48, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9635, + "step": 62855 + }, + { + "epoch": 3.48006015037594, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 62856 + }, + { + "epoch": 3.4801203007518797, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 62857 + }, + { + "epoch": 3.4801804511278194, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 62858 + }, + { + "epoch": 3.4802406015037595, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9612, + "step": 62859 + }, + { + "epoch": 3.480300751879699, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9489, + "step": 62860 + }, + { + "epoch": 3.480360902255639, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0087, + "step": 62861 + }, + { + "epoch": 3.480421052631579, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9367, + "step": 62862 + }, + { + "epoch": 3.480481203007519, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0693, + "step": 62863 + }, + { + "epoch": 3.4805413533834586, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 62864 + }, + { + "epoch": 3.4806015037593987, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0766, + "step": 62865 + }, + { + "epoch": 3.4806616541353383, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 62866 + }, + { + "epoch": 3.480721804511278, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 62867 + }, + { + "epoch": 3.480781954887218, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0301, + "step": 62868 + }, + { + "epoch": 3.4808421052631577, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 62869 + }, + { + "epoch": 3.480902255639098, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 62870 + }, + { + "epoch": 3.4809624060150375, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 62871 + }, + { + "epoch": 3.4810225563909776, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.8952, + "step": 62872 + }, + { + "epoch": 3.4810827067669172, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0614, + "step": 62873 + }, + { + "epoch": 3.4811428571428573, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9427, + "step": 62874 + }, + { + "epoch": 3.481203007518797, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9819, + "step": 62875 + }, + { + "epoch": 3.4812631578947366, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9252, + "step": 62876 + }, + { + "epoch": 3.4813233082706767, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 62877 + }, + { + "epoch": 3.4813834586466164, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0379, + "step": 62878 + }, + { + "epoch": 3.4814436090225565, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 62879 + }, + { + "epoch": 3.481503759398496, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 62880 + }, + { + "epoch": 3.481563909774436, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.941, + "step": 62881 + }, + { + "epoch": 3.481624060150376, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1217, + "step": 62882 + }, + { + "epoch": 3.481684210526316, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.1745, + "step": 62883 + }, + { + "epoch": 3.4817443609022556, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 62884 + }, + { + "epoch": 3.4818045112781957, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.8869, + "step": 62885 + }, + { + "epoch": 3.4818646616541353, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9704, + "step": 62886 + }, + { + "epoch": 3.4819248120300754, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9044, + "step": 62887 + }, + { + "epoch": 3.481984962406015, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9104, + "step": 62888 + }, + { + "epoch": 3.4820451127819547, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0347, + "step": 62889 + }, + { + "epoch": 3.482105263157895, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9377, + "step": 62890 + }, + { + "epoch": 3.4821654135338345, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 62891 + }, + { + "epoch": 3.4822255639097746, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9482, + "step": 62892 + }, + { + "epoch": 3.482285714285714, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 62893 + }, + { + "epoch": 3.4823458646616543, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0543, + "step": 62894 + }, + { + "epoch": 3.482406015037594, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 62895 + }, + { + "epoch": 3.482466165413534, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9613, + "step": 62896 + }, + { + "epoch": 3.4825263157894737, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 0.9713, + "step": 62897 + }, + { + "epoch": 3.4825864661654133, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9466, + "step": 62898 + }, + { + "epoch": 3.4826466165413534, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 62899 + }, + { + "epoch": 3.482706766917293, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 62900 + }, + { + "epoch": 3.482766917293233, + "grad_norm": 0.51171875, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 62901 + }, + { + "epoch": 3.482827067669173, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9672, + "step": 62902 + }, + { + "epoch": 3.482887218045113, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 62903 + }, + { + "epoch": 3.4829473684210526, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.1999, + "step": 62904 + }, + { + "epoch": 3.4830075187969927, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9846, + "step": 62905 + }, + { + "epoch": 3.4830676691729323, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9652, + "step": 62906 + }, + { + "epoch": 3.483127819548872, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 62907 + }, + { + "epoch": 3.483187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0276, + "step": 62908 + }, + { + "epoch": 3.4832481203007517, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 62909 + }, + { + "epoch": 3.483308270676692, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 62910 + }, + { + "epoch": 3.4833684210526314, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0191, + "step": 62911 + }, + { + "epoch": 3.4834285714285715, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0049, + "step": 62912 + }, + { + "epoch": 3.483488721804511, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.1108, + "step": 62913 + }, + { + "epoch": 3.4835488721804513, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 62914 + }, + { + "epoch": 3.483609022556391, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9488, + "step": 62915 + }, + { + "epoch": 3.4836691729323306, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 62916 + }, + { + "epoch": 3.4837293233082707, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9453, + "step": 62917 + }, + { + "epoch": 3.4837894736842108, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0067, + "step": 62918 + }, + { + "epoch": 3.4838496240601504, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 62919 + }, + { + "epoch": 3.48390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9338, + "step": 62920 + }, + { + "epoch": 3.48396992481203, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 62921 + }, + { + "epoch": 3.48403007518797, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 62922 + }, + { + "epoch": 3.48409022556391, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9948, + "step": 62923 + }, + { + "epoch": 3.4841503759398496, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 62924 + }, + { + "epoch": 3.4842105263157896, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0713, + "step": 62925 + }, + { + "epoch": 3.4842706766917293, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9138, + "step": 62926 + }, + { + "epoch": 3.4843308270676694, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 62927 + }, + { + "epoch": 3.484390977443609, + "grad_norm": 0.953125, + "learning_rate": 3e-05, + "loss": 1.031, + "step": 62928 + }, + { + "epoch": 3.4844511278195487, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0894, + "step": 62929 + }, + { + "epoch": 3.484511278195489, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 62930 + }, + { + "epoch": 3.4845714285714284, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 62931 + }, + { + "epoch": 3.4846315789473685, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0472, + "step": 62932 + }, + { + "epoch": 3.484691729323308, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9011, + "step": 62933 + }, + { + "epoch": 3.4847518796992483, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9565, + "step": 62934 + }, + { + "epoch": 3.484812030075188, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 62935 + }, + { + "epoch": 3.484872180451128, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 62936 + }, + { + "epoch": 3.4849323308270677, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 62937 + }, + { + "epoch": 3.4849924812030073, + "grad_norm": 1.328125, + "learning_rate": 3e-05, + "loss": 1.1977, + "step": 62938 + }, + { + "epoch": 3.4850526315789474, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.101, + "step": 62939 + }, + { + "epoch": 3.485112781954887, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0613, + "step": 62940 + }, + { + "epoch": 3.485172932330827, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.1556, + "step": 62941 + }, + { + "epoch": 3.485233082706767, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 62942 + }, + { + "epoch": 3.485293233082707, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.013, + "step": 62943 + }, + { + "epoch": 3.4853533834586465, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0645, + "step": 62944 + }, + { + "epoch": 3.4854135338345866, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9311, + "step": 62945 + }, + { + "epoch": 3.4854736842105263, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 62946 + }, + { + "epoch": 3.485533834586466, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0162, + "step": 62947 + }, + { + "epoch": 3.485593984962406, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0745, + "step": 62948 + }, + { + "epoch": 3.4856541353383457, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9649, + "step": 62949 + }, + { + "epoch": 3.4857142857142858, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 62950 + }, + { + "epoch": 3.4857744360902254, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9066, + "step": 62951 + }, + { + "epoch": 3.4858345864661655, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9389, + "step": 62952 + }, + { + "epoch": 3.485894736842105, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9456, + "step": 62953 + }, + { + "epoch": 3.4859548872180453, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1246, + "step": 62954 + }, + { + "epoch": 3.486015037593985, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.0962, + "step": 62955 + }, + { + "epoch": 3.486075187969925, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9484, + "step": 62956 + }, + { + "epoch": 3.4861353383458646, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.9321, + "step": 62957 + }, + { + "epoch": 3.4861954887218047, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 62958 + }, + { + "epoch": 3.4862556390977444, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 62959 + }, + { + "epoch": 3.486315789473684, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 62960 + }, + { + "epoch": 3.486375939849624, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9882, + "step": 62961 + }, + { + "epoch": 3.4864360902255638, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 62962 + }, + { + "epoch": 3.486496240601504, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 62963 + }, + { + "epoch": 3.4865563909774435, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9273, + "step": 62964 + }, + { + "epoch": 3.4866165413533836, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.9137, + "step": 62965 + }, + { + "epoch": 3.4866766917293233, + "grad_norm": 0.5546875, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 62966 + }, + { + "epoch": 3.4867368421052634, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 62967 + }, + { + "epoch": 3.486796992481203, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.04, + "step": 62968 + }, + { + "epoch": 3.4868571428571427, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9088, + "step": 62969 + }, + { + "epoch": 3.4869172932330827, + "grad_norm": 1.3828125, + "learning_rate": 3e-05, + "loss": 1.0963, + "step": 62970 + }, + { + "epoch": 3.4869774436090224, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 62971 + }, + { + "epoch": 3.4870375939849625, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.8276, + "step": 62972 + }, + { + "epoch": 3.487097744360902, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.037, + "step": 62973 + }, + { + "epoch": 3.4871578947368422, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0075, + "step": 62974 + }, + { + "epoch": 3.487218045112782, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 62975 + }, + { + "epoch": 3.487278195488722, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9318, + "step": 62976 + }, + { + "epoch": 3.4873383458646616, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 62977 + }, + { + "epoch": 3.4873984962406013, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0364, + "step": 62978 + }, + { + "epoch": 3.4874586466165414, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9897, + "step": 62979 + }, + { + "epoch": 3.487518796992481, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9831, + "step": 62980 + }, + { + "epoch": 3.487578947368421, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0872, + "step": 62981 + }, + { + "epoch": 3.4876390977443608, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0079, + "step": 62982 + }, + { + "epoch": 3.487699248120301, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 0.9962, + "step": 62983 + }, + { + "epoch": 3.4877593984962405, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0072, + "step": 62984 + }, + { + "epoch": 3.4878195488721806, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9474, + "step": 62985 + }, + { + "epoch": 3.4878796992481202, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9836, + "step": 62986 + }, + { + "epoch": 3.4879398496240603, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9789, + "step": 62987 + }, + { + "epoch": 3.488, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.1346, + "step": 62988 + }, + { + "epoch": 3.48806015037594, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9328, + "step": 62989 + }, + { + "epoch": 3.4881203007518797, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0418, + "step": 62990 + }, + { + "epoch": 3.4881804511278194, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0602, + "step": 62991 + }, + { + "epoch": 3.4882406015037595, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0697, + "step": 62992 + }, + { + "epoch": 3.488300751879699, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9468, + "step": 62993 + }, + { + "epoch": 3.488360902255639, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9052, + "step": 62994 + }, + { + "epoch": 3.488421052631579, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0515, + "step": 62995 + }, + { + "epoch": 3.488481203007519, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0971, + "step": 62996 + }, + { + "epoch": 3.4885413533834586, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.2812, + "step": 62997 + }, + { + "epoch": 3.4886015037593987, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9533, + "step": 62998 + }, + { + "epoch": 3.4886616541353384, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 62999 + }, + { + "epoch": 3.488721804511278, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9164, + "step": 63000 + }, + { + "epoch": 3.488781954887218, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.049, + "step": 63001 + }, + { + "epoch": 3.4888421052631577, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0968, + "step": 63002 + }, + { + "epoch": 3.488902255639098, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0728, + "step": 63003 + }, + { + "epoch": 3.4889624060150375, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0581, + "step": 63004 + }, + { + "epoch": 3.4890225563909776, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0082, + "step": 63005 + }, + { + "epoch": 3.4890827067669172, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 63006 + }, + { + "epoch": 3.4891428571428573, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 63007 + }, + { + "epoch": 3.489203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 63008 + }, + { + "epoch": 3.4892631578947366, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 63009 + }, + { + "epoch": 3.4893233082706767, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 63010 + }, + { + "epoch": 3.4893834586466164, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 63011 + }, + { + "epoch": 3.4894436090225565, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.1036, + "step": 63012 + }, + { + "epoch": 3.489503759398496, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0167, + "step": 63013 + }, + { + "epoch": 3.489563909774436, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0657, + "step": 63014 + }, + { + "epoch": 3.489624060150376, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9401, + "step": 63015 + }, + { + "epoch": 3.489684210526316, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0687, + "step": 63016 + }, + { + "epoch": 3.4897443609022556, + "grad_norm": 1.15625, + "learning_rate": 3e-05, + "loss": 0.991, + "step": 63017 + }, + { + "epoch": 3.4898045112781952, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 63018 + }, + { + "epoch": 3.4898646616541353, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 63019 + }, + { + "epoch": 3.4899248120300754, + "grad_norm": 1.3359375, + "learning_rate": 3e-05, + "loss": 0.9911, + "step": 63020 + }, + { + "epoch": 3.489984962406015, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 0.9267, + "step": 63021 + }, + { + "epoch": 3.4900451127819547, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 63022 + }, + { + "epoch": 3.490105263157895, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0738, + "step": 63023 + }, + { + "epoch": 3.4901654135338345, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 63024 + }, + { + "epoch": 3.4902255639097746, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9595, + "step": 63025 + }, + { + "epoch": 3.490285714285714, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 63026 + }, + { + "epoch": 3.4903458646616543, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9856, + "step": 63027 + }, + { + "epoch": 3.490406015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 63028 + }, + { + "epoch": 3.490466165413534, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.993, + "step": 63029 + }, + { + "epoch": 3.4905263157894737, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9756, + "step": 63030 + }, + { + "epoch": 3.4905864661654133, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9239, + "step": 63031 + }, + { + "epoch": 3.4906466165413534, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9381, + "step": 63032 + }, + { + "epoch": 3.490706766917293, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 1.0871, + "step": 63033 + }, + { + "epoch": 3.490766917293233, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 63034 + }, + { + "epoch": 3.490827067669173, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 63035 + }, + { + "epoch": 3.490887218045113, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0344, + "step": 63036 + }, + { + "epoch": 3.4909473684210526, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 63037 + }, + { + "epoch": 3.4910075187969927, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 63038 + }, + { + "epoch": 3.4910676691729323, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9342, + "step": 63039 + }, + { + "epoch": 3.491127819548872, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9251, + "step": 63040 + }, + { + "epoch": 3.491187969924812, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0015, + "step": 63041 + }, + { + "epoch": 3.4912481203007517, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0144, + "step": 63042 + }, + { + "epoch": 3.491308270676692, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.891, + "step": 63043 + }, + { + "epoch": 3.4913684210526315, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9546, + "step": 63044 + }, + { + "epoch": 3.4914285714285715, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 63045 + }, + { + "epoch": 3.491488721804511, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9992, + "step": 63046 + }, + { + "epoch": 3.4915488721804513, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9352, + "step": 63047 + }, + { + "epoch": 3.491609022556391, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0237, + "step": 63048 + }, + { + "epoch": 3.4916691729323306, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9682, + "step": 63049 + }, + { + "epoch": 3.4917293233082707, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 0.9609, + "step": 63050 + }, + { + "epoch": 3.4917894736842103, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 63051 + }, + { + "epoch": 3.4918496240601504, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 63052 + }, + { + "epoch": 3.49190977443609, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9335, + "step": 63053 + }, + { + "epoch": 3.49196992481203, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.1505, + "step": 63054 + }, + { + "epoch": 3.49203007518797, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.989, + "step": 63055 + }, + { + "epoch": 3.49209022556391, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9881, + "step": 63056 + }, + { + "epoch": 3.4921503759398496, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0719, + "step": 63057 + }, + { + "epoch": 3.4922105263157897, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9691, + "step": 63058 + }, + { + "epoch": 3.4922706766917293, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0052, + "step": 63059 + }, + { + "epoch": 3.4923308270676694, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 63060 + }, + { + "epoch": 3.492390977443609, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 63061 + }, + { + "epoch": 3.4924511278195487, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9406, + "step": 63062 + }, + { + "epoch": 3.492511278195489, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0817, + "step": 63063 + }, + { + "epoch": 3.4925714285714284, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 63064 + }, + { + "epoch": 3.4926315789473685, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0112, + "step": 63065 + }, + { + "epoch": 3.492691729323308, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 63066 + }, + { + "epoch": 3.4927518796992483, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 63067 + }, + { + "epoch": 3.492812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0002, + "step": 63068 + }, + { + "epoch": 3.492872180451128, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9378, + "step": 63069 + }, + { + "epoch": 3.4929323308270677, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9443, + "step": 63070 + }, + { + "epoch": 3.4929924812030073, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.8847, + "step": 63071 + }, + { + "epoch": 3.4930526315789474, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9945, + "step": 63072 + }, + { + "epoch": 3.493112781954887, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.1051, + "step": 63073 + }, + { + "epoch": 3.493172932330827, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 63074 + }, + { + "epoch": 3.493233082706767, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 0.9541, + "step": 63075 + }, + { + "epoch": 3.493293233082707, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0704, + "step": 63076 + }, + { + "epoch": 3.4933533834586465, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 63077 + }, + { + "epoch": 3.4934135338345866, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 63078 + }, + { + "epoch": 3.4934736842105263, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9471, + "step": 63079 + }, + { + "epoch": 3.493533834586466, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 63080 + }, + { + "epoch": 3.493593984962406, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 63081 + }, + { + "epoch": 3.4936541353383457, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 63082 + }, + { + "epoch": 3.4937142857142858, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0012, + "step": 63083 + }, + { + "epoch": 3.4937744360902254, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.8995, + "step": 63084 + }, + { + "epoch": 3.4938345864661655, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 63085 + }, + { + "epoch": 3.493894736842105, + "grad_norm": 1.4296875, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 63086 + }, + { + "epoch": 3.4939548872180453, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0243, + "step": 63087 + }, + { + "epoch": 3.494015037593985, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 0.9629, + "step": 63088 + }, + { + "epoch": 3.494075187969925, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9012, + "step": 63089 + }, + { + "epoch": 3.4941353383458646, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0921, + "step": 63090 + }, + { + "epoch": 3.4941954887218047, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9089, + "step": 63091 + }, + { + "epoch": 3.4942556390977444, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0372, + "step": 63092 + }, + { + "epoch": 3.494315789473684, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0654, + "step": 63093 + }, + { + "epoch": 3.494375939849624, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0679, + "step": 63094 + }, + { + "epoch": 3.494436090225564, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0385, + "step": 63095 + }, + { + "epoch": 3.494496240601504, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 63096 + }, + { + "epoch": 3.4945563909774435, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 63097 + }, + { + "epoch": 3.4946165413533836, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.907, + "step": 63098 + }, + { + "epoch": 3.4946766917293233, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9968, + "step": 63099 + }, + { + "epoch": 3.4947368421052634, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 63100 + }, + { + "epoch": 3.494796992481203, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 63101 + }, + { + "epoch": 3.4948571428571427, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9662, + "step": 63102 + }, + { + "epoch": 3.4949172932330828, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0762, + "step": 63103 + }, + { + "epoch": 3.4949774436090224, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 0.9661, + "step": 63104 + }, + { + "epoch": 3.4950375939849625, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9131, + "step": 63105 + }, + { + "epoch": 3.495097744360902, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 1.0866, + "step": 63106 + }, + { + "epoch": 3.4951578947368422, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9226, + "step": 63107 + }, + { + "epoch": 3.495218045112782, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.1305, + "step": 63108 + }, + { + "epoch": 3.495278195488722, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 63109 + }, + { + "epoch": 3.4953383458646616, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9447, + "step": 63110 + }, + { + "epoch": 3.4953984962406013, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9818, + "step": 63111 + }, + { + "epoch": 3.4954586466165414, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 63112 + }, + { + "epoch": 3.495518796992481, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 63113 + }, + { + "epoch": 3.495578947368421, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 63114 + }, + { + "epoch": 3.4956390977443608, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 63115 + }, + { + "epoch": 3.495699248120301, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0239, + "step": 63116 + }, + { + "epoch": 3.4957593984962405, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.0038, + "step": 63117 + }, + { + "epoch": 3.4958195488721806, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 63118 + }, + { + "epoch": 3.4958796992481203, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9127, + "step": 63119 + }, + { + "epoch": 3.49593984962406, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 63120 + }, + { + "epoch": 3.496, + "grad_norm": 1.1328125, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 63121 + }, + { + "epoch": 3.49606015037594, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.11, + "step": 63122 + }, + { + "epoch": 3.4961203007518797, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9685, + "step": 63123 + }, + { + "epoch": 3.4961804511278194, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0224, + "step": 63124 + }, + { + "epoch": 3.4962406015037595, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9969, + "step": 63125 + }, + { + "epoch": 3.496300751879699, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 63126 + }, + { + "epoch": 3.4963609022556392, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0323, + "step": 63127 + }, + { + "epoch": 3.496421052631579, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.012, + "step": 63128 + }, + { + "epoch": 3.496481203007519, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1029, + "step": 63129 + }, + { + "epoch": 3.4965413533834586, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9418, + "step": 63130 + }, + { + "epoch": 3.4966015037593987, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9841, + "step": 63131 + }, + { + "epoch": 3.4966616541353384, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.058, + "step": 63132 + }, + { + "epoch": 3.496721804511278, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0106, + "step": 63133 + }, + { + "epoch": 3.496781954887218, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 63134 + }, + { + "epoch": 3.4968421052631578, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0278, + "step": 63135 + }, + { + "epoch": 3.496902255639098, + "grad_norm": 1.4375, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 63136 + }, + { + "epoch": 3.4969624060150375, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 63137 + }, + { + "epoch": 3.4970225563909776, + "grad_norm": 0.9921875, + "learning_rate": 3e-05, + "loss": 1.0584, + "step": 63138 + }, + { + "epoch": 3.4970827067669172, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9491, + "step": 63139 + }, + { + "epoch": 3.4971428571428573, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.8582, + "step": 63140 + }, + { + "epoch": 3.497203007518797, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0664, + "step": 63141 + }, + { + "epoch": 3.4972631578947366, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.954, + "step": 63142 + }, + { + "epoch": 3.4973233082706767, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.0322, + "step": 63143 + }, + { + "epoch": 3.4973834586466164, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 63144 + }, + { + "epoch": 3.4974436090225565, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.1219, + "step": 63145 + }, + { + "epoch": 3.497503759398496, + "grad_norm": 0.93359375, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 63146 + }, + { + "epoch": 3.497563909774436, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 63147 + }, + { + "epoch": 3.497624060150376, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 63148 + }, + { + "epoch": 3.497684210526316, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.0389, + "step": 63149 + }, + { + "epoch": 3.4977443609022556, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 63150 + }, + { + "epoch": 3.4978045112781952, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9379, + "step": 63151 + }, + { + "epoch": 3.4978646616541353, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0346, + "step": 63152 + }, + { + "epoch": 3.4979248120300754, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 63153 + }, + { + "epoch": 3.497984962406015, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9626, + "step": 63154 + }, + { + "epoch": 3.4980451127819547, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9614, + "step": 63155 + }, + { + "epoch": 3.498105263157895, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 63156 + }, + { + "epoch": 3.4981654135338345, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9742, + "step": 63157 + }, + { + "epoch": 3.4982255639097746, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9718, + "step": 63158 + }, + { + "epoch": 3.498285714285714, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0535, + "step": 63159 + }, + { + "epoch": 3.4983458646616543, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.985, + "step": 63160 + }, + { + "epoch": 3.498406015037594, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 63161 + }, + { + "epoch": 3.498466165413534, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9059, + "step": 63162 + }, + { + "epoch": 3.4985263157894737, + "grad_norm": 1.0390625, + "learning_rate": 3e-05, + "loss": 1.1626, + "step": 63163 + }, + { + "epoch": 3.4985864661654134, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0097, + "step": 63164 + }, + { + "epoch": 3.4986466165413534, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0111, + "step": 63165 + }, + { + "epoch": 3.498706766917293, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 63166 + }, + { + "epoch": 3.498766917293233, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.1519, + "step": 63167 + }, + { + "epoch": 3.498827067669173, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.1092, + "step": 63168 + }, + { + "epoch": 3.498887218045113, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0225, + "step": 63169 + }, + { + "epoch": 3.4989473684210526, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.016, + "step": 63170 + }, + { + "epoch": 3.4990075187969927, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9078, + "step": 63171 + }, + { + "epoch": 3.4990676691729323, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9256, + "step": 63172 + }, + { + "epoch": 3.499127819548872, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9774, + "step": 63173 + }, + { + "epoch": 3.499187969924812, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9674, + "step": 63174 + }, + { + "epoch": 3.4992481203007517, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.1126, + "step": 63175 + }, + { + "epoch": 3.499308270676692, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 63176 + }, + { + "epoch": 3.4993684210526315, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.102, + "step": 63177 + }, + { + "epoch": 3.4994285714285716, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0218, + "step": 63178 + }, + { + "epoch": 3.499488721804511, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 63179 + }, + { + "epoch": 3.4995488721804513, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 63180 + }, + { + "epoch": 3.499609022556391, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9632, + "step": 63181 + }, + { + "epoch": 3.4996691729323306, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 63182 + }, + { + "epoch": 3.4997293233082707, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 63183 + }, + { + "epoch": 3.4997894736842103, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 1.0488, + "step": 63184 + }, + { + "epoch": 3.4998496240601504, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0429, + "step": 63185 + }, + { + "epoch": 3.49990977443609, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 63186 + }, + { + "epoch": 3.49996992481203, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 63187 + }, + { + "epoch": 3.50003007518797, + "grad_norm": 1.140625, + "learning_rate": 3e-05, + "loss": 1.0868, + "step": 63188 + }, + { + "epoch": 3.50009022556391, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9748, + "step": 63189 + }, + { + "epoch": 3.5001503759398496, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0473, + "step": 63190 + }, + { + "epoch": 3.500210526315789, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9913, + "step": 63191 + }, + { + "epoch": 3.5002706766917293, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0469, + "step": 63192 + }, + { + "epoch": 3.5003308270676694, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 63193 + }, + { + "epoch": 3.500390977443609, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0056, + "step": 63194 + }, + { + "epoch": 3.5004511278195487, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9373, + "step": 63195 + }, + { + "epoch": 3.500511278195489, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0493, + "step": 63196 + }, + { + "epoch": 3.5005714285714284, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0125, + "step": 63197 + }, + { + "epoch": 3.5006315789473685, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 63198 + }, + { + "epoch": 3.500691729323308, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9099, + "step": 63199 + }, + { + "epoch": 3.5007518796992483, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9117, + "step": 63200 + }, + { + "epoch": 3.500812030075188, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0092, + "step": 63201 + }, + { + "epoch": 3.500872180451128, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9918, + "step": 63202 + }, + { + "epoch": 3.5009323308270677, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1082, + "step": 63203 + }, + { + "epoch": 3.5009924812030073, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9727, + "step": 63204 + }, + { + "epoch": 3.5010526315789474, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 63205 + }, + { + "epoch": 3.501112781954887, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 63206 + }, + { + "epoch": 3.501172932330827, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0174, + "step": 63207 + }, + { + "epoch": 3.501233082706767, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9446, + "step": 63208 + }, + { + "epoch": 3.501293233082707, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9924, + "step": 63209 + }, + { + "epoch": 3.5013533834586466, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9436, + "step": 63210 + }, + { + "epoch": 3.5014135338345866, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.1048, + "step": 63211 + }, + { + "epoch": 3.5014736842105263, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9857, + "step": 63212 + }, + { + "epoch": 3.501533834586466, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0013, + "step": 63213 + }, + { + "epoch": 3.501593984962406, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 63214 + }, + { + "epoch": 3.501654135338346, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0804, + "step": 63215 + }, + { + "epoch": 3.501714285714286, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0341, + "step": 63216 + }, + { + "epoch": 3.5017744360902254, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 63217 + }, + { + "epoch": 3.5018345864661655, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9387, + "step": 63218 + }, + { + "epoch": 3.501894736842105, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 63219 + }, + { + "epoch": 3.5019548872180453, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9486, + "step": 63220 + }, + { + "epoch": 3.502015037593985, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9238, + "step": 63221 + }, + { + "epoch": 3.5020751879699246, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 63222 + }, + { + "epoch": 3.5021353383458647, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 63223 + }, + { + "epoch": 3.5021954887218048, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.2315, + "step": 63224 + }, + { + "epoch": 3.5022556390977444, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 63225 + }, + { + "epoch": 3.502315789473684, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9669, + "step": 63226 + }, + { + "epoch": 3.502375939849624, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 63227 + }, + { + "epoch": 3.502436090225564, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0629, + "step": 63228 + }, + { + "epoch": 3.502496240601504, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0489, + "step": 63229 + }, + { + "epoch": 3.5025563909774435, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 0.9579, + "step": 63230 + }, + { + "epoch": 3.502616541353383, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0955, + "step": 63231 + }, + { + "epoch": 3.5026766917293233, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0219, + "step": 63232 + }, + { + "epoch": 3.5027368421052634, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0238, + "step": 63233 + }, + { + "epoch": 3.502796992481203, + "grad_norm": 1.2734375, + "learning_rate": 3e-05, + "loss": 1.0993, + "step": 63234 + }, + { + "epoch": 3.5028571428571427, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0376, + "step": 63235 + }, + { + "epoch": 3.5029172932330828, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9256, + "step": 63236 + }, + { + "epoch": 3.5029774436090224, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 63237 + }, + { + "epoch": 3.5030375939849625, + "grad_norm": 1.2890625, + "learning_rate": 3e-05, + "loss": 1.0591, + "step": 63238 + }, + { + "epoch": 3.503097744360902, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0226, + "step": 63239 + }, + { + "epoch": 3.5031578947368422, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0893, + "step": 63240 + }, + { + "epoch": 3.503218045112782, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 63241 + }, + { + "epoch": 3.503278195488722, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 63242 + }, + { + "epoch": 3.5033383458646616, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 1.2503, + "step": 63243 + }, + { + "epoch": 3.5033984962406013, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 63244 + }, + { + "epoch": 3.5034586466165414, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0223, + "step": 63245 + }, + { + "epoch": 3.5035187969924815, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 63246 + }, + { + "epoch": 3.503578947368421, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0577, + "step": 63247 + }, + { + "epoch": 3.5036390977443608, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0397, + "step": 63248 + }, + { + "epoch": 3.503699248120301, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0316, + "step": 63249 + }, + { + "epoch": 3.5037593984962405, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9219, + "step": 63250 + }, + { + "epoch": 3.5038195488721806, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.946, + "step": 63251 + }, + { + "epoch": 3.5038796992481203, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 63252 + }, + { + "epoch": 3.50393984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9243, + "step": 63253 + }, + { + "epoch": 3.504, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9171, + "step": 63254 + }, + { + "epoch": 3.50406015037594, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 63255 + }, + { + "epoch": 3.5041203007518797, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9776, + "step": 63256 + }, + { + "epoch": 3.5041804511278194, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 63257 + }, + { + "epoch": 3.5042406015037595, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.0057, + "step": 63258 + }, + { + "epoch": 3.504300751879699, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 63259 + }, + { + "epoch": 3.5043609022556392, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.06, + "step": 63260 + }, + { + "epoch": 3.504421052631579, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9845, + "step": 63261 + }, + { + "epoch": 3.5044812030075185, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 63262 + }, + { + "epoch": 3.5045413533834586, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9482, + "step": 63263 + }, + { + "epoch": 3.5046015037593987, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0152, + "step": 63264 + }, + { + "epoch": 3.5046616541353384, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 63265 + }, + { + "epoch": 3.504721804511278, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 63266 + }, + { + "epoch": 3.504781954887218, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 63267 + }, + { + "epoch": 3.5048421052631578, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 63268 + }, + { + "epoch": 3.504902255639098, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1321, + "step": 63269 + }, + { + "epoch": 3.5049624060150375, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0432, + "step": 63270 + }, + { + "epoch": 3.5050225563909776, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.999, + "step": 63271 + }, + { + "epoch": 3.5050827067669172, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 63272 + }, + { + "epoch": 3.5051428571428573, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0637, + "step": 63273 + }, + { + "epoch": 3.505203007518797, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 63274 + }, + { + "epoch": 3.5052631578947366, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9741, + "step": 63275 + }, + { + "epoch": 3.5053233082706767, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.1427, + "step": 63276 + }, + { + "epoch": 3.5053834586466164, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.964, + "step": 63277 + }, + { + "epoch": 3.5054436090225565, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 1.0912, + "step": 63278 + }, + { + "epoch": 3.505503759398496, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9391, + "step": 63279 + }, + { + "epoch": 3.505563909774436, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.1041, + "step": 63280 + }, + { + "epoch": 3.505624060150376, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9569, + "step": 63281 + }, + { + "epoch": 3.505684210526316, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0007, + "step": 63282 + }, + { + "epoch": 3.5057443609022556, + "grad_norm": 1.5078125, + "learning_rate": 3e-05, + "loss": 1.1209, + "step": 63283 + }, + { + "epoch": 3.5058045112781953, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0132, + "step": 63284 + }, + { + "epoch": 3.5058646616541354, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.8979, + "step": 63285 + }, + { + "epoch": 3.5059248120300754, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0428, + "step": 63286 + }, + { + "epoch": 3.505984962406015, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9966, + "step": 63287 + }, + { + "epoch": 3.5060451127819547, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0716, + "step": 63288 + }, + { + "epoch": 3.506105263157895, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.0205, + "step": 63289 + }, + { + "epoch": 3.5061654135338345, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 63290 + }, + { + "epoch": 3.5062255639097746, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9086, + "step": 63291 + }, + { + "epoch": 3.5062857142857142, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.008, + "step": 63292 + }, + { + "epoch": 3.506345864661654, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 63293 + }, + { + "epoch": 3.506406015037594, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0189, + "step": 63294 + }, + { + "epoch": 3.506466165413534, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0623, + "step": 63295 + }, + { + "epoch": 3.5065263157894737, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0574, + "step": 63296 + }, + { + "epoch": 3.5065864661654134, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.083, + "step": 63297 + }, + { + "epoch": 3.5066466165413535, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9765, + "step": 63298 + }, + { + "epoch": 3.506706766917293, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 63299 + }, + { + "epoch": 3.506766917293233, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 63300 + }, + { + "epoch": 3.506827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0906, + "step": 63301 + }, + { + "epoch": 3.506887218045113, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 1.0355, + "step": 63302 + }, + { + "epoch": 3.5069473684210526, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.044, + "step": 63303 + }, + { + "epoch": 3.5070075187969927, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9938, + "step": 63304 + }, + { + "epoch": 3.5070676691729323, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0166, + "step": 63305 + }, + { + "epoch": 3.507127819548872, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0486, + "step": 63306 + }, + { + "epoch": 3.507187969924812, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9989, + "step": 63307 + }, + { + "epoch": 3.5072481203007517, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0176, + "step": 63308 + }, + { + "epoch": 3.507308270676692, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.063, + "step": 63309 + }, + { + "epoch": 3.5073684210526315, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 63310 + }, + { + "epoch": 3.5074285714285716, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9469, + "step": 63311 + }, + { + "epoch": 3.507488721804511, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9496, + "step": 63312 + }, + { + "epoch": 3.5075488721804513, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.087, + "step": 63313 + }, + { + "epoch": 3.507609022556391, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9572, + "step": 63314 + }, + { + "epoch": 3.5076691729323306, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 63315 + }, + { + "epoch": 3.5077293233082707, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9811, + "step": 63316 + }, + { + "epoch": 3.507789473684211, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0212, + "step": 63317 + }, + { + "epoch": 3.5078496240601504, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9754, + "step": 63318 + }, + { + "epoch": 3.50790977443609, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0876, + "step": 63319 + }, + { + "epoch": 3.50796992481203, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9654, + "step": 63320 + }, + { + "epoch": 3.50803007518797, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0175, + "step": 63321 + }, + { + "epoch": 3.50809022556391, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9567, + "step": 63322 + }, + { + "epoch": 3.5081503759398496, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0587, + "step": 63323 + }, + { + "epoch": 3.5082105263157892, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 63324 + }, + { + "epoch": 3.5082706766917293, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9221, + "step": 63325 + }, + { + "epoch": 3.5083308270676694, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 63326 + }, + { + "epoch": 3.508390977443609, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9946, + "step": 63327 + }, + { + "epoch": 3.5084511278195487, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1253, + "step": 63328 + }, + { + "epoch": 3.508511278195489, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9766, + "step": 63329 + }, + { + "epoch": 3.5085714285714285, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0895, + "step": 63330 + }, + { + "epoch": 3.5086315789473685, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9928, + "step": 63331 + }, + { + "epoch": 3.508691729323308, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9316, + "step": 63332 + }, + { + "epoch": 3.508751879699248, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 63333 + }, + { + "epoch": 3.508812030075188, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9528, + "step": 63334 + }, + { + "epoch": 3.508872180451128, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 63335 + }, + { + "epoch": 3.5089323308270677, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0743, + "step": 63336 + }, + { + "epoch": 3.5089924812030073, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0201, + "step": 63337 + }, + { + "epoch": 3.5090526315789474, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.003, + "step": 63338 + }, + { + "epoch": 3.509112781954887, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.1053, + "step": 63339 + }, + { + "epoch": 3.509172932330827, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.934, + "step": 63340 + }, + { + "epoch": 3.509233082706767, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0332, + "step": 63341 + }, + { + "epoch": 3.509293233082707, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0107, + "step": 63342 + }, + { + "epoch": 3.5093533834586466, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 63343 + }, + { + "epoch": 3.5094135338345867, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 63344 + }, + { + "epoch": 3.5094736842105263, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 63345 + }, + { + "epoch": 3.509533834586466, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9806, + "step": 63346 + }, + { + "epoch": 3.509593984962406, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 63347 + }, + { + "epoch": 3.509654135338346, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0399, + "step": 63348 + }, + { + "epoch": 3.509714285714286, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0556, + "step": 63349 + }, + { + "epoch": 3.5097744360902254, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0458, + "step": 63350 + }, + { + "epoch": 3.5098345864661655, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.0122, + "step": 63351 + }, + { + "epoch": 3.509894736842105, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0791, + "step": 63352 + }, + { + "epoch": 3.5099548872180453, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.1167, + "step": 63353 + }, + { + "epoch": 3.510015037593985, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0294, + "step": 63354 + }, + { + "epoch": 3.5100751879699246, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.097, + "step": 63355 + }, + { + "epoch": 3.5101353383458647, + "grad_norm": 0.9140625, + "learning_rate": 3e-05, + "loss": 0.9892, + "step": 63356 + }, + { + "epoch": 3.5101954887218048, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0146, + "step": 63357 + }, + { + "epoch": 3.5102556390977444, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 63358 + }, + { + "epoch": 3.510315789473684, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.957, + "step": 63359 + }, + { + "epoch": 3.510375939849624, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 63360 + }, + { + "epoch": 3.510436090225564, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.8489, + "step": 63361 + }, + { + "epoch": 3.510496240601504, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0798, + "step": 63362 + }, + { + "epoch": 3.5105563909774435, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9849, + "step": 63363 + }, + { + "epoch": 3.510616541353383, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.1987, + "step": 63364 + }, + { + "epoch": 3.5106766917293233, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9277, + "step": 63365 + }, + { + "epoch": 3.5107368421052634, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 63366 + }, + { + "epoch": 3.510796992481203, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 0.9914, + "step": 63367 + }, + { + "epoch": 3.5108571428571427, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9277, + "step": 63368 + }, + { + "epoch": 3.5109172932330828, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.9038, + "step": 63369 + }, + { + "epoch": 3.5109774436090224, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9024, + "step": 63370 + }, + { + "epoch": 3.5110375939849625, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0446, + "step": 63371 + }, + { + "epoch": 3.511097744360902, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9864, + "step": 63372 + }, + { + "epoch": 3.5111578947368423, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 63373 + }, + { + "epoch": 3.511218045112782, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.0767, + "step": 63374 + }, + { + "epoch": 3.511278195488722, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 63375 + }, + { + "epoch": 3.5113383458646616, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0337, + "step": 63376 + }, + { + "epoch": 3.5113984962406013, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0529, + "step": 63377 + }, + { + "epoch": 3.5114586466165414, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9066, + "step": 63378 + }, + { + "epoch": 3.511518796992481, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9467, + "step": 63379 + }, + { + "epoch": 3.511578947368421, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9547, + "step": 63380 + }, + { + "epoch": 3.511639097744361, + "grad_norm": 1.125, + "learning_rate": 3e-05, + "loss": 1.1758, + "step": 63381 + }, + { + "epoch": 3.511699248120301, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.0265, + "step": 63382 + }, + { + "epoch": 3.5117593984962405, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1017, + "step": 63383 + }, + { + "epoch": 3.5118195488721806, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0014, + "step": 63384 + }, + { + "epoch": 3.5118796992481203, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9731, + "step": 63385 + }, + { + "epoch": 3.51193984962406, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 63386 + }, + { + "epoch": 3.512, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9663, + "step": 63387 + }, + { + "epoch": 3.51206015037594, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9759, + "step": 63388 + }, + { + "epoch": 3.5121203007518798, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 1.0558, + "step": 63389 + }, + { + "epoch": 3.5121804511278194, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.9638, + "step": 63390 + }, + { + "epoch": 3.5122406015037595, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0986, + "step": 63391 + }, + { + "epoch": 3.512300751879699, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0023, + "step": 63392 + }, + { + "epoch": 3.5123609022556392, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9355, + "step": 63393 + }, + { + "epoch": 3.512421052631579, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0203, + "step": 63394 + }, + { + "epoch": 3.5124812030075185, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0594, + "step": 63395 + }, + { + "epoch": 3.5125413533834586, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.9618, + "step": 63396 + }, + { + "epoch": 3.5126015037593987, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9356, + "step": 63397 + }, + { + "epoch": 3.5126616541353384, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.9747, + "step": 63398 + }, + { + "epoch": 3.512721804511278, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.248, + "step": 63399 + }, + { + "epoch": 3.512781954887218, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0395, + "step": 63400 + }, + { + "epoch": 3.5128421052631578, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9422, + "step": 63401 + }, + { + "epoch": 3.512902255639098, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 63402 + }, + { + "epoch": 3.5129624060150375, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.0027, + "step": 63403 + }, + { + "epoch": 3.5130225563909776, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9991, + "step": 63404 + }, + { + "epoch": 3.5130827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0041, + "step": 63405 + }, + { + "epoch": 3.5131428571428573, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 0.984, + "step": 63406 + }, + { + "epoch": 3.513203007518797, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0513, + "step": 63407 + }, + { + "epoch": 3.5132631578947366, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9611, + "step": 63408 + }, + { + "epoch": 3.5133233082706767, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.8868, + "step": 63409 + }, + { + "epoch": 3.5133834586466164, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0187, + "step": 63410 + }, + { + "epoch": 3.5134436090225565, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9077, + "step": 63411 + }, + { + "epoch": 3.513503759398496, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 1.1111, + "step": 63412 + }, + { + "epoch": 3.5135639097744362, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0658, + "step": 63413 + }, + { + "epoch": 3.513624060150376, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 63414 + }, + { + "epoch": 3.513684210526316, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9021, + "step": 63415 + }, + { + "epoch": 3.5137443609022556, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 63416 + }, + { + "epoch": 3.5138045112781953, + "grad_norm": 0.546875, + "learning_rate": 3e-05, + "loss": 0.9653, + "step": 63417 + }, + { + "epoch": 3.5138646616541354, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9781, + "step": 63418 + }, + { + "epoch": 3.5139248120300755, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 63419 + }, + { + "epoch": 3.513984962406015, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 63420 + }, + { + "epoch": 3.5140451127819547, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 0.96, + "step": 63421 + }, + { + "epoch": 3.514105263157895, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 63422 + }, + { + "epoch": 3.5141654135338345, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9597, + "step": 63423 + }, + { + "epoch": 3.5142255639097746, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9397, + "step": 63424 + }, + { + "epoch": 3.5142857142857142, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9307, + "step": 63425 + }, + { + "epoch": 3.514345864661654, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.944, + "step": 63426 + }, + { + "epoch": 3.514406015037594, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0408, + "step": 63427 + }, + { + "epoch": 3.514466165413534, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 63428 + }, + { + "epoch": 3.5145263157894737, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 63429 + }, + { + "epoch": 3.5145864661654134, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 0.9671, + "step": 63430 + }, + { + "epoch": 3.5146466165413535, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.026, + "step": 63431 + }, + { + "epoch": 3.514706766917293, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 0.9826, + "step": 63432 + }, + { + "epoch": 3.514766917293233, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9673, + "step": 63433 + }, + { + "epoch": 3.514827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9814, + "step": 63434 + }, + { + "epoch": 3.5148872180451125, + "grad_norm": 1.578125, + "learning_rate": 3e-05, + "loss": 0.971, + "step": 63435 + }, + { + "epoch": 3.5149473684210526, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.867, + "step": 63436 + }, + { + "epoch": 3.5150075187969927, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0417, + "step": 63437 + }, + { + "epoch": 3.5150676691729323, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 63438 + }, + { + "epoch": 3.515127819548872, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0331, + "step": 63439 + }, + { + "epoch": 3.515187969924812, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9712, + "step": 63440 + }, + { + "epoch": 3.5152481203007517, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9711, + "step": 63441 + }, + { + "epoch": 3.515308270676692, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0384, + "step": 63442 + }, + { + "epoch": 3.5153684210526315, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0254, + "step": 63443 + }, + { + "epoch": 3.5154285714285716, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0485, + "step": 63444 + }, + { + "epoch": 3.515488721804511, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9769, + "step": 63445 + }, + { + "epoch": 3.5155488721804513, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0154, + "step": 63446 + }, + { + "epoch": 3.515609022556391, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0576, + "step": 63447 + }, + { + "epoch": 3.5156691729323306, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.082, + "step": 63448 + }, + { + "epoch": 3.5157293233082707, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 63449 + }, + { + "epoch": 3.515789473684211, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9973, + "step": 63450 + }, + { + "epoch": 3.5158496240601504, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.1201, + "step": 63451 + }, + { + "epoch": 3.51590977443609, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 63452 + }, + { + "epoch": 3.51596992481203, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0981, + "step": 63453 + }, + { + "epoch": 3.51603007518797, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.015, + "step": 63454 + }, + { + "epoch": 3.51609022556391, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.128, + "step": 63455 + }, + { + "epoch": 3.5161503759398496, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 63456 + }, + { + "epoch": 3.5162105263157892, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9285, + "step": 63457 + }, + { + "epoch": 3.5162706766917293, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0139, + "step": 63458 + }, + { + "epoch": 3.5163308270676694, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9903, + "step": 63459 + }, + { + "epoch": 3.516390977443609, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9999, + "step": 63460 + }, + { + "epoch": 3.5164511278195487, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.0287, + "step": 63461 + }, + { + "epoch": 3.516511278195489, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 63462 + }, + { + "epoch": 3.5165714285714285, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9643, + "step": 63463 + }, + { + "epoch": 3.5166315789473686, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9716, + "step": 63464 + }, + { + "epoch": 3.516691729323308, + "grad_norm": 0.98828125, + "learning_rate": 3e-05, + "loss": 0.9788, + "step": 63465 + }, + { + "epoch": 3.516751879699248, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9869, + "step": 63466 + }, + { + "epoch": 3.516812030075188, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.045, + "step": 63467 + }, + { + "epoch": 3.516872180451128, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9557, + "step": 63468 + }, + { + "epoch": 3.5169323308270677, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 63469 + }, + { + "epoch": 3.5169924812030073, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9428, + "step": 63470 + }, + { + "epoch": 3.5170526315789474, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.0477, + "step": 63471 + }, + { + "epoch": 3.517112781954887, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 63472 + }, + { + "epoch": 3.517172932330827, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9867, + "step": 63473 + }, + { + "epoch": 3.517233082706767, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.983, + "step": 63474 + }, + { + "epoch": 3.517293233082707, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0381, + "step": 63475 + }, + { + "epoch": 3.5173533834586466, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0382, + "step": 63476 + }, + { + "epoch": 3.5174135338345867, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9405, + "step": 63477 + }, + { + "epoch": 3.5174736842105263, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 63478 + }, + { + "epoch": 3.517533834586466, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 0.9633, + "step": 63479 + }, + { + "epoch": 3.517593984962406, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 63480 + }, + { + "epoch": 3.5176541353383457, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 63481 + }, + { + "epoch": 3.517714285714286, + "grad_norm": 0.53125, + "learning_rate": 3e-05, + "loss": 0.9317, + "step": 63482 + }, + { + "epoch": 3.5177744360902254, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9866, + "step": 63483 + }, + { + "epoch": 3.5178345864661655, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 63484 + }, + { + "epoch": 3.517894736842105, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0222, + "step": 63485 + }, + { + "epoch": 3.5179548872180453, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.1173, + "step": 63486 + }, + { + "epoch": 3.518015037593985, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 63487 + }, + { + "epoch": 3.5180751879699246, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 63488 + }, + { + "epoch": 3.5181353383458647, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0171, + "step": 63489 + }, + { + "epoch": 3.5181954887218048, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 63490 + }, + { + "epoch": 3.5182556390977444, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0286, + "step": 63491 + }, + { + "epoch": 3.518315789473684, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.009, + "step": 63492 + }, + { + "epoch": 3.518375939849624, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9803, + "step": 63493 + }, + { + "epoch": 3.518436090225564, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 63494 + }, + { + "epoch": 3.518496240601504, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 63495 + }, + { + "epoch": 3.5185563909774435, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.076, + "step": 63496 + }, + { + "epoch": 3.518616541353383, + "grad_norm": 0.92578125, + "learning_rate": 3e-05, + "loss": 1.1675, + "step": 63497 + }, + { + "epoch": 3.5186766917293233, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.946, + "step": 63498 + }, + { + "epoch": 3.5187368421052634, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0936, + "step": 63499 + }, + { + "epoch": 3.518796992481203, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 63500 + }, + { + "epoch": 3.5188571428571427, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 0.9853, + "step": 63501 + }, + { + "epoch": 3.5189172932330828, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 63502 + }, + { + "epoch": 3.5189774436090224, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9377, + "step": 63503 + }, + { + "epoch": 3.5190375939849625, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0101, + "step": 63504 + }, + { + "epoch": 3.519097744360902, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0, + "step": 63505 + }, + { + "epoch": 3.5191578947368423, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0683, + "step": 63506 + }, + { + "epoch": 3.519218045112782, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 63507 + }, + { + "epoch": 3.519278195488722, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 63508 + }, + { + "epoch": 3.5193383458646617, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0266, + "step": 63509 + }, + { + "epoch": 3.5193984962406013, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9708, + "step": 63510 + }, + { + "epoch": 3.5194586466165414, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 1.0434, + "step": 63511 + }, + { + "epoch": 3.519518796992481, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9566, + "step": 63512 + }, + { + "epoch": 3.519578947368421, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0845, + "step": 63513 + }, + { + "epoch": 3.519639097744361, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0307, + "step": 63514 + }, + { + "epoch": 3.519699248120301, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0151, + "step": 63515 + }, + { + "epoch": 3.5197593984962405, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 63516 + }, + { + "epoch": 3.5198195488721806, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9878, + "step": 63517 + }, + { + "epoch": 3.5198796992481203, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 63518 + }, + { + "epoch": 3.51993984962406, + "grad_norm": 1.3125, + "learning_rate": 3e-05, + "loss": 0.9504, + "step": 63519 + }, + { + "epoch": 3.52, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 63520 + }, + { + "epoch": 3.52006015037594, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9227, + "step": 63521 + }, + { + "epoch": 3.5201203007518798, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9927, + "step": 63522 + }, + { + "epoch": 3.5201804511278194, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 63523 + }, + { + "epoch": 3.5202406015037595, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.992, + "step": 63524 + }, + { + "epoch": 3.520300751879699, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9642, + "step": 63525 + }, + { + "epoch": 3.5203609022556392, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0401, + "step": 63526 + }, + { + "epoch": 3.520421052631579, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9917, + "step": 63527 + }, + { + "epoch": 3.5204812030075185, + "grad_norm": 0.56640625, + "learning_rate": 3e-05, + "loss": 0.9699, + "step": 63528 + }, + { + "epoch": 3.5205413533834586, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0099, + "step": 63529 + }, + { + "epoch": 3.5206015037593987, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.123, + "step": 63530 + }, + { + "epoch": 3.5206616541353384, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0336, + "step": 63531 + }, + { + "epoch": 3.520721804511278, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9657, + "step": 63532 + }, + { + "epoch": 3.520781954887218, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0503, + "step": 63533 + }, + { + "epoch": 3.5208421052631578, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0843, + "step": 63534 + }, + { + "epoch": 3.520902255639098, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 63535 + }, + { + "epoch": 3.5209624060150375, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9476, + "step": 63536 + }, + { + "epoch": 3.5210225563909776, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0299, + "step": 63537 + }, + { + "epoch": 3.5210827067669173, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0086, + "step": 63538 + }, + { + "epoch": 3.5211428571428574, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0169, + "step": 63539 + }, + { + "epoch": 3.521203007518797, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0022, + "step": 63540 + }, + { + "epoch": 3.5212631578947367, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 63541 + }, + { + "epoch": 3.5213233082706767, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0823, + "step": 63542 + }, + { + "epoch": 3.5213834586466164, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0033, + "step": 63543 + }, + { + "epoch": 3.5214436090225565, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.0048, + "step": 63544 + }, + { + "epoch": 3.521503759398496, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 1.0258, + "step": 63545 + }, + { + "epoch": 3.5215639097744362, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9352, + "step": 63546 + }, + { + "epoch": 3.521624060150376, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 63547 + }, + { + "epoch": 3.521684210526316, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0124, + "step": 63548 + }, + { + "epoch": 3.5217443609022556, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0457, + "step": 63549 + }, + { + "epoch": 3.5218045112781953, + "grad_norm": 0.57421875, + "learning_rate": 3e-05, + "loss": 0.988, + "step": 63550 + }, + { + "epoch": 3.5218646616541354, + "grad_norm": 0.88671875, + "learning_rate": 3e-05, + "loss": 1.0898, + "step": 63551 + }, + { + "epoch": 3.5219248120300755, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9603, + "step": 63552 + }, + { + "epoch": 3.521984962406015, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0694, + "step": 63553 + }, + { + "epoch": 3.5220451127819548, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 63554 + }, + { + "epoch": 3.522105263157895, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9775, + "step": 63555 + }, + { + "epoch": 3.5221654135338345, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.1727, + "step": 63556 + }, + { + "epoch": 3.5222255639097746, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 63557 + }, + { + "epoch": 3.5222857142857142, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9534, + "step": 63558 + }, + { + "epoch": 3.522345864661654, + "grad_norm": 0.98046875, + "learning_rate": 3e-05, + "loss": 0.9487, + "step": 63559 + }, + { + "epoch": 3.522406015037594, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 63560 + }, + { + "epoch": 3.522466165413534, + "grad_norm": 0.515625, + "learning_rate": 3e-05, + "loss": 0.8838, + "step": 63561 + }, + { + "epoch": 3.5225263157894737, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 63562 + }, + { + "epoch": 3.5225864661654134, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 1.0761, + "step": 63563 + }, + { + "epoch": 3.5226466165413535, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0386, + "step": 63564 + }, + { + "epoch": 3.522706766917293, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.048, + "step": 63565 + }, + { + "epoch": 3.522766917293233, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.8833, + "step": 63566 + }, + { + "epoch": 3.522827067669173, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0666, + "step": 63567 + }, + { + "epoch": 3.5228872180451125, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9895, + "step": 63568 + }, + { + "epoch": 3.5229473684210526, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0549, + "step": 63569 + }, + { + "epoch": 3.5230075187969927, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 63570 + }, + { + "epoch": 3.5230676691729323, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.9848, + "step": 63571 + }, + { + "epoch": 3.523127819548872, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9625, + "step": 63572 + }, + { + "epoch": 3.523187969924812, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9246, + "step": 63573 + }, + { + "epoch": 3.5232481203007517, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0388, + "step": 63574 + }, + { + "epoch": 3.523308270676692, + "grad_norm": 1.28125, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 63575 + }, + { + "epoch": 3.5233684210526315, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 63576 + }, + { + "epoch": 3.5234285714285716, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 63577 + }, + { + "epoch": 3.5234887218045112, + "grad_norm": 0.99609375, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 63578 + }, + { + "epoch": 3.5235488721804513, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.9942, + "step": 63579 + }, + { + "epoch": 3.523609022556391, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.9599, + "step": 63580 + }, + { + "epoch": 3.5236691729323306, + "grad_norm": 0.9375, + "learning_rate": 3e-05, + "loss": 1.0145, + "step": 63581 + }, + { + "epoch": 3.5237293233082707, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 0.9889, + "step": 63582 + }, + { + "epoch": 3.5237894736842104, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0396, + "step": 63583 + }, + { + "epoch": 3.5238496240601505, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 63584 + }, + { + "epoch": 3.52390977443609, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0005, + "step": 63585 + }, + { + "epoch": 3.52396992481203, + "grad_norm": 0.62109375, + "learning_rate": 3e-05, + "loss": 0.9344, + "step": 63586 + }, + { + "epoch": 3.52403007518797, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.1034, + "step": 63587 + }, + { + "epoch": 3.52409022556391, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9287, + "step": 63588 + }, + { + "epoch": 3.5241503759398496, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9797, + "step": 63589 + }, + { + "epoch": 3.5242105263157892, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 0.9493, + "step": 63590 + }, + { + "epoch": 3.5242706766917293, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 1.0407, + "step": 63591 + }, + { + "epoch": 3.5243308270676694, + "grad_norm": 0.55078125, + "learning_rate": 3e-05, + "loss": 0.95, + "step": 63592 + }, + { + "epoch": 3.524390977443609, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 0.998, + "step": 63593 + }, + { + "epoch": 3.5244511278195487, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 63594 + }, + { + "epoch": 3.524511278195489, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.8856, + "step": 63595 + }, + { + "epoch": 3.5245714285714285, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.0402, + "step": 63596 + }, + { + "epoch": 3.5246315789473686, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9813, + "step": 63597 + }, + { + "epoch": 3.524691729323308, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 63598 + }, + { + "epoch": 3.524751879699248, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 63599 + }, + { + "epoch": 3.524812030075188, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9834, + "step": 63600 + }, + { + "epoch": 3.524872180451128, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0479, + "step": 63601 + }, + { + "epoch": 3.5249323308270677, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9965, + "step": 63602 + }, + { + "epoch": 3.5249924812030073, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 63603 + }, + { + "epoch": 3.5250526315789474, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 63604 + }, + { + "epoch": 3.525112781954887, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0409, + "step": 63605 + }, + { + "epoch": 3.525172932330827, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 1.1175, + "step": 63606 + }, + { + "epoch": 3.525233082706767, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 1.1229, + "step": 63607 + }, + { + "epoch": 3.525293233082707, + "grad_norm": 0.91015625, + "learning_rate": 3e-05, + "loss": 1.1295, + "step": 63608 + }, + { + "epoch": 3.5253533834586466, + "grad_norm": 1.0546875, + "learning_rate": 3e-05, + "loss": 0.9508, + "step": 63609 + }, + { + "epoch": 3.5254135338345867, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 63610 + }, + { + "epoch": 3.5254736842105263, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9777, + "step": 63611 + }, + { + "epoch": 3.525533834586466, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 63612 + }, + { + "epoch": 3.525593984962406, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 63613 + }, + { + "epoch": 3.5256541353383457, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0948, + "step": 63614 + }, + { + "epoch": 3.525714285714286, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.017, + "step": 63615 + }, + { + "epoch": 3.5257744360902254, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9768, + "step": 63616 + }, + { + "epoch": 3.5258345864661655, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.025, + "step": 63617 + }, + { + "epoch": 3.525894736842105, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0879, + "step": 63618 + }, + { + "epoch": 3.5259548872180453, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0058, + "step": 63619 + }, + { + "epoch": 3.526015037593985, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0416, + "step": 63620 + }, + { + "epoch": 3.5260751879699246, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.1184, + "step": 63621 + }, + { + "epoch": 3.5261353383458647, + "grad_norm": 1.0, + "learning_rate": 3e-05, + "loss": 1.0367, + "step": 63622 + }, + { + "epoch": 3.5261954887218048, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 0.9733, + "step": 63623 + }, + { + "epoch": 3.5262556390977444, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0589, + "step": 63624 + }, + { + "epoch": 3.526315789473684, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 0.9604, + "step": 63625 + }, + { + "epoch": 3.526375939849624, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0016, + "step": 63626 + }, + { + "epoch": 3.526436090225564, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 63627 + }, + { + "epoch": 3.526496240601504, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9108, + "step": 63628 + }, + { + "epoch": 3.5265563909774436, + "grad_norm": 0.578125, + "learning_rate": 3e-05, + "loss": 0.9901, + "step": 63629 + }, + { + "epoch": 3.526616541353383, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.018, + "step": 63630 + }, + { + "epoch": 3.5266766917293233, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9899, + "step": 63631 + }, + { + "epoch": 3.5267368421052634, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0251, + "step": 63632 + }, + { + "epoch": 3.526796992481203, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9564, + "step": 63633 + }, + { + "epoch": 3.5268571428571427, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.1104, + "step": 63634 + }, + { + "epoch": 3.526917293233083, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.068, + "step": 63635 + }, + { + "epoch": 3.5269774436090224, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9996, + "step": 63636 + }, + { + "epoch": 3.5270375939849625, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9722, + "step": 63637 + }, + { + "epoch": 3.527097744360902, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9249, + "step": 63638 + }, + { + "epoch": 3.5271578947368423, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 0.9907, + "step": 63639 + }, + { + "epoch": 3.527218045112782, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 63640 + }, + { + "epoch": 3.527278195488722, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9985, + "step": 63641 + }, + { + "epoch": 3.5273383458646617, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 0.973, + "step": 63642 + }, + { + "epoch": 3.5273984962406013, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 1.0373, + "step": 63643 + }, + { + "epoch": 3.5274586466165414, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9898, + "step": 63644 + }, + { + "epoch": 3.527518796992481, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9575, + "step": 63645 + }, + { + "epoch": 3.527578947368421, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 63646 + }, + { + "epoch": 3.527639097744361, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0135, + "step": 63647 + }, + { + "epoch": 3.527699248120301, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0784, + "step": 63648 + }, + { + "epoch": 3.5277593984962405, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.0915, + "step": 63649 + }, + { + "epoch": 3.5278195488721806, + "grad_norm": 1.0234375, + "learning_rate": 3e-05, + "loss": 0.9605, + "step": 63650 + }, + { + "epoch": 3.5278796992481203, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.9191, + "step": 63651 + }, + { + "epoch": 3.52793984962406, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9646, + "step": 63652 + }, + { + "epoch": 3.528, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9844, + "step": 63653 + }, + { + "epoch": 3.52806015037594, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 0.966, + "step": 63654 + }, + { + "epoch": 3.5281203007518798, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 63655 + }, + { + "epoch": 3.5281804511278194, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0621, + "step": 63656 + }, + { + "epoch": 3.5282406015037595, + "grad_norm": 0.9296875, + "learning_rate": 3e-05, + "loss": 0.9558, + "step": 63657 + }, + { + "epoch": 3.528300751879699, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.043, + "step": 63658 + }, + { + "epoch": 3.5283609022556393, + "grad_norm": 0.7421875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 63659 + }, + { + "epoch": 3.528421052631579, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.0842, + "step": 63660 + }, + { + "epoch": 3.5284812030075186, + "grad_norm": 0.921875, + "learning_rate": 3e-05, + "loss": 1.038, + "step": 63661 + }, + { + "epoch": 3.5285413533834586, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.972, + "step": 63662 + }, + { + "epoch": 3.5286015037593987, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9336, + "step": 63663 + }, + { + "epoch": 3.5286616541353384, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.957, + "step": 63664 + }, + { + "epoch": 3.528721804511278, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.1777, + "step": 63665 + }, + { + "epoch": 3.528781954887218, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 0.9884, + "step": 63666 + }, + { + "epoch": 3.528842105263158, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.918, + "step": 63667 + }, + { + "epoch": 3.528902255639098, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 1.0604, + "step": 63668 + }, + { + "epoch": 3.5289624060150375, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 63669 + }, + { + "epoch": 3.529022556390977, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.942, + "step": 63670 + }, + { + "epoch": 3.5290827067669173, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.1336, + "step": 63671 + }, + { + "epoch": 3.5291428571428574, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.921, + "step": 63672 + }, + { + "epoch": 3.529203007518797, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0345, + "step": 63673 + }, + { + "epoch": 3.5292631578947367, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 1.0983, + "step": 63674 + }, + { + "epoch": 3.5293233082706768, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0452, + "step": 63675 + }, + { + "epoch": 3.5293834586466164, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9407, + "step": 63676 + }, + { + "epoch": 3.5294436090225565, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0508, + "step": 63677 + }, + { + "epoch": 3.529503759398496, + "grad_norm": 0.53515625, + "learning_rate": 3e-05, + "loss": 1.0553, + "step": 63678 + }, + { + "epoch": 3.5295639097744362, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0268, + "step": 63679 + }, + { + "epoch": 3.529624060150376, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 0.8999, + "step": 63680 + }, + { + "epoch": 3.529684210526316, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.1344, + "step": 63681 + }, + { + "epoch": 3.5297443609022556, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 1.011, + "step": 63682 + }, + { + "epoch": 3.5298045112781953, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 63683 + }, + { + "epoch": 3.5298646616541354, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.02, + "step": 63684 + }, + { + "epoch": 3.5299248120300755, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9165, + "step": 63685 + }, + { + "epoch": 3.529984962406015, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9651, + "step": 63686 + }, + { + "epoch": 3.5300451127819548, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 63687 + }, + { + "epoch": 3.530105263157895, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 1.0231, + "step": 63688 + }, + { + "epoch": 3.5301654135338345, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0659, + "step": 63689 + }, + { + "epoch": 3.5302255639097746, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0982, + "step": 63690 + }, + { + "epoch": 3.5302857142857142, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.0325, + "step": 63691 + }, + { + "epoch": 3.530345864661654, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0365, + "step": 63692 + }, + { + "epoch": 3.530406015037594, + "grad_norm": 0.91796875, + "learning_rate": 3e-05, + "loss": 1.0024, + "step": 63693 + }, + { + "epoch": 3.530466165413534, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9062, + "step": 63694 + }, + { + "epoch": 3.5305263157894737, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9235, + "step": 63695 + }, + { + "epoch": 3.5305864661654134, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0611, + "step": 63696 + }, + { + "epoch": 3.5306466165413535, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0403, + "step": 63697 + }, + { + "epoch": 3.530706766917293, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.9921, + "step": 63698 + }, + { + "epoch": 3.530766917293233, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0011, + "step": 63699 + }, + { + "epoch": 3.530827067669173, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0368, + "step": 63700 + }, + { + "epoch": 3.5308872180451125, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 63701 + }, + { + "epoch": 3.5309473684210526, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0031, + "step": 63702 + }, + { + "epoch": 3.5310075187969927, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0924, + "step": 63703 + }, + { + "epoch": 3.5310676691729324, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.9838, + "step": 63704 + }, + { + "epoch": 3.531127819548872, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0444, + "step": 63705 + }, + { + "epoch": 3.531187969924812, + "grad_norm": 0.875, + "learning_rate": 3e-05, + "loss": 1.107, + "step": 63706 + }, + { + "epoch": 3.5312481203007517, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0003, + "step": 63707 + }, + { + "epoch": 3.531308270676692, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0293, + "step": 63708 + }, + { + "epoch": 3.5313684210526315, + "grad_norm": 0.90625, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 63709 + }, + { + "epoch": 3.5314285714285716, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 63710 + }, + { + "epoch": 3.5314887218045112, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 1.0164, + "step": 63711 + }, + { + "epoch": 3.5315488721804513, + "grad_norm": 0.765625, + "learning_rate": 3e-05, + "loss": 0.959, + "step": 63712 + }, + { + "epoch": 3.531609022556391, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0277, + "step": 63713 + }, + { + "epoch": 3.5316691729323306, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 1.0533, + "step": 63714 + }, + { + "epoch": 3.5317293233082707, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9676, + "step": 63715 + }, + { + "epoch": 3.5317894736842104, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.1144, + "step": 63716 + }, + { + "epoch": 3.5318496240601505, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 63717 + }, + { + "epoch": 3.53190977443609, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0813, + "step": 63718 + }, + { + "epoch": 3.53196992481203, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 63719 + }, + { + "epoch": 3.53203007518797, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 63720 + }, + { + "epoch": 3.53209022556391, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0155, + "step": 63721 + }, + { + "epoch": 3.5321503759398496, + "grad_norm": 0.97265625, + "learning_rate": 3e-05, + "loss": 1.0443, + "step": 63722 + }, + { + "epoch": 3.5322105263157892, + "grad_norm": 0.94921875, + "learning_rate": 3e-05, + "loss": 1.0028, + "step": 63723 + }, + { + "epoch": 3.5322706766917293, + "grad_norm": 0.96484375, + "learning_rate": 3e-05, + "loss": 0.9815, + "step": 63724 + }, + { + "epoch": 3.5323308270676694, + "grad_norm": 0.87890625, + "learning_rate": 3e-05, + "loss": 0.954, + "step": 63725 + }, + { + "epoch": 3.532390977443609, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9607, + "step": 63726 + }, + { + "epoch": 3.5324511278195487, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0425, + "step": 63727 + }, + { + "epoch": 3.532511278195489, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9886, + "step": 63728 + }, + { + "epoch": 3.5325714285714285, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.1099, + "step": 63729 + }, + { + "epoch": 3.5326315789473686, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.0685, + "step": 63730 + }, + { + "epoch": 3.532691729323308, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9821, + "step": 63731 + }, + { + "epoch": 3.532751879699248, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 63732 + }, + { + "epoch": 3.532812030075188, + "grad_norm": 0.8125, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 63733 + }, + { + "epoch": 3.532872180451128, + "grad_norm": 0.6015625, + "learning_rate": 3e-05, + "loss": 0.9571, + "step": 63734 + }, + { + "epoch": 3.5329323308270677, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.034, + "step": 63735 + }, + { + "epoch": 3.5329924812030074, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.953, + "step": 63736 + }, + { + "epoch": 3.5330526315789474, + "grad_norm": 0.9453125, + "learning_rate": 3e-05, + "loss": 1.0352, + "step": 63737 + }, + { + "epoch": 3.533112781954887, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9519, + "step": 63738 + }, + { + "epoch": 3.533172932330827, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9114, + "step": 63739 + }, + { + "epoch": 3.533233082706767, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 1.0302, + "step": 63740 + }, + { + "epoch": 3.533293233082707, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9459, + "step": 63741 + }, + { + "epoch": 3.5333533834586466, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 1.0025, + "step": 63742 + }, + { + "epoch": 3.5334135338345867, + "grad_norm": 0.482421875, + "learning_rate": 3e-05, + "loss": 0.9804, + "step": 63743 + }, + { + "epoch": 3.5334736842105263, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 63744 + }, + { + "epoch": 3.533533834586466, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9238, + "step": 63745 + }, + { + "epoch": 3.533593984962406, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 63746 + }, + { + "epoch": 3.5336541353383457, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0578, + "step": 63747 + }, + { + "epoch": 3.533714285714286, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0159, + "step": 63748 + }, + { + "epoch": 3.5337744360902255, + "grad_norm": 0.76953125, + "learning_rate": 3e-05, + "loss": 1.0177, + "step": 63749 + }, + { + "epoch": 3.5338345864661656, + "grad_norm": 0.58984375, + "learning_rate": 3e-05, + "loss": 0.9872, + "step": 63750 + }, + { + "epoch": 3.533894736842105, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 1.0141, + "step": 63751 + }, + { + "epoch": 3.5339548872180453, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0156, + "step": 63752 + }, + { + "epoch": 3.534015037593985, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9592, + "step": 63753 + }, + { + "epoch": 3.5340751879699246, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9933, + "step": 63754 + }, + { + "epoch": 3.5341353383458647, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9393, + "step": 63755 + }, + { + "epoch": 3.534195488721805, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0793, + "step": 63756 + }, + { + "epoch": 3.5342556390977444, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 1.0017, + "step": 63757 + }, + { + "epoch": 3.534315789473684, + "grad_norm": 0.75390625, + "learning_rate": 3e-05, + "loss": 0.974, + "step": 63758 + }, + { + "epoch": 3.534375939849624, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9883, + "step": 63759 + }, + { + "epoch": 3.534436090225564, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0799, + "step": 63760 + }, + { + "epoch": 3.534496240601504, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 0.9512, + "step": 63761 + }, + { + "epoch": 3.5345563909774436, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0779, + "step": 63762 + }, + { + "epoch": 3.534616541353383, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 63763 + }, + { + "epoch": 3.5346766917293233, + "grad_norm": 1.078125, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 63764 + }, + { + "epoch": 3.5347368421052634, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9158, + "step": 63765 + }, + { + "epoch": 3.534796992481203, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9477, + "step": 63766 + }, + { + "epoch": 3.5348571428571427, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9539, + "step": 63767 + }, + { + "epoch": 3.534917293233083, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9461, + "step": 63768 + }, + { + "epoch": 3.5349774436090224, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 63769 + }, + { + "epoch": 3.5350375939849625, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9398, + "step": 63770 + }, + { + "epoch": 3.535097744360902, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0686, + "step": 63771 + }, + { + "epoch": 3.535157894736842, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0081, + "step": 63772 + }, + { + "epoch": 3.535218045112782, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.01, + "step": 63773 + }, + { + "epoch": 3.535278195488722, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0227, + "step": 63774 + }, + { + "epoch": 3.5353383458646617, + "grad_norm": 1.1015625, + "learning_rate": 3e-05, + "loss": 1.0128, + "step": 63775 + }, + { + "epoch": 3.5353984962406013, + "grad_norm": 1.21875, + "learning_rate": 3e-05, + "loss": 1.0698, + "step": 63776 + }, + { + "epoch": 3.5354586466165414, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 63777 + }, + { + "epoch": 3.535518796992481, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.97, + "step": 63778 + }, + { + "epoch": 3.535578947368421, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0763, + "step": 63779 + }, + { + "epoch": 3.535639097744361, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 63780 + }, + { + "epoch": 3.535699248120301, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0045, + "step": 63781 + }, + { + "epoch": 3.5357593984962405, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0042, + "step": 63782 + }, + { + "epoch": 3.5358195488721806, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 63783 + }, + { + "epoch": 3.5358796992481203, + "grad_norm": 0.95703125, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 63784 + }, + { + "epoch": 3.53593984962406, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0526, + "step": 63785 + }, + { + "epoch": 3.536, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0652, + "step": 63786 + }, + { + "epoch": 3.53606015037594, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9452, + "step": 63787 + }, + { + "epoch": 3.5361203007518798, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9994, + "step": 63788 + }, + { + "epoch": 3.5361804511278194, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 1.033, + "step": 63789 + }, + { + "epoch": 3.5362406015037595, + "grad_norm": 0.84765625, + "learning_rate": 3e-05, + "loss": 1.0149, + "step": 63790 + }, + { + "epoch": 3.536300751879699, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.027, + "step": 63791 + }, + { + "epoch": 3.5363609022556393, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 63792 + }, + { + "epoch": 3.536421052631579, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0448, + "step": 63793 + }, + { + "epoch": 3.5364812030075186, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.9515, + "step": 63794 + }, + { + "epoch": 3.5365413533834587, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.064, + "step": 63795 + }, + { + "epoch": 3.5366015037593987, + "grad_norm": 1.8203125, + "learning_rate": 3e-05, + "loss": 0.9787, + "step": 63796 + }, + { + "epoch": 3.5366616541353384, + "grad_norm": 0.5859375, + "learning_rate": 3e-05, + "loss": 0.9859, + "step": 63797 + }, + { + "epoch": 3.536721804511278, + "grad_norm": 0.859375, + "learning_rate": 3e-05, + "loss": 1.0034, + "step": 63798 + }, + { + "epoch": 3.536781954887218, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0456, + "step": 63799 + }, + { + "epoch": 3.536842105263158, + "grad_norm": 0.52734375, + "learning_rate": 3e-05, + "loss": 0.8142, + "step": 63800 + }, + { + "epoch": 3.536902255639098, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0121, + "step": 63801 + }, + { + "epoch": 3.5369624060150375, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.1169, + "step": 63802 + }, + { + "epoch": 3.537022556390977, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 0.979, + "step": 63803 + }, + { + "epoch": 3.5370827067669173, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.975, + "step": 63804 + }, + { + "epoch": 3.5371428571428574, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.0061, + "step": 63805 + }, + { + "epoch": 3.537203007518797, + "grad_norm": 0.5703125, + "learning_rate": 3e-05, + "loss": 0.87, + "step": 63806 + }, + { + "epoch": 3.5372631578947367, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0836, + "step": 63807 + }, + { + "epoch": 3.5373233082706768, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9955, + "step": 63808 + }, + { + "epoch": 3.5373834586466164, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 1.071, + "step": 63809 + }, + { + "epoch": 3.5374436090225565, + "grad_norm": 0.54296875, + "learning_rate": 3e-05, + "loss": 0.9097, + "step": 63810 + }, + { + "epoch": 3.537503759398496, + "grad_norm": 1.046875, + "learning_rate": 3e-05, + "loss": 1.032, + "step": 63811 + }, + { + "epoch": 3.5375639097744362, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 0.9602, + "step": 63812 + }, + { + "epoch": 3.537624060150376, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9939, + "step": 63813 + }, + { + "epoch": 3.537684210526316, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 0.982, + "step": 63814 + }, + { + "epoch": 3.5377443609022556, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0622, + "step": 63815 + }, + { + "epoch": 3.5378045112781953, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9717, + "step": 63816 + }, + { + "epoch": 3.5378646616541354, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9586, + "step": 63817 + }, + { + "epoch": 3.537924812030075, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0343, + "step": 63818 + }, + { + "epoch": 3.537984962406015, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0181, + "step": 63819 + }, + { + "epoch": 3.5380451127819548, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 0.9758, + "step": 63820 + }, + { + "epoch": 3.538105263157895, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9052, + "step": 63821 + }, + { + "epoch": 3.5381654135338345, + "grad_norm": 0.8671875, + "learning_rate": 3e-05, + "loss": 1.1166, + "step": 63822 + }, + { + "epoch": 3.5382255639097746, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.0296, + "step": 63823 + }, + { + "epoch": 3.5382857142857143, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 0.9279, + "step": 63824 + }, + { + "epoch": 3.538345864661654, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0531, + "step": 63825 + }, + { + "epoch": 3.538406015037594, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 0.9798, + "step": 63826 + }, + { + "epoch": 3.538466165413534, + "grad_norm": 0.60546875, + "learning_rate": 3e-05, + "loss": 1.0741, + "step": 63827 + }, + { + "epoch": 3.5385263157894737, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 63828 + }, + { + "epoch": 3.5385864661654134, + "grad_norm": 1.1953125, + "learning_rate": 3e-05, + "loss": 0.9934, + "step": 63829 + }, + { + "epoch": 3.5386466165413535, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0349, + "step": 63830 + }, + { + "epoch": 3.538706766917293, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9967, + "step": 63831 + }, + { + "epoch": 3.5387669172932332, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0235, + "step": 63832 + }, + { + "epoch": 3.538827067669173, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9437, + "step": 63833 + }, + { + "epoch": 3.5388872180451125, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0168, + "step": 63834 + }, + { + "epoch": 3.5389473684210526, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9912, + "step": 63835 + }, + { + "epoch": 3.5390075187969927, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0259, + "step": 63836 + }, + { + "epoch": 3.5390676691729324, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0392, + "step": 63837 + }, + { + "epoch": 3.539127819548872, + "grad_norm": 0.8046875, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 63838 + }, + { + "epoch": 3.539187969924812, + "grad_norm": 0.796875, + "learning_rate": 3e-05, + "loss": 1.0326, + "step": 63839 + }, + { + "epoch": 3.5392481203007518, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9329, + "step": 63840 + }, + { + "epoch": 3.539308270676692, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 1.1055, + "step": 63841 + }, + { + "epoch": 3.5393684210526315, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 0.9615, + "step": 63842 + }, + { + "epoch": 3.5394285714285716, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 1.0522, + "step": 63843 + }, + { + "epoch": 3.5394887218045112, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9817, + "step": 63844 + }, + { + "epoch": 3.5395488721804513, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 1.0216, + "step": 63845 + }, + { + "epoch": 3.539609022556391, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9935, + "step": 63846 + }, + { + "epoch": 3.5396691729323306, + "grad_norm": 0.6953125, + "learning_rate": 3e-05, + "loss": 1.0391, + "step": 63847 + }, + { + "epoch": 3.5397293233082707, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.0109, + "step": 63848 + }, + { + "epoch": 3.5397894736842104, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 0.9983, + "step": 63849 + }, + { + "epoch": 3.5398496240601505, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 0.9686, + "step": 63850 + }, + { + "epoch": 3.53990977443609, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.006, + "step": 63851 + }, + { + "epoch": 3.53996992481203, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.8825, + "step": 63852 + }, + { + "epoch": 3.54003007518797, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 1.042, + "step": 63853 + }, + { + "epoch": 3.54009022556391, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9956, + "step": 63854 + }, + { + "epoch": 3.5401503759398496, + "grad_norm": 0.55859375, + "learning_rate": 3e-05, + "loss": 0.9133, + "step": 63855 + }, + { + "epoch": 3.5402105263157893, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 0.9117, + "step": 63856 + }, + { + "epoch": 3.5402706766917293, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9807, + "step": 63857 + }, + { + "epoch": 3.5403308270676694, + "grad_norm": 0.47265625, + "learning_rate": 3e-05, + "loss": 0.9404, + "step": 63858 + }, + { + "epoch": 3.540390977443609, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 0.9581, + "step": 63859 + }, + { + "epoch": 3.5404511278195487, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 1.0247, + "step": 63860 + }, + { + "epoch": 3.540511278195489, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 1.0348, + "step": 63861 + }, + { + "epoch": 3.5405714285714285, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.925, + "step": 63862 + }, + { + "epoch": 3.5406315789473686, + "grad_norm": 0.63671875, + "learning_rate": 3e-05, + "loss": 0.9636, + "step": 63863 + }, + { + "epoch": 3.5406917293233082, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 63864 + }, + { + "epoch": 3.540751879699248, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9523, + "step": 63865 + }, + { + "epoch": 3.540812030075188, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.958, + "step": 63866 + }, + { + "epoch": 3.540872180451128, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9213, + "step": 63867 + }, + { + "epoch": 3.5409323308270677, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.0536, + "step": 63868 + }, + { + "epoch": 3.5409924812030074, + "grad_norm": 0.75, + "learning_rate": 3e-05, + "loss": 1.0029, + "step": 63869 + }, + { + "epoch": 3.5410526315789475, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0369, + "step": 63870 + }, + { + "epoch": 3.541112781954887, + "grad_norm": 0.89453125, + "learning_rate": 3e-05, + "loss": 0.9752, + "step": 63871 + }, + { + "epoch": 3.541172932330827, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 0.9959, + "step": 63872 + }, + { + "epoch": 3.541233082706767, + "grad_norm": 0.72265625, + "learning_rate": 3e-05, + "loss": 1.0178, + "step": 63873 + }, + { + "epoch": 3.5412932330827065, + "grad_norm": 0.7734375, + "learning_rate": 3e-05, + "loss": 0.9181, + "step": 63874 + }, + { + "epoch": 3.5413533834586466, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 63875 + }, + { + "epoch": 3.5414135338345867, + "grad_norm": 0.83203125, + "learning_rate": 3e-05, + "loss": 0.9724, + "step": 63876 + }, + { + "epoch": 3.5414736842105263, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0561, + "step": 63877 + }, + { + "epoch": 3.541533834586466, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 1.0592, + "step": 63878 + }, + { + "epoch": 3.541593984962406, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 0.9584, + "step": 63879 + }, + { + "epoch": 3.5416541353383457, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 63880 + }, + { + "epoch": 3.541714285714286, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0478, + "step": 63881 + }, + { + "epoch": 3.5417744360902255, + "grad_norm": 0.90234375, + "learning_rate": 3e-05, + "loss": 1.0361, + "step": 63882 + }, + { + "epoch": 3.5418345864661656, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9606, + "step": 63883 + }, + { + "epoch": 3.541894736842105, + "grad_norm": 1.1484375, + "learning_rate": 3e-05, + "loss": 1.2144, + "step": 63884 + }, + { + "epoch": 3.5419548872180453, + "grad_norm": 1.0703125, + "learning_rate": 3e-05, + "loss": 1.055, + "step": 63885 + }, + { + "epoch": 3.542015037593985, + "grad_norm": 0.78125, + "learning_rate": 3e-05, + "loss": 1.0723, + "step": 63886 + }, + { + "epoch": 3.5420751879699246, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 1.0835, + "step": 63887 + }, + { + "epoch": 3.5421353383458647, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.061, + "step": 63888 + }, + { + "epoch": 3.542195488721805, + "grad_norm": 0.8984375, + "learning_rate": 3e-05, + "loss": 1.0234, + "step": 63889 + }, + { + "epoch": 3.5422556390977444, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9458, + "step": 63890 + }, + { + "epoch": 3.542315789473684, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 0.9664, + "step": 63891 + }, + { + "epoch": 3.542375939849624, + "grad_norm": 0.7578125, + "learning_rate": 3e-05, + "loss": 1.0362, + "step": 63892 + }, + { + "epoch": 3.542436090225564, + "grad_norm": 0.74609375, + "learning_rate": 3e-05, + "loss": 0.9809, + "step": 63893 + }, + { + "epoch": 3.542496240601504, + "grad_norm": 0.609375, + "learning_rate": 3e-05, + "loss": 0.9295, + "step": 63894 + }, + { + "epoch": 3.5425563909774436, + "grad_norm": 0.79296875, + "learning_rate": 3e-05, + "loss": 0.9893, + "step": 63895 + }, + { + "epoch": 3.542616541353383, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.8888, + "step": 63896 + }, + { + "epoch": 3.5426766917293233, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 0.9764, + "step": 63897 + }, + { + "epoch": 3.5427368421052634, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9578, + "step": 63898 + }, + { + "epoch": 3.542796992481203, + "grad_norm": 0.6875, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 63899 + }, + { + "epoch": 3.5428571428571427, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0066, + "step": 63900 + }, + { + "epoch": 3.542917293233083, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 63901 + }, + { + "epoch": 3.5429774436090224, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 1.0158, + "step": 63902 + }, + { + "epoch": 3.5430375939849625, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 1.0137, + "step": 63903 + }, + { + "epoch": 3.543097744360902, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.0726, + "step": 63904 + }, + { + "epoch": 3.543157894736842, + "grad_norm": 0.71484375, + "learning_rate": 3e-05, + "loss": 1.0669, + "step": 63905 + }, + { + "epoch": 3.543218045112782, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9977, + "step": 63906 + }, + { + "epoch": 3.543278195488722, + "grad_norm": 0.83984375, + "learning_rate": 3e-05, + "loss": 1.0241, + "step": 63907 + }, + { + "epoch": 3.5433383458646617, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 0.9252, + "step": 63908 + }, + { + "epoch": 3.5433984962406013, + "grad_norm": 0.66015625, + "learning_rate": 3e-05, + "loss": 1.0283, + "step": 63909 + }, + { + "epoch": 3.5434586466165414, + "grad_norm": 0.62890625, + "learning_rate": 3e-05, + "loss": 1.0422, + "step": 63910 + }, + { + "epoch": 3.543518796992481, + "grad_norm": 0.59375, + "learning_rate": 3e-05, + "loss": 1.065, + "step": 63911 + }, + { + "epoch": 3.543578947368421, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.9648, + "step": 63912 + }, + { + "epoch": 3.543639097744361, + "grad_norm": 0.77734375, + "learning_rate": 3e-05, + "loss": 0.9745, + "step": 63913 + }, + { + "epoch": 3.543699248120301, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0291, + "step": 63914 + }, + { + "epoch": 3.5437593984962406, + "grad_norm": 1.03125, + "learning_rate": 3e-05, + "loss": 1.0074, + "step": 63915 + }, + { + "epoch": 3.5438195488721806, + "grad_norm": 0.61328125, + "learning_rate": 3e-05, + "loss": 0.9634, + "step": 63916 + }, + { + "epoch": 3.5438796992481203, + "grad_norm": 0.94140625, + "learning_rate": 3e-05, + "loss": 1.0285, + "step": 63917 + }, + { + "epoch": 3.54393984962406, + "grad_norm": 0.734375, + "learning_rate": 3e-05, + "loss": 1.005, + "step": 63918 + }, + { + "epoch": 3.544, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 0.9561, + "step": 63919 + }, + { + "epoch": 3.5440601503759397, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.8988, + "step": 63920 + }, + { + "epoch": 3.54412030075188, + "grad_norm": 0.671875, + "learning_rate": 3e-05, + "loss": 0.9338, + "step": 63921 + }, + { + "epoch": 3.5441804511278194, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.2988, + "step": 63922 + }, + { + "epoch": 3.5442406015037595, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9861, + "step": 63923 + }, + { + "epoch": 3.544300751879699, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9225, + "step": 63924 + }, + { + "epoch": 3.5443609022556393, + "grad_norm": 1.234375, + "learning_rate": 3e-05, + "loss": 0.9891, + "step": 63925 + }, + { + "epoch": 3.544421052631579, + "grad_norm": 0.890625, + "learning_rate": 3e-05, + "loss": 0.9386, + "step": 63926 + }, + { + "epoch": 3.5444812030075186, + "grad_norm": 0.66796875, + "learning_rate": 3e-05, + "loss": 1.1181, + "step": 63927 + }, + { + "epoch": 3.5445413533834587, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 0.9463, + "step": 63928 + }, + { + "epoch": 3.5446015037593988, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 0.9879, + "step": 63929 + }, + { + "epoch": 3.5446616541353384, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.869, + "step": 63930 + }, + { + "epoch": 3.544721804511278, + "grad_norm": 0.625, + "learning_rate": 3e-05, + "loss": 1.0648, + "step": 63931 + }, + { + "epoch": 3.544781954887218, + "grad_norm": 0.9609375, + "learning_rate": 3e-05, + "loss": 1.1043, + "step": 63932 + }, + { + "epoch": 3.544842105263158, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0695, + "step": 63933 + }, + { + "epoch": 3.544902255639098, + "grad_norm": 0.70703125, + "learning_rate": 3e-05, + "loss": 1.0078, + "step": 63934 + }, + { + "epoch": 3.5449624060150375, + "grad_norm": 1.109375, + "learning_rate": 3e-05, + "loss": 1.0315, + "step": 63935 + }, + { + "epoch": 3.545022556390977, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0606, + "step": 63936 + }, + { + "epoch": 3.5450827067669173, + "grad_norm": 0.80859375, + "learning_rate": 3e-05, + "loss": 1.07, + "step": 63937 + }, + { + "epoch": 3.5451428571428574, + "grad_norm": 0.85546875, + "learning_rate": 3e-05, + "loss": 1.1562, + "step": 63938 + }, + { + "epoch": 3.545203007518797, + "grad_norm": 0.76171875, + "learning_rate": 3e-05, + "loss": 1.0809, + "step": 63939 + }, + { + "epoch": 3.5452631578947367, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0113, + "step": 63940 + }, + { + "epoch": 3.5453233082706768, + "grad_norm": 0.8515625, + "learning_rate": 3e-05, + "loss": 0.9551, + "step": 63941 + }, + { + "epoch": 3.5453834586466164, + "grad_norm": 0.65234375, + "learning_rate": 3e-05, + "loss": 0.9786, + "step": 63942 + }, + { + "epoch": 3.5454436090225565, + "grad_norm": 0.64453125, + "learning_rate": 3e-05, + "loss": 1.0245, + "step": 63943 + }, + { + "epoch": 3.545503759398496, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0068, + "step": 63944 + }, + { + "epoch": 3.5455639097744363, + "grad_norm": 1.0625, + "learning_rate": 3e-05, + "loss": 1.0792, + "step": 63945 + }, + { + "epoch": 3.545624060150376, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9949, + "step": 63946 + }, + { + "epoch": 3.545684210526316, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 0.965, + "step": 63947 + }, + { + "epoch": 3.5457443609022556, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 0.9532, + "step": 63948 + }, + { + "epoch": 3.5458045112781953, + "grad_norm": 0.96875, + "learning_rate": 3e-05, + "loss": 1.0783, + "step": 63949 + }, + { + "epoch": 3.5458646616541354, + "grad_norm": 0.6328125, + "learning_rate": 3e-05, + "loss": 0.9483, + "step": 63950 + }, + { + "epoch": 3.545924812030075, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0491, + "step": 63951 + }, + { + "epoch": 3.545984962406015, + "grad_norm": 0.82421875, + "learning_rate": 3e-05, + "loss": 1.0312, + "step": 63952 + }, + { + "epoch": 3.5460451127819548, + "grad_norm": 0.828125, + "learning_rate": 3e-05, + "loss": 0.9923, + "step": 63953 + }, + { + "epoch": 3.546105263157895, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0684, + "step": 63954 + }, + { + "epoch": 3.5461654135338345, + "grad_norm": 0.6484375, + "learning_rate": 3e-05, + "loss": 0.9623, + "step": 63955 + }, + { + "epoch": 3.5462255639097746, + "grad_norm": 0.65625, + "learning_rate": 3e-05, + "loss": 1.0019, + "step": 63956 + }, + { + "epoch": 3.5462857142857143, + "grad_norm": 0.58203125, + "learning_rate": 3e-05, + "loss": 0.9783, + "step": 63957 + }, + { + "epoch": 3.546345864661654, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 1.0414, + "step": 63958 + }, + { + "epoch": 3.546406015037594, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 0.9563, + "step": 63959 + }, + { + "epoch": 3.546466165413534, + "grad_norm": 0.69140625, + "learning_rate": 3e-05, + "loss": 1.0453, + "step": 63960 + }, + { + "epoch": 3.5465263157894737, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9762, + "step": 63961 + }, + { + "epoch": 3.5465864661654134, + "grad_norm": 0.67578125, + "learning_rate": 3e-05, + "loss": 1.0108, + "step": 63962 + }, + { + "epoch": 3.5466466165413535, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 0.9926, + "step": 63963 + }, + { + "epoch": 3.546706766917293, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9827, + "step": 63964 + }, + { + "epoch": 3.5467669172932332, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 0.9666, + "step": 63965 + }, + { + "epoch": 3.546827067669173, + "grad_norm": 0.71875, + "learning_rate": 3e-05, + "loss": 1.096, + "step": 63966 + }, + { + "epoch": 3.5468872180451125, + "grad_norm": 0.984375, + "learning_rate": 3e-05, + "loss": 0.9795, + "step": 63967 + }, + { + "epoch": 3.5469473684210526, + "grad_norm": 0.86328125, + "learning_rate": 3e-05, + "loss": 1.0242, + "step": 63968 + }, + { + "epoch": 3.5470075187969927, + "grad_norm": 1.375, + "learning_rate": 3e-05, + "loss": 1.0703, + "step": 63969 + }, + { + "epoch": 3.5470676691729324, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 1.0439, + "step": 63970 + }, + { + "epoch": 3.547127819548872, + "grad_norm": 0.73828125, + "learning_rate": 3e-05, + "loss": 0.9519, + "step": 63971 + }, + { + "epoch": 3.547187969924812, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0085, + "step": 63972 + }, + { + "epoch": 3.5472481203007518, + "grad_norm": 0.78515625, + "learning_rate": 3e-05, + "loss": 1.0427, + "step": 63973 + }, + { + "epoch": 3.547308270676692, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 1.0228, + "step": 63974 + }, + { + "epoch": 3.5473684210526315, + "grad_norm": 0.6171875, + "learning_rate": 3e-05, + "loss": 0.9359, + "step": 63975 + }, + { + "epoch": 3.547428571428571, + "grad_norm": 0.87109375, + "learning_rate": 3e-05, + "loss": 0.9832, + "step": 63976 + }, + { + "epoch": 3.5474887218045112, + "grad_norm": 0.703125, + "learning_rate": 3e-05, + "loss": 0.9353, + "step": 63977 + }, + { + "epoch": 3.5475488721804513, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 1.022, + "step": 63978 + }, + { + "epoch": 3.547609022556391, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0051, + "step": 63979 + }, + { + "epoch": 3.5476691729323306, + "grad_norm": 0.6796875, + "learning_rate": 3e-05, + "loss": 1.0102, + "step": 63980 + }, + { + "epoch": 3.5477293233082707, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.9971, + "step": 63981 + }, + { + "epoch": 3.5477894736842104, + "grad_norm": 0.59765625, + "learning_rate": 3e-05, + "loss": 1.0474, + "step": 63982 + }, + { + "epoch": 3.5478496240601505, + "grad_norm": 0.81640625, + "learning_rate": 3e-05, + "loss": 0.987, + "step": 63983 + }, + { + "epoch": 3.54790977443609, + "grad_norm": 0.69921875, + "learning_rate": 3e-05, + "loss": 1.0338, + "step": 63984 + }, + { + "epoch": 3.54796992481203, + "grad_norm": 0.7890625, + "learning_rate": 3e-05, + "loss": 0.943, + "step": 63985 + }, + { + "epoch": 3.54803007518797, + "grad_norm": 0.7109375, + "learning_rate": 3e-05, + "loss": 1.0559, + "step": 63986 + }, + { + "epoch": 3.54809022556391, + "grad_norm": 0.5390625, + "learning_rate": 3e-05, + "loss": 0.9524, + "step": 63987 + }, + { + "epoch": 3.5481503759398496, + "grad_norm": 0.68359375, + "learning_rate": 3e-05, + "loss": 0.9979, + "step": 63988 + }, + { + "epoch": 3.5482105263157893, + "grad_norm": 0.7265625, + "learning_rate": 3e-05, + "loss": 1.0118, + "step": 63989 + }, + { + "epoch": 3.5482706766917294, + "grad_norm": 0.6640625, + "learning_rate": 3e-05, + "loss": 0.931, + "step": 63990 + }, + { + "epoch": 3.5483308270676694, + "grad_norm": 0.640625, + "learning_rate": 3e-05, + "loss": 1.0309, + "step": 63991 + }, + { + "epoch": 3.548390977443609, + "grad_norm": 0.8203125, + "learning_rate": 3e-05, + "loss": 0.9117, + "step": 63992 + }, + { + "epoch": 3.5484511278195487, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 1.0084, + "step": 63993 + }, + { + "epoch": 3.548511278195489, + "grad_norm": 0.8828125, + "learning_rate": 3e-05, + "loss": 0.9237, + "step": 63994 + }, + { + "epoch": 3.5485714285714285, + "grad_norm": 1.0078125, + "learning_rate": 3e-05, + "loss": 1.0246, + "step": 63995 + }, + { + "epoch": 3.5486315789473686, + "grad_norm": 0.84375, + "learning_rate": 3e-05, + "loss": 1.0676, + "step": 63996 + }, + { + "epoch": 3.5486917293233082, + "grad_norm": 0.80078125, + "learning_rate": 3e-05, + "loss": 0.9824, + "step": 63997 + }, + { + "epoch": 3.548751879699248, + "grad_norm": 0.73046875, + "learning_rate": 3e-05, + "loss": 0.9449, + "step": 63998 + }, + { + "epoch": 3.548812030075188, + "grad_norm": 0.8359375, + "learning_rate": 3e-05, + "loss": 0.9455, + "step": 63999 + }, + { + "epoch": 3.548872180451128, + "grad_norm": 1.015625, + "learning_rate": 3e-05, + "loss": 1.0923, + "step": 64000 + } + ], + "logging_steps": 1, + "max_steps": 166250, + "num_input_tokens_seen": 0, + "num_train_epochs": 10, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.2294460011790991e+18, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +}